Tining For Speed

THIIID
'DITloN
......M
,. ,,"" ...,1.. ~...
,
.........
.....
TUNING FOR SPEED

Now in its third edition, this book has been
revised, enlarged and brought up to date in such
matters as racing fuels, "tuned" exhaust systems
and the theory of carburation and lubrication.
Two new chapters, one on twin-cylinder engines
and the other concerned with two-stroke tuning,
have been added and the tables of valve and
ignition tunings, of sparking plugs and carburetter sizes and settings now include data
applicable to the most recent types of racing
engines. Based upon many years of actual racing
experience, this work is invaluable to owners of
racing motorcycles and of interest to every keen
motorcyclist.
fJ.uning
ALSO PUBLISHED IN ASSOCIATION WITH
"MOTOR CYCLING"
"MOTOR CYCLING" MANUAL
t04
SPEED
"MOTOR CYCLING" ELECTRICAL MANUAL

MODERN MOTORCYCLE MAINTENANCE
LIGHTWEIGHT MOTORCYCLES
"MOTOR CYCLING" YEAR BOOK
HolY to increase the performance of a standard

motorcycle engine for racing and
competition work
"MOTOR CYCLING" ROAD TESTS

THE CYCLEMOTOR MANUAL
MOTORCYCLING ABROAD
TRIALS RIDING
MOTORCYCLE ENGINEERING FOR SCHOOLS
BY
"MOTOR CYCLING" SERVICE SERIES
P. E. IRVING
IN ASSOCIATION WITH
THE STAFF OF
TEMPLE PRESS LIMITED

BOWLING GREEN LANE
LONDON, E.c.l
. 'I '.
..
'j
.::J,
FIRST PUDLiSHED 1948

REPRINTED 1949
SECOND EDITION 1951
REPRINTED 1952
THIRD EDITION 1956
CONTENTS
FOREWORD
vii
A PRELIMINARY SURVEY
II
III
ATTENTION TO O.H.V. CYLINDER HEADS
32
ADJUSTING COMPRESSION RATIOS
46
CHECKING PISTON AND RING WEAR
64
VI
THE CONNECTING-ROD AND BIG-END
79
VII
TRUING AND BALANCING FLYWHEELS
IX
X
REASSEMBLING CRANKCASE COMPONENTS
94
109
122
"TOP-HALF" REASSEMBLY
132
WORK ON THE CRANKCASE
LIGHTENING THE VALVE MECHANISM
141
ADJUSTMENTS TO VALVE GEAR
153
XIII
IMPROVED CYLINDER FILLING
XIV
IGNITION TOPICS
XI
XII
TWIN-CYLINDER ENGINES
168
182
189
TWO-STROKE ENGINES
207
XVII
ASSEMBLING THE ENGINE IN THE FRAME
234
XVIII
MAINTENANCE OF RACING CARBURETTERS
240
XV
XVI
XlX
XX
MADE IN GREAT BRITAIN
16
IV
VIII
Prill/ed by Purnell alld SOliS, Ltd.

Pail/toil (Somerset) alld LOlldoll
CARBURATION AND INDUCTION FACTORS
"HARD" AND "SOFT" TYPES OF SPARKING PLUG
258
TESTING THE FINISHED WORK
264
ApPENDIX
273
INDEX
285
FOREWORD
SINCE time immemorial man has sought to travel faster than his
fellow man. This inborn human urge gained tremendous impetus
some 60 years ago with the introduction of the petrol engine as
a practical proposition for vehicle propulsion. Inevitably, users
of the new form of transport felt their competitive instincts
aroused. Their demands for more power and more miles per
hour were met initially by the simple expedient of fitting larger
and still larger engines. In this" brute force" era, in which sheer
volume counted for more than volumetric efficiency, the results
of hit-or-miss experiments were all too obvious, the scientific
reasons for those results all too seldom understood.
Clearly a halt had to be called to this" Frankenstein's Monster"
method of progression. With a growing understanding of basic
principles in internal-combustion power production, allied with
practical experiments on the road, designers turned their attention to obtaining increased perfomlance from engines of more
moderate cubic capacity. So far as motorcycles were concerned,
this more scientific form of development was encouraged by the
introduction of the Isle of Man Tourist Trophy Races with their
strict limitation of engine sizes. This second era was one in which
the observant mechanically-minded amateur rider sometimes discovered a "secret of tune" which enabled him for a time to score
over his professional rivals.
With advances in metallurgy and a complete appreciation of
basic principles, the point has now been reached in this, the third
era, where there are no longer any "secrets". Science has replaced brute force and guesswork. Success in international
events depends upon the combined efforts of the designer and
development engineer, plus a host of specialist technicians
responsible for ancillary equipment.
But if the day has passed when the lone amateur could hope to
beat the factory representative in a major event, tuning skill,
combined with riding ability, continues to bring success to the
private owner in a wide variety of competitions. Nor need he
possess elaborate workshop facilities to improve the performance
of llis standard sports model or production type racer. What he
must have, however, is the necessary "know how" and that is
vii
FOREWORD
supplied in this' Manual. Written by a well-known technician

responsible for the design, development and racing preparation
of highly successful "factory" models, it embodies a wealth of
hard-earned knowledge yet reflects throughout the author's ready
understanding of the private owner's problems. Now in its third
edition and incorporating two new chapters devoted respectively
to twin-cylinder four-stroke and two-stroke power units, plus an
extended index and many additional illustrations, the book has
been fully revised and brought completely up-to-date.
Although intended primarily for the racing motorcyclist,
much of the text will prove equally useful to those drivers of
Formula III racing cars who may lack practical experience in the
tuning of high-performance air-cooled single- and twin-cylinder
power units. Moreover the ordinary everyday motorcyclist,
perhaps not in the least degree interested in racing, can benefit
from a study of the special assembly methods employed by racing
mechanics; a touring engine rebuilt to racing standards of
accuracy could well be more efficient in overall performance,
smoother running, cleaner and quieter in operation than hitherto.
The author emphasizes that in this textbook it has not been
possible to delve deeply into details of specific makes of machines
and power units. Thus, should the reader find that his advice
runs counter to that given in some manufacturers' instruction
manual, the latter should always be heeded or the maker consulted before modifications are undertaken. Similarly, where
reference is made to machine tools, heat treatments and so forth,
the information is given so that the reader may explain what is
required to the engineer deputed to carry out the work; specialist
technicians are not always well versed in the idiosyncrasies of
motorcycles and will appreciate expert advice.
Consulting Editor
.. Motor Cycling"
viii
CHAPTER I
SUCCESS in tuning an engine for speed depends upon three
main factors. They are: (1) getting the maximum combustible charge into the cylinder the maximum number of
times per minute; (2) turning as much as possible of the
heat liberated into useful work, instead of absorbing it into
the combustion-chamber walls or losing it down the exhaust
pipe; and (3) eliminating all unnecessary sources of internal
friction.
With these three points fixed firmly in mind, some time
can be spent very profitably in planning a course of action
which will ultimately give the maximum benefit without
wasting too much time, or money, on non-essentials.
It must be assumed that certain indispensable equipment
is available, such as a good workshop bench, well lighted,
and in surroundings of such a nature that scrupulous cleanliness can be observed, particularly during the final stages of
assembly. It is also a good plan to prepare suitable washing
and draining trays, large enough to accommodate a complete crankcase, before commencing work.
External and internal micrometers of assorted sizes are
useful, but expensive. As one will be concerned primarily
with clearances and not actual dimensions, they are not
absolutely vital, but a good set of narrow-bladed feeler
gauges most definitely will be. Another very useful gadget
is a cast-iron surface plate at least 12 in. square. Really
accurate scraped plates, too, are expensive, but a slab of
cast iron or mild steel plate, surface ground on both sides,
makes an excellent substitute, and there are plenty of
machine shops which will carry out the grinding at quite
reasonable rates. In conjunction with the surface plate will
..;-! . '(,
TUNING FOR SPEED
be needed a scribing block and dial gauge reading to .001

in., for checking squareness and parallelism of various
components and assemblies, and a burette, or graduated
glass jar, for measuring compression ratios.
A lot can be done without any equipment additional to
the usual workshop hand tools, but a little capital invested
at the outset is almost certain to pay a dividend later on.
When lack of cash is a deciding factor, it is a good plan for
several aspirants to speed honours to pool their resources
and thus obtain a well-equipped workshop without an
excessive drain on the pocket of each individuaL
Then one must decide for what sort of event the machine
is to be used-or, rather, for what it is suitable-grass,
scrambles, hill-climb or road events. However well pre4.'ared, very few engines other than modern types can compete with any hope of even moderate success in long-distance racing when limited to non-alcoholic fuel, but they
may do quite well in short-distance work-dirt or grasstracking, hill-climbs or sprint stuff-or even long-distance
track or road-racing where alcohol is permitted. Some of
the engines built several years before the 1939-45 war are
excellent for such work; they are usually lighter than their
modern brethren, and although sometimes rather deficient
in fin area judged by modern standards, this is not to their
detriment if the engine is to run exclusively on "dope" fuel;
in fact, it may even be an asset in short-distance events
when a rapid warm-up to operating temperature is necessary.
Right from the start it is best to keep a record, rather
than to trust to memory, of work done and the results which
have been obtained, and, in order to know where you have
started, commence operations by noting carefully the
ignition timing, valve timing and compression ratio.
The latter is determined by setting the piston at t.d.c.
with the valves closed, then filling the cylinder head from
the burette until the liquid just reaches the lowest threads
of tbe plug bole, the engine being positioned so that the
plug hole is vertical. Then rock the mainshaft very slightly
until the liquid reaches its maximum height and top-up,
2
if necessary, until it reaches half-way up the hole. Now

read off on the burette the amount of liquid poured into
the head. Assuming this to be 80 c.c. and the correct volume
of the engine to be 498 c.c., the C.R. would be
498 + 80
578
7.22 to 1
and so on for any other capacity (see Chapter IV), The

level of liquid in the hole should be watched closely for
a while, to verify that no leakage is taking place past the
rings or valves. If there is a leakage, then the real volume
The method 0/ measur-
~
<
illg combustiol/-chamber
I'oiume with a burette

alld liquid-eilher paraffill alld thill oil or water
is lIsed.
of the head will be somewhat less than the apparent volume

but, unless the engine is in a very bad state indeed, the
inaccuracy will rarely be bad enough to matter at this
stage of the proceedings. If you really want to do the job
accurately, then lift off the head and carefully apply grease
around the valve heads and between the piston crown and
barrel to form a temporary seal.
As to the liquid used, water is the best when there is any
doubt about the condition of the valves or rings, because it
is less likely to leak past these points than paraffin or light
oil, particularly if a drop or two of engine oil is first
3
TUNING FOR SPEED
squirted up the inlet and exhaust ports. For checking the

ratio of a newly assembled engine, a mixture of paraffin
and thin oil is preferable, as it does not leave a legacy of
rust behind it.
Although burettes (which can be obtained from most
wholesale chemists or scientific instrument suppliers) are
mostly graduated in C.c.s, there is of course no absolute
need to use either that instrument or that scale. The head
volume could just as easily be measured in fluid ounces or
cubic inches, provided that the same measure is used for
the cylinder capacity. In working out the ratio remember
that 35.2 fluid ounces equals one litre (1,000 c.c.) and, using
this conversion factor, the volume of any cylinder can easily
be determined.
If a timing disc is fitted to the mainshaft before measuring the clearance volume and a wire pointer is fixed at some
convenient location, the pointer can be bent to fix accurately zero on the disc when the piston is absolutely at
t.d.c., as indicated by the liquid in the plug hole having
reached maximum height.
The ignition and valve timings can now be checked, and to
avoid inconsistencies it is best to adopt a standard system of
measurement, preferably that recommended by the makers.
For measuring the ignition, first set the points to the correct
gap (:012 in. in the case of most racing magnetos), insert a 1i-thou. feeler or a piece of tissue paper between
the points, and, while applying slight resistance to the
armature spindle (by holding the sprocket nut between
the fingers or some such method), in order to take up any
back-lash in the drive, turn the motor until the feeler is just
released as the points commence to open, the control
being, of course, set to full advance.
On machines fitted with automatic centrifugal advance
mechanisms, it is best to lock the mechanism into full
advance with a piece of metal or wood jammed between
one pair of the fixed and moving stops; checking at full
retard and adding the theoretical amount of advance given
by the A.T.D. is not really accurate, although it is a good
4
plan to determine for subsequent reference what the actual

advance range is.
For the valves, methods vary: some cams are ground
with " quietening curves " at the start and finish of lift,
which give a very slow take-up, making the actual" lift"
and "close" points difficult to determine. The recom
mendation in such cases is to measure the timing with the
tappets set to a clearance much greater than the normal
running clearance. This has the effect of making the
opening and closing points more sharply defined, but, of
course, it does not give the true figures, a point which must
be borne in mind when comparing the timings for engines
of different makes.
A table of representative valve timings and clearances is
given in Chapter XII from which it will be seen that, on
o.h.c. Nortons, the timing is quoted with less than the
running clearance. As the clearance is adjusted by varying
thicknesses of shims inside,the valve caps on these models,
the recommended checking clearance is obtained simply by
using feeler gauges of the appropriate thickness between
valve stem and cam follower.
On Vincent engines, it is not possible to "measure the
clearances with feelers; the method adopted is to set up a
dial gauge in contact with the valve stem, and to adjust the
tappets so that the valve is lifted one to two thous., making
sure, of course, that the follower is on the base circle of the
cam and not just at the start of lift. Then turn the engine
until the valve lifts a further .005 in., and read off the timing
at that point. This method is the one which should be used
on any engine to get a really accurate reading of the valve
timing.
The crank angle at which each valve reaches full lift, too,
should be measured, and also the full height of lift of
both valves and the amount each is open at both bottom and
top dead centres with correct running clearances. You will
then have a fairly complete picture of the valve-timing
diagram, though to do the job properly, readings should
be taken at, say, 2" intervals: the lift curve can then be
5
TUNING FOR SPEED
plotted on squared paper and the effect of any subsequent

cam modifications will be shown very clearly by plotting
the new lift curves on the same sheet (see Chapter XIII).
All this measuring and note-taking may seem very upstage and time-wasting, but if you are to do any good at
the game you must know what you are about, and you may
*as well start off correctly. You may, for instance, do something very silly later on, but will have all the information
necessary to check back on earlier work and thus be able
to put the motor back to a previously successful state, and
then start again from there but along different lines.
The way is now clear to get down to business, the first
step, obviously, being to strip the engine completely, keepjng watch all the time for places where undue wear bas
occurred, for wear implies friction, which is one of the
main things to be eliminated.
Mechanical Condition
Regarding the general mechanical condition, any parts
which show signs of bad wear should be replaced, the main
thing being to decide what constitutes "severe" wear.
Because a bit of mechanical noise does not matter, a certain
amount of slackness here and there can be disregarded; but
it is a mistake to think, as some do, that everything should be
cam gear and small-end
sloppy. Plain bearings in
should have a clearance of between .001 in. and .002 in.
per inch of diameter, with a maximum of .003 in. Ball or
roller bearings should have practically no slack in them at
all, and the race tracks must be examined to see that no
pitting has taken place.
When dismantling any engine containing more than one
set of rollers, take particular care to avoid mixing up the
sets. Take each bearing apart carefully, place each set
of rollers in a matchbox, or tin, and make a note on the
container stating the location of the contents. Injecting
fresh oil into each bearing prior to dismantling will usually
serve to retain the rollers in place long enough to stop them
cascading all over the bench.
the
Sometimes a bearing will be slightly pitted in one small

area but is otherwise in good condition: In this case" as
a temporary expedient if no spa~e i.s avaIla?le, the beanng
can be refitted on its shaft or III Its hO~SI~g so th~t the
damaged area comes into a position w~ere It IS ?ot subjected
to heavy load. The line of least load ~s ~t 9 ~ clock ,on the
drive-side main, at 3 o'clock on the tImmg-side !Dam, and
9 o'clock on the crankpin-in each case lookmg at the
crankcase from the drive side.
, ,
One point which needs to be watched closely IS blgend lubrication, and another is whether the. rod or cage
shows any tendency to work over to one sI,de a~d bear
heavily against the flywheel. The lat~er state IS a SIgn that
either the crankpin or big-end sleeve IS tapered, or that the
roller-cage slots are not absolutely parallel with the ,axle;
but, in any case, the trouble is liab.le to be aggravated If the
oil is fed in at one side of the cage mstead o~ through a ~ole,
or holes, immediately under the rollers. SIde-fed, beanngs
are a potential source of bother at prolonged high revs;
if possib1~, steps should be taken to ~b~ain a central feed,
either by modifying the pin or obtammg ano.the~ of the
approved pattern. This point will be dealt With III more
.'
detail in Chapter VI.
If the compression is already faIrly high, say, up to 7 to 1
or so, it is probable that enough I?etal can be turned off
the cylinder to raise the C.R. Suffi,cI~ntly ~or pet!ol-~eIl:zo!e
to be used while retaining the ongmal pIston, l~ thl~ IS m
good condition and shows no signs of crackmg IIl: the
gudgeon bosses. If" dope" is to be used, the ,old pIston
can be put quietly on the shelf, for a. new hIg~-domed
pattern will be required in order to obtam the deSIrable 14
to 1, or so, ratio.
,
The condition of barrel and rings is especially Important,
for it is no use sucking in a lot of gas just to let it blow past
the piston. If an area of wear exists ~t the top end of the
barrel, the rings will lose contact WIth the walls at very
high speeds, allowing pressure to, be reduce~.
Unfortunately, few cylinders will stand bemg rebored to
7
TUNING FOR SPEED
the full standard oversize without bringing the engine outside the class cubic capacity limit, and you may have to look
into the question of obtaining a new barrel or having the old
one re-lined. The latter technique is quite a good idea for
the arduous conditions of dirt-track or sand racing, because
the liner can first be bored out .005 in. under-size, and, later,
honed out to standard diameter: after that a new liner can
be fitted and thus the expense of new barrels is avoided. If
the wear is only to the extent of two or three tho us. the
barrel can be lapped out by hand.
The condition of the valves, guides and seats is important.
Inlet valve stems should have between .002 in. and .004 in.
clearance, but exhaust valves, if made of K.E. 965 or
Jessop's G2. steel, need more, .002 per 1 in. of stem diameter
being a good figure to work to. If the clearance is less than
this, the valve may bind at high temperatures; if more; it
cannot get rid of heat properly, owing to insufficient contact
with the guide.
This matter of heat dissipation is vital, and if the existing
guide is of cast iron (which is a comparatively poor conductor of heat) it can, with advantage, be replaced by
one of aluminium-bronze or phosphor-bronze, though the
latter has been known to give trouble under extreme beat
conditions tbrough breaking away at the hot end near
the port.
An alloy composed of copper with I per cent. chromium
added, and known as chromium-copper, is even better for
heat transference and its rate of wear is low; itis, however,
difficult to machine and particularly hard to ream. It must
be remembered that none of these non-ferrous materials
will function satisfactorily without lubrication; cast-iron
will work dry, but even it will benefit from the presence of
a little oil. Cast iron is, however, quite satisfactory for
the inlet valve guide, as heat conditions here are much
less severe, and iron material can with advantage be used for
exhaust guides in place of aluminium bronze in engines
with aluminium heads using methanol, Shell A or 811, or
similar fuels arid a castor-base oil. This type of oil is soluble
8
in methanol and, as the mixture in such engines is often

excessively rich, the guide lubrication suffers and wear is
rapid.
Valve and Seat Modifications
If the exhaust valve has worn to a very thin edge or shows
signs of being badly burnt under the head, it should be discarded in favour of a new one.
The silchrome or cobalt-chrome valves frequently fitted
to sports engines are very satisfactory in ordinary road use
for both inlet and exhaust, but where racing is concerned it
is better to play safe and go in for an austenitic steel, i.e.,
K.E. 965 or Jessop's G2. for the exhaust valve, although the
standard steel should be satisfactory for the inlet.
Apart from the faint chance that a crack may have
occurred between the exhaust valve seat and plug hole (a
defect which is sometimes difficult to detect), an old cylinder
head which has seen much service usually shows a good
deal of valve-seat wear-particularly on the inlet sidewhich causes the valves to be slightly masked and reduces
their effective lift. If, as is frequently the case in sports
engines, the valves are of equal size, the best way to cure this
trouble will be to obtain a valve about 1 in. larger in
diameter, open out the inlet port to suit and recut the seat.
This procedure may not be possible in some engines
where the valve is already of the maximum permissible
diameter, as dictated by the position of the plug hole or the
liability of fouling the edge of the exhaust valve when both
valves are partially lifted. In such a case the masking must
be removed by cutting it away with a tool similar to a valve
seat cutter but with a much flatter angle; if the valve seat
angle is 45, it is often possible to utilize a cutter intended
for a larger valve with a 30 seat. Some welding firms make
a speciality of building-up worn seats by welding, and in
some cases also will re-machine the head, though not of
course to racing standards; both iron and aluminiumbronze heads can be reclaimed by this method, but there is
a grave danger of melting the aluminium jacket or at least
9
TUNING FOR SPEED
destroying the close adherence between the two metals in

the case of bi-metal heads which have aluminium fins cast
round a bronze" skull".
Boring out the seats and fitting inserts is an attractive
method of head reclamation, since there is no risk of
cracking or spoiling the head through excessive heat, and
also the inserts can be made of material which is less prone
to seat-sinkage than the original; cast aluminium-bronze
is bad in this respect whereas wrought aluminium-bronze
in the form of tube or extruded bar is good. Seat-rings
made in plain cast-iron or austenitic cast-iron which is highexpansion, high-wear-resisting material, can be obtained
from various firms-" Brico" and Wellworthy to mention
only two-in standard sizes which may, however, need to
be reduced in outsize diameter to be usable in some heads.
Seat rings can also be purchased as spares from firms who
supply aluminium "heads with inserted seats as standard
equipment. The nickel-iron anoy known as "Ni-resist" is
of the austenitic type and is excellent raw material from
which to make seat-rings.
When fitting seat-rings, the limiting factor is usually the
thickness of metal which will be left between the recesses
and the plug-hole and at the top of the sphere between
the two recesses; if the thickness at the surface of the
sphere is less than 1 in. there is a risk that a crack will
occur in service, particularly if there is much difference
between the thermal coefficients of expansion between the
metals of the seats and the head. The most likely place
for such a crack to start is between the plug-hole and
exhaust seat, and it may be possible to improve the situain.
tion by using a seat-ring to suit an exhaust valve
smaller; admittedly this will leave a small ridge in the
port but this can be subsequently blended out and in any
case the exact shape of the exhaust port is not of great
importance.
Seat-rings of the same coefficient of expansion as the
head do not need to be fitted very tightly : 1t to 2 thous. per
inch of diameter will retain them indefinitely, provided that
10
n-
the mating surfaces are very accurate and smooth and the
bottom of each recess and ring are square to the axis.
For aluminium-bronze heads, use either Monel-metal or
K-Mon~l (~btainable from Henry Wiggin and Company,
Ltd., Blrmmgham) or wrought aluminium bronze preferably in the form of tube.
Ri,ngs should always be shrunk in, never pressed in cold,

as wlth the latter method there is always a risk of scoring
the surfaces. With .002 in. per inch of diameter interference a ring should just drop into a cast-iron head at
200 C. temperature difference, but, to make sure, 250 0 C.
is advisable. This can be obtained either by just heating the
head to 250 0 C. or by heating it to 200" C. and freezing the
ring in dry ice (frozen carbon dioxide; 200 0 C. (400 F.) head
temperature without freezing is sufficient for a bronze head'
this temperature can be obtained in a domestic gas
electric stove. To ensure quick and accurate fitting a
mandrel (or mandrels) should first be made up from steel
bar, with a pilot which fits the guide bore and a register
on which the ring fits with a little more slack than the
interference of the ring in the recess (otherwise the mandrel
may be gripped by the ring). In use, a light coating of
grease on the register will retain the ring whilst the mandrel
is pushed quickly home and pressed down for some seconds
until the ring is gripped by the head.
Aluminium-alloy heads with inserted seats have been
used on racing engines for many years and are a feature
of all post-war production racers. Various methods of
fitting the rings have been tried, but the plain parallel
form just described is much the simplest to make and to
replace and never comes loose if properly fitted in the
first place. Aluminium, besides having a high coefficient
of expansion loses a large proportion of its strength at
high temperatures, though some alloys such as Y-alloy
and R.R..53b' are much better than others in this respect
and are In consequence much used. Two conflicting
factors accentuate the problem; if the initial fit is not
tight enough the head will expand away from the ring
11
Q
0;
TUNING FOR SPEED
and it will loosen ; but if the fit is too tight, under certain
circumstances the' head metal may stretch or else the
ring may be contained so tightly that it collapses and in
either event the ring will loosen. The most satisfactory
seat materials seem to be austenitic cast iron for the inlet
and wrought aluminium-bronze for the exhaust, and with
these materials a cold interference of .003 to .004 in. per
inch of diameter gives perfectly satisfactory operation in
Y-alloy or R.R. 53b heads. Austenitic calit iron work
hardens very rapidly and any machining operations must
be done with a perceptible depth of clit ata low speed;
if the tool is allowed to skid on the work it will be very
difficult to re-start the cut. For that reason, ordinary
valve-seating cutters operated by hand are useless; the
only satisfactory way to form the seats true to the valve
axis in the first place is by means of a seat-grinder of the
Black and Decker type.
Inserted seats which are worn but still tight cannot easily
be removed by heating the head, as the temperature required
to get the seat loose enough to drop out is so high that the
heat-treatment of the head metal may be affected. The best
method is to turn out the old seat very carefully to avoid
damaging the recess, or else to run a flat-ended drill down
through the ring at two or three places and break the ring
into pieces. This operation must, however, be done very
carefully to avoid making any cuts in the wall of the recess,
or in the bottom face. Local damage of this sort is almost
bound to lead to gas leakage and severe damage to the head.
Sometimes a seat ring will be found to be loose, or there
will be some evidence that it has been 10Dse when the head
has been at running temperature. If the recess and ring are
still otherwise in good condition, a satisfactory repair can
be made by copper plating the ring and finishing the
deposited metal by turning or filing in a lathe to a size which
will give the correct interference fit. All but the outside
diameter of the ring should be "stopped off" with wax
bef-;,e plating to eliminate any possibility of plated metal
flaking off under the action of heat inside the head.
12
Should the plug hole be tapped to take the old-fashioned

18 mm. size, which is now almost unobtainable in " hard"
grades, it can be fitted with a 14 mm. adaptor sleeve, usually
obtainable from one of the accessory houses. Perhaps a
better plan is to screw in a solid steel plug, lying ilr in.
or so below the inner surface of the head, and to get this
recess filled in by an acetylene welder with Sifbronze. This
can be done at a dull-red heat, and will not harm the head
if it is made of iron. A new plug hole can then be drilled
slightly off-centre and at a different angle to the original,
so as to get the plug points as near to the centre of the head
as possible. The diagram gives the machining dimensions.
The remaining components still to be considered are the
cams and rockers. Supposing that the valve-timing figures
have worked out like this: Inlet opens 35 before t.d.c.,
closes 60 after b.d.c. Exhaust opens 65 before b.d.c.,
closes 30 after t.d.c.; the cams are of such a type as to be
worth retaining-at least for the time being-unless they
are badly worn or exhibit wavy grooves on the flanks. But
if the timing is slower (i.e., of shorter duration) than that,
and particularly if the overlap at t.d.c. is less, then a new
set of cams should be obtained.
!/lnH. D
TO SUIT LONG
18 mm. " I.S PITCH

THREAD
fACE SQUARE
TO THREAD
All old-type cylinder head fitted with all 18 mm. sparking

plug call befitted with this special adaptor to carry a 14 mm.
plug. The adaptor is drilled and tapped at all angle to
brillg the plug points illto a more favourable position.
13
~~5~~!li~l!iiliJil,'I!fi~lffii!i!ii!!)!il1!1~.{i!i!iti,l'i~~!!IJ_~~_'l!:i1IIiiili'lilimMIIJllllliIlIi'_ _ _ _ _ _ _ _ _i!Il!Il!I_ _
Bil&2&Iim_li111li._1:!m~"""'=;""'~,_;.,;",,'";;;;';'-".0>
1
TUNING FOR SPEED
If ),ollr cams give a

timing approximate to
this diagram and with
not less thall the overlap shown, they call be
retailled temporarily
alld the performance
improved by olher
means.
indication that cams giving longer valve openings are

required for maximum power at high revs.
The foregoing more or less covers the preliminary survey
of the new parts which may be required and the work to be
done. The methods of doing that work form the subject of
subsequent chapters.
The lift must also be considered and should be slightly

greater than one-quarter of the port diameter.
Many makers will supply racing cams which can be
fitted directly in place of touring or sports cams, and it is
often possible to obtain higher lifts by utilizing cams made
for an engine of similar make but of larger capacity.
The valve-timing diagram is detetmined not only by the
cam contours but also by that of the cam-follower feet, and
wear of their surfaces will affect the timing quite considerably. The grooves which eventually appear in cam followers
with non-rotating solid feet have the effect of opening the
valve earlier and closing it later; therefore, if considerable
wear of this nature is discovered, it is best to recheck the
timing with new rockers, or with the old ones reground to
the correct contour.
If the wear is more than 10 or 12 thous. in depth the
rockers will, in any event, need to be reground and casehardened anew unless new ones can be obtained, and this is
obviously a job best done by the makers. In this connection
it is sometimes found that an engine which has been going
very well despite severely worn rockers, falls off in performance after new rockers have been installed. This is a clear
14
15
CHAPTER II

AT this juncture it is worth noting that a lot of fun can
be obtained by "hotting-up" a side-valve engine' there
have been quite a number of 90 m.p.h. s.v. "5008" in
existen:e at. one time or another, and readers with long
memones WIll recollect the astonishing little 250 C.c. J.A.P.
engines which were reputed to be able to rev. at 8,000
r.p.m. The s.v. unit is usually much lighter than its o.h.v.
c~unterpart, and can be screamed up to maximum revs.
WIthout danger of the valves hitting the piston or fouling
each other.
Irrespective of the location of the valves or their method
of operation, the prime necessity is to have the inlet passage
as unobstructed as possible. An easy exit for the exhaust is
also ~ghly desirable, but this is more for the purpose of
ensunng cool running than to aid in power production.
There are three main factors to be considered in the inlet
tract: (a) the size of the valve ; (b) the shape and length of
the port; and (c) the size of the carburetter. To some
extent .these are allied to each other. For example, it is no
l:se usmg a very large-bore carburetter if the rest of the
induction system is inadequate; such an instrument will
give little or no extra power and will certainly spoil the
acceleration. Improving the breathing capacity of the
valve and port without increasing the choke size will on
the other hand, improve the power a little at high :evs.
with?ut affecting the low speed end much; but, of course,
the Ideal arrangement is to make the whole induction
system match, thus carrying out a principle once aptly
described as " proportional tuning".
A 500 c.c. engine with a conventional cast-iron o.h.v.
head will usually have a carburetter of Ii-in. or 11\-in. bore;
16
the latter is about the smallest size which will give anything
like good results, and as progress is made, larger sizes can
be tried. At the commencement it is unwise to use too
big a choke, and the following ranges are recommended:
500 c.c. single o.h.v., Ii in.-l-& in.; 500 c.c. twin o.h.v.,
in.-l in.; 350 C.c. o.h.v., 500 C.c. s.v., 1 in.-1J.\ in.;
250 C.c. o.h.v., 350 C.c. s.v.,
in.-~ in.; 250 C.c. s.v.,
~ in.
As to valve size, it is scarcely possible to lay down any
hard and fast rule as so many factors are involved, but the
following table gives an idea of the throat sizes which are
advisable: 500 C.c., 1i in.; 350 C.c., I! in.; 250 C.c.,
1-& in.
The capacities refer to the size of individual cylinders
and not to the engine as a whole. At best, however, these
figures can only be a rough guide, because other factors
besides the cylinder capacity enter into the matter; the
type of performance (i.e. the shape of the power curve) .
required, the valve timing, compression ratio, the design
of carburetter and the nature of the exhaust system all
exercise an effect. Generally speaking, the faster the
engine is capable of revolving, judged from a purely
mechanical sense, the larger the choke which can eventually
be used, though a good deal of modification may be
required before- the ultimate size is attained; this line of
development almost inevitably leads to a type of engine
which, whilst it develops very high power at high r.p.m.,
is virtually useless below a certain minimum speed and in
consequence has to be used in conjunction with very carefully chosen gear ratios and driven with great skill in order
to keep it always pulling above its minimum speed. Such
a power unit will show up at its best on a fast circuit, but
may be equalled or even outclassed by an engine with less
maximum power but with greater torque or pulling power
at low or medium speeds, on a short course with many
corners or in a vehicle such as a racing car where a very
close set of gear ratios cannot be used. The Italians are
notable exponents of large-bore induction systems; the
17
"*
TUNING FOR SPEED
250 Guzzi has a-choke of 32 mm. diameter, or just over

Ii in. and, therefore, 16 in. larger than that normally fitted
to post-war 500 C.c. " Manx" Nortons, and this bore is continued right through the long induction pipe and inlet port
up to the throat of the valve. The 500 C.C. twill of the same
make, which is virtually two 250 sinales has even been
used with 35 mm. chokes and the sam; si;e is used on the
four-cylinder Gilera, in which one carburetter feeds a pair
of cylinders of which each is of course only 125 C.c. So
far, even the fastest English" works" machines have never
gone to these e~tremes; the twin AJ.S. had one 1i in.
carburetter feedIng each of its 250 c.c. cylinders, whilst
the "works" 500 c.c: Nortons and Velocettes used 1ilr in.
chokes. The foregoIng data refer to enaines built circa
195~52. Si?ce. then e~lgine speeds ha~e gone higher
p~rticular1y In sIngle-cylInder types, and chokes of liB in.
dlam: h~ve been used on Norton and Vincent engines,
runnIng In excess of 7,000 r.p.m.
Even these sizes, though smaller in relation to the
cylinder c~pacity than th~, Italian examples, furnish a gas
speed conSIderably lower th~ those stated by some authorities to be t~e theoretical optimum, this somewhat puzzling
par~dox bemg probably caused by the aforesaid optimum
hav~ng ?een . determined in manifold-type multi-cylinder
engmes In whIch the fuel has to be vaporized and maintained
i~ t~at s~ate before reaching the cylinders to obtain equal
dIstrIbutIOn, whereas in motorcycle engines the bulk of
the fuel is not actually vaporized until after it enters the
hot cylinder; this process takes full advantage of the
la~er:t heat of. evaporation of the fuel to cool the charge
wIthin the cylmder and so results in a greater weight of
charge being induced.
Work on the Inlet Port
If your engine measures up to these standards and the
valve head and seat are in good condition, it may be well to
leave things as they are temporarily and to. experiment with
18
larger valves at a later stage, but in the meantime several

jobs can be done to improve the breathing ability. As
the valve is operating only partially open for much of the
time, anything which will ease the passage of gas past the
seat at low lifts is highly beneficial, and one step towards
this is to reduce the seat width from the usual j~ in. or so
down to a mere :lz in. by carefully radiusing the corners of
the valve at the seat and the top edge, and performing a
similar operation on the port.
RADIUS EDGES
7~O"NWOU?
AT WGE OF SEAT
How to reduce the width of the seatings on

valve face and inlet port.
If the seat is very wide, as it frequently is when interchangeable valves are used, it is also possible to open out the
port by perhaps -fir in., which Plovides a useful increase in
total area. The correct method is to use a cutter, similar to
a valve-seat cutter, but with an included angle of 45,
but the job can be done by careful work with a file and
scraper.
F or rounding off the valve edges, hold the stem in the
chuck of a lathe or drill, remove the bulk of the metal with
a smooth file and finish off with emery cloth. Should the
valve-seat in the head be very badly pocketed and you
decide to fit a larger valve, it is often possible to utilize
one from a larger engine of the same make, as it is a frequent
practice to employ the same size of stem throughout the
range; alternatively, racing editions of sports engines are
often fitted with larger valves, and one of these could be
employed.
The shape of the inlet valve under the head is of more
importance than is often realized and depends upon the
shape of port. It used to be the fashion to have a com19
TUNING FOR SPEED
paratively fiat, straight port-a shape which directs the gas

flow past the valve stem into the region nearest the exhaust
valve-and in this case the valve head must be fairly flat
underneath and joined to the stem with a small radius, so
as to allow plenty of room for the gas to flow past into the
operative region. This port shape-see sketch below-is by
no means the best, partly because the area (B) is relatively
ineffective owing to the acute change in direction which the
gas is compelled to take, and partly because much of the
charge coming in at (A) is liable to rush straight across and
out through the still-open exhaust valve, particularly if a
highly extractive exhaust system is in use.
LEAKAGE PATH
Limitations inherent ill the oldtype "flat" inlet

port are depicted ill
this sketch. Note
holY the fresh
charge tends to
.{low inlo the exhaust port.
~;
Ii
11
'I
INEFFECTIVE AREA B
The shape shown in the second sketch-as used in

modern engines-may not look so well to the casual eye,
but is, in fact, better because almost the whole of the seat
circumference is brought into effective use. This design of
port works best in conjunction with a semi-tulip valve the
shape of which has the effect of steering the gas towards
the lower side of the seating. Some of the fresh charge is
thus directed away from the exhaust valve, less will escape,
and it will pick up less heat from that hot-headed component. Conversely, the latter is inclined to run hotter, a
point which will have to be. watched in petrol-benzole
motors but is of less consequence with alcohol fuel.
By going up in valve size, it is possible to improve the
shape of a flat port (as indicated in the third sketch by dotted
20
The modern dOlYndraught port used

ill conjunction with
a til lip-shaped inlet
valve overcomes the
defects of the flatport design.
lines) in addition to increa~ing the valve area, but it may. be

better to leave this operation until a later stage, thus bemg
able to pull something out of the bag which others may not
suspect. In any case, the valve guide must be so treated. as
to offer the minimum obstruction, the usual method bemg
to cut it off flush with the roof of the port. This sometimes
reduces the guide length rather too drastically; equally
good results can be obtained by fairing the guide off to a
knife edge on the entering side and filing out the sharp
corner between the guide and port, reducing the guide
wall thickness down to h in. or so at the sid~s. When a
larger inlet valve is fitted, or sometimes even 1ll standard
heads, a certain amount of masking or interruption to smooth
gas flow may exist in the area around the lower edg~ of
the valve which may prevent this area being fully effective.
GUIDE NOT CUT OFF FLUSH
rF~I,~!~T~HI~S~C~UiR,VEiDEfLECTS
j
EX.VALVECHARGE
How gasflow can be improved by enlarging the inlet tract

to suit a bigger valve. The dotted lines indicate where
metal should be removed.
21
TUNING FOR SPEED

OBSTRUCTION
~~Of-"'--- TO GAS FLOW
Gas flo IV may be

improved by reshaping the mating
edges of the head
and cylinder barrel
adjacel/f to the inlel
valve.
flexible shaft chuck, and a strip of emery cloth or better

still, "Durex" abrasive tape, is slipped into the slot and
wrapped round tht'f rod up to about! in. diameter. The
wound-up tape must be inserted into the port before the
power is turned on after which the flying end of the tape
rough-polishes the port very rapidly. For final polishing,
use felt bobs moistened with ordinary metal polish, or
cotton material wound round the slotted bar. There is
some controversy regarding the value of an absolutely
mirror-like surface, but there is no doubt that a finish of
this nature cannot do any harm, and is at least a source
of envy to chance beholders; however, shape is of greater
importance than mere polish.
PARTICULARLY AT
lOW VALVE LIfT
It is often possible to cut away some metal from the head

and in some instances the mating edge of the cylinder
barrel also, and generally shape and smooth up the surfaces so that gas coming past the seat radially to the valve
is gently guided into the cylinder instead of impinging
against an irregular surface, or one which is too close to the
valve to permit easy entry. For these port and head modifications, rotary files or emery wheels driven by a flexible
shaft from an electric motor are invaluable, but the same
effect can be produced with "riffiers" or bent files.
Ordinary round files can be bent to shape after heating to
re~ness, and re-hardened' by plunging into water; after
thIS treatment the steel is very hard and brittle, but if not
knocked about, the teeth remain keen for a long time.
~rom the carburetter flange face or stub up to the valve
gUIde boss there should be a gradual taper in the port,
entirely frce from ridges or hollows, and after the whole
port has been brought to a satisfactory shape it must be
polished with rough and fine emery cloth to a dead smooth
finish. For rough port-polishing, one of the handiest tools
is a piece of steel rod slotted across a diameter with a hacksaw for about i in. in from the end. This rod is held in the
22
Side-valve Inlet Ports
The treatment described for o.h.v. inlet ports applies

equally well to side-valve types .. Except in those cases
where the port has been purposely elevated in order to
obtain clearance over a rear-mounted electrical installation,
most s.v. ports have reasonably easy curves, and, so far as
the approach of the gas to the underside of the valve is
concerned, may be as good or better than the average o.h.v.
type. As to valve shape, the same rules apply: a fiattish
GRIND AWAY HERE TO
III a side-valve ullit the gas flow call usually be improved

considerably by grillding away metal ill the cyJinder head
at the poillt illdicated. Great care is required in this operatioll. Do not overdo it.
23
TUNING FOR SPEED
port requires a flattish underhead profile, but a combination of a swept port with a semi-tulip shape of valve gives
the best results.
The valve guide can be shortened a little or even cut off
flush with the port without detriment, but the amount to
which this is done depends largely on the port shape. It
may be better to leave the guide fairly long and streamline
it by tapering off to a very thin section at the upper end.
Certain V-twin engines were built with the inlet ports
lying horizontally in order to employ a plain T-shaped
induction manifold, but an incidental advantage gained is
the reduction in the angle of bend in the port itself. The
front cylinder off such a motor mounted on a single-cylinder
crankcase and equipped with a carburetter arranged to
work at 25 upward inclination offers interesting possibilities.
It is obvious that after passing the valve the charge has
to undergo another change in direction before it can reach
the cylinder, and this, plus the fact that a third-possibly
one-half-of the valve opening is partially masked by the
valve chest wall is bound to impair the breathing. The
change in direction cannot be avoided, but the effect of the
masking can be reduced by clearing away the chest wall
with a port-grinder and eliminating as far as possible all
sharp angles.
.
MINIMUM CLEARANCE
THIS AREA NOT TO ASSIST IN RAISING C RAilO
CONSIDERED AS
RAISED ROOF WITH SMOOTH
TO DIRECT GAS FREELY
INTO CYLINDER
Further Aids to "Breathing"

In detachable-head engines, care must be taken to see
that the inner edge of the gasket merges properly with the
interior of the head. It may be advisable to cut a new
gasket, from soft sheet aluminium or copper, which fits the
head studs accurately, and to trim it up to the correct internal
contour after bolting to the head.
As a further aid to " breathing", reduce the valve head
diameter until the seat width is only -fir in. and round off
the edges of seat and valve as previously described. On
account of the proximity of the chest wall, it is rarely of
much use to attempt to use a larger valve, although it may
be possible to do so if there is ample room available around
24
FLAT STRAIGHT
DIRECTING
STREAM TOWARDS
PORTION OF VALVE OPENING
1I
A section of the inlet tract of a side-valve engine modified

for special high-performance valve timing.
the head, or alternatively if there is sufficient metal in the

casting to allow for the necessary grinding-away.
Most side-valve engines have "slow" timings and low
valve lifts, with the intention of giving good low-speed
performance and, in keeping with this idea, are fitted with
small-bore carburetters. After due attention to the induction tract and timing (a point still to be dealt with), a larger
carburetter can be used, although it is not possible to go up
to quite the same size as an o.h.v. unit will take. As a
commencement, use the next size lower to that recommended
for an o.h.v. engine in the capacity choke scale on page 17.
Naturally, the inlet port must be increased in diameter to
match up with the larger-bore carburetter. One of the
difficulties associated with the obtaining of good breathing
in a side-valve engine on the lines just described is that, if
ample gas-flow areas are provided all round the valve head,
the compression ratio may become much too low. A possible
line of development worth pursuing only if valve-timing
modifications can also be carried out would be to regard
B
25
TUNING FOR SPEED
CARBURATIO N AND INDUCTION FACTORS
that portion of the valve opening furthest from the cylinder

as being relatively useless anyway, and to re-design the
combustion chamber with the space immediately above
the inlet valve machined parallel to a radius only slightly
greater than the largest inlet valve that can be worked in.
The dep~h of this poc~et should be about 3h in. greater than
the maXlmum valve ltft, and the cam contour modified to
slam the valve up to fuIllift as rapidly as possible and hold
it at full lift for a considerable period.
The port likely to be most effective in this instance would
be of the flattish type, as shown in the diagram on page 25,
directing the gas stream mainly in the direction of the effective
portion of the valve opening. To assist the gas in changing
direction to enter the cylinder, the head may be hollowed out
slightly and the top edge of the cylinder bore radiused off,
though not so far down the bore that the top ring is deprived
of the protection of the upper ring land.
pletely in order to get the maximum port area; by so doing

more of the valve will be exposed to the outgoing gas and
the stem will be robbed of much valuable contact area
through which most of the heat picked up has to be con
ducted.
On the other hand, if the port is so shaped that a large
proportion of the extremely hot gas is caused to impinge
on the stem without a free path of escape, there is a grave
risk of valve-failure, through overheating at the junction
of the stem and head. Cutting away the side walls so that
the port is considerably wider than the diameter at the
throat reduces this effect very considerably, and is in fact
the only course open in some engines where for the sake of
compactness the port has purposely been kept fiat.
Before doing much modification to port shape, the thickness of metal between the port and neighbouring pockets
such as rocker-boxes, spring-boxes or the recess sometimes
provided to accommodate coil valve springs, should be
explored, preferably with indicating callipers, to ensure
that there is ample metal present; nothing is more annoying
(or expensive) than to find the port breaking through into
a place where it should not do so, particularly just as the
job is nearing completion. Also, on inserted-seat heads,
The Exhaust Port
The exhaust valve and port still remain to be dealt with,

and it should be appreciated that the gas-flow conditions
are quite different from those existing in the inlet. In the
latter, the gas approaches from the back of the valve, and
the pressure-difference between port and cylinder is (or
should be) of the order of 5 or 6 lb. per sq. in. only
as a maximum value, and must be much less than this at
the finish of the induction stroke.
With the exhaust valve, however, the gas at the moment
of opening is at high pressure-70 or 80 lb. per sq. in.and is waiting aU round the periphery of the head, ready to
escape the moment the valve begins to lift. Thus, whatever
the port shape, the whole of the circumference of the valve
is operative, and 50% of the gas can escape straight down
the port; only half of the gas is obstructed by the valve
stem and guide, consequently these can be permitted to
occupy quite a lot of space without detriment.
It is a mistake to carve away the guide and boss com-
(Right) The ideal exhaust-p:JrI arrangemeJl( showing, ill comparisoll with

the inset sketch, the seat ring well supported alld the valve guide protected
from feed-back of heat from the exhallst pipe area.
26
27
C AND A WASHER
FINNEO NUT
INSUFFICIENT BACKING
IIEHIND SEAT RING
TUNING FOR SPEED

~;ever
remove metal in any but the smallest amount from

the seat ring itself, and also leave plenty of metal under the
ring where it is closest to the port, as there is a possibility
that explosion pressure may actually force the seat-ring
back due to lack of support if there is insufficient portmetal backing up the ring at this point; the valve will then
cease to be gas-tight and may shortly fail through overheating.
It is not always appreciated that one of the most prolific
causes of trouble in I.-c. engines is the large amount of
waste heat absorbed by the exhaust-port walls and fed
back into the head. Once the exhaust valve has opened,
any heat left in the gas ceases to be of any value and
becomes a waste product, and the less it is re-absorbed
into the engine the better. Consequently, exhaust ports
should be as short as possible, and there should be a path
for beat conduction between the port and the exhaust pipe
nut, but not between nut and pipe. These conditions are
all assisted by using a male-threaded nut, with a perceptible
clearance between the bore of the nut and the pipe; at one
time long ports furnished with many fins to cool the port
down were fashionable, but it is far better to prevent the
heat being absorbed rather than to dissipate it afterwards;
an examination of almost every engine which has been a
successful performer on petrol or petrol-benzole will show
that they all possess this feature of short, direct ports.
The point is ofless importance on engines designed specifically for alcohol fuel, but that does not impair the essential
correctness of the principle.
Double-port heads which were popular some years ago
are, in general, not nearly as efficient as those with single
ports, and most engines so designed will perform better if
one of the ports is blanked off completely and as close up
to the valve as possible. The term "double port" is used
here as applying to a single valve; if there are two exhaust
valves, it is usual for each to have its own port, and in this
event the criticism, of course, does not apply.
Once an exhaust port has been correctly shaped and
28
polished, it is best not to clean the carbon out of it, but

to allow a thin skin to remain. Carbon is a poor heat
conductor and acts as quite an effective insulator,
which helps to limit the absorption of waste heat by the
port walls.
Python engines (the name under which Rudge power
units were supplied to other makers) were sometimes
equipped with adaptors forced into the ports to suit Ii-in.
pipes (far too big, even for the 500 c.c. models), and these
should be removed. This is done by removing the hexagonheaded setscrews which lock the adaptors in place; the
latter can then be twisted out with the aid of a large Footprint pipe-wrench, an operation best conducted before
detaching the head from the cylinder. The right pipe size
is that which just fits in the counterbore in the port from
which the adaptors have been removed.
Much of the heat absorbed by the exhaust valve head gets
away by conduction through the seat; the latter should,
therefore. be reasonably wide-between ~ in. and tin.,
according to valve size. For sprint work, or on engines
which will be regularly stripped down after every event,
narrower seats can be used with safety, but more frequent
valve-clearance adjustment may be necessary. It may pay on
heads originally designed with both valves of equal size to
reduce the diameter of the exhaust valve, and, therefore,
the seat width, in order to fit a larger inlet valve and still
maintain enough clearance between the two to prevent
mutual interference in the overlap period when both are
partially lifted. A slightly convex head is preferable to
one which is dead flat or even concave, although it means
a slight increase in weight. The edge must not be too
thin, otherwise it will get very hot and may warp out of
shape at running temperature; as with the inlet valve, the
top edge should be slightly rounded off to improve gasflow and to avoid a local hot area.
If there is any doubt as to the quality or age of the valve,
play safe-discard it and get a new one of K.E. 965 or
similar austenitic steel, although these materials are difficult
29
..
"!
:.
"
.~!
,;"
TUNING FOR SPEED
in stem diameter, and have cast-iron guides; an improvement in heat dissipation can be effected by increasing the
stem diameter by
in. and using aluminium-bronze or
Barronia-metal guides, but it will be necessary with these
metals to ensure a small though positive supply of lubricant
to prevent seizure. A flash-coating of hard chrome plate
.0005 in-.OOI in. thick on the valve stem will also assist
in this direction.
Absorption of heat by the head is lessened by contouring
the valve chest so that the outgoing gas is allowed to flow
away as smoothly as possible; this is achieved by eliminating sharp corners in the valve chest and allowing plenty of
room over the head of the valve when it is at full lift.
The scheme previously outlined for deliberately masking
part of the inlet valve is definitely not applicable to the
exhaust valve, for it is imperative that the whole circumference of this component is brought into action at the
commencement of lift in order to liberate the high-pressure
gas immediately.
nr
RIGHT
IE~
Correct and incorrect techniques ill the process

of izard-jacillg m/l'e seat areas.
to come by to-day. Better still, if possible get hold of a

sodium-filled valve of the type commonly fitted to aeroengines and used by our leading racing firms during the
last two or three years of racing before the war. These
valves are larger in the stem than the " solid" ones they
replace, and in consequence the guides will need to be
renewed or bored out. Valves which have worn thin near
the seats but are otherwise sound can be reclaimed by
building up the seat with "Stellite" No. 6 grade or
.. Brightray". Both these alloys are non-ferrous and have
the advantage that they are very resistant to attack by the
by-products of Ethyl fluid; in countries where heavilyleaded fuel is obtainable and permissible for racing it is a
wise precaution to have the valves" Stellited before they
become damaged by lead attack. There are other materials
such as "Cobalite" which can be used for hard-facing valves,
but it must be remembered that the deposited material
should always be backed up by the parent metal of the
valve, as shown in the diagram; it is inadvisable to use hardfacing to increase the diameter of the head or to build up a
very badly worn head to its original
H
Side Exhaust Valves

Side exhaust valves are worse off than the o.h.v. type
as regards heat, because they get little cooling from the
incoming charge, and the part of the seat next the barrel is
bound to run very hot. Also they afe usually rather small
30
Valve Seat Angles

The majority of valve seats are formed at 45 0 to the axis,
but some makers favour the flatter 30 seat, which is more
effective at low lifts because the actual openin'g between
valve and seat is greater by 22% or in the ratio of 86
to 70. This advantage ceases after the valve has lifted by an
amount approximately equal to Ii times the seat width, and
is, therefore, of less importance with narrow seats than
with wide ones. But with the widths commonly employed
the gain is worthwhile because much of the breathing must
perforce be accomplished when either valve is only partially
lifted. Conversely, the 45 seat has a greater self-centring
effect on the valve, and may give better results mechanically
in heads with short guides, Of those in which distortion may
occur at high temperatures.
0
31
CHAPTER III
So far we have progressed to the stage where the valve

seats and ports have been attended to-at least, so far
as the written word is concerned: carrying out the work in
a really thorough manner takes a much longer tBne and
far more patience than simply describing it.
Before proceeding further, the carburetter should be
fitted temporarily to make sure that it lines up accurately
with the port without a ,. step" anywhere around the
circumference. If the mixing chamber pushes on to a stub,
or has a spigoted flange. the two bores are almost bound to
be concentric, but a " step" will exist if the diameters are
unequal or if the port has been filed a trifle out of round
during the cleaning-up process.
Once lined up, either of these types will always reassemble
correctly, but the standard flange. which has no spigot,
is sometimes a rather slack fit on its two studs. If the studholes are worn or have been filed out at some time, it pays
to ream them out to i in. and fit bushes on to the studs;
Showing how all old~type inductioll stub can be modified to

accept {J modem flange-:!ittil/g carburelter. Note holl' the
inductioJl pipe is blended smoothly into the inlet porI.
32
the carburetter and porl bores can then be blended into

each other and you will be sure that, even if the carburetter
has to be refitted hastily at some future date, it will go
back correctly and cannot possibly work out of position at
any time through vibration.
Sometimes these flanges become bent (usually through the
use of a packing washer which is too soft and thick), and
this fault must be rectified by truing up the face. For use
in later experiments, it is worth while making up some
extension blocks to go between the flange faces, in order to
vary the length of the induction system; the easiest way
is to machine them from aluminium c;;st'ngs made from a
wooden pattern the same shape as the flange and i-in.
finished thickness. Four of these, plus the necessary longer
studs, will enable the induction pipe length to be varied in
i-in. steps up to 2 in.
The cylinder heads of early models were fitted with
screwed-in stubs to take clip-fitting carburetters, and the
engine makers concerned can often supply longer patterns;
alternatively, in order to use a flange-fitting carburetter, a
steel flange can be welded or brazed to the stub.
It may be advisable to make up a new stub altogether
rather than to modify an old one; many induction systems
are neither long enough nor large enough for maximum
power at very high f.p.m. (see Chapter XVIII). New pipes
can be made up quite easily from steel tube with flanges
brazed or welded on. Another simple method is to use a
piece of canvas-reinforced hose attached to the existing
induction pipe and to an extension piece by Jubilee clips,
though some care has to be taken to avoid a nasty step
or change in diameter at each change from pipe to hose,
particularly if an inch or so of hose is exposed; in that case,
the best thing is to line the hose with a third piece of
metallic tube to preserve a constant diameter. This scheme
insulates the carburetter from the high-speed vibrations
which sometimes causes a mysterious loss of r.p.m. through
unsuspected flooding at a certain critical speed and so is
referred to again in Chapter XVIII.
33
TUNING FOR SPEED
Amal carburetters are made in both flange- and clipfitting forms. The largest bore which can safely be used
with the standard 2-in. stud centre-distance is iii in., and
even then there is only just clearance for the nuts. Clipfitting carburetters, i.e. those which fit on stubs, do n?t !13ve
this limitation, and ""hen for reasons of space or to ehmmate
bends in the induction system, the mixing chamber has to be
mounted out of the vertical, the clip-fitting type is easier to
install because it can simply be swivelled round the stub
into the correct position. The angle of the mixing chamber
is immaterial, with limits, provided the float level is correct
in relation to the jet, though if the inclination is great~r
than 15 flooding from the pilot jet into the engine WIll
occur when standing if the fuel supply is not turned off.
Standard stub diameters are Ii in., It in., and Ii in.,
according to the bore of the carburetter. Adaptor stubs with
any of these diameters can easily be made up to fit on
standard flange studs, but it is best to use socket head
screws of the Allen or . Unbrako" type with the largest size
to avoid having to cut too far into the wall to obtain head
clearance. By this method it is also possible to make up a
stub of 1.422 in. diameter to utilize the 32 mm. carburetter,
which is the largest size commercially available at the
time of writing.
.
That just about completes the work on tbe head itself,
and if you are really satisfied that all is as it should be, the
next thing is to examine the valve springs, collars and split
collets, to see if any improvement can be effected.
Coil and Hair-pin Springs
Hair-pin springs came into use many years ago for racing,
because the coil springs in use at that period were not
always reliable and became less so as engine speeds rose
above 6,000 r.p.m. and valve lifts increased above i in.;
the combination of these two factors creates a difficult
situation for any spring and even hair-pin springs have
undergone considerable design modifications before reach34
ATTENTION TO O. H. V. CYLINDER HEADS
ing their present state of reliability. Most sports models

and genuine" racers" of 1934-39 vintage were fitted with
hair-pin springs with the notable exception of the speedway
J.A.P. which has always retained coils, these being, in any
case, sufficiently reliable for an engine intended primarily
for short-distance events. Since 1945 there has been a
considerable change back to coils, largely because of the
difficulty of enclosing hair-pins without serious interruption to air-flow through the head fins, particularly on a
parallel twin. Coil-spring fracture is mainly caused through
" surging" or a state of rapid vibration of the centre coils.
and the way to overcome this is by using only a limited
number of coils; all modern coil springs are duplex or
even triplex with not more than six free coils in the outer;
the inner spring often has one or two more coils of lighter
gauge wire and it is an advantage for it to be a push fit
within the outer one as this provides a certain amount of
friction damping and so helps to prevent surging.
Side-valve engines are usually equipped with single
springs of fairly light poundage, sometimes of the" floppy"
type. with a large number of coils. These should be replaced by others of similar proportions to those just
described, which will probably entail replacing the top
and bottom spring caps and the split collets. It is also
probable that pieces will be required to make up for the
reduced length of spring and these should be made with the
smallest possible area in contact with the head to minimise
transfer of heat to the uppermost coils.
With regard to spring pressure, it is a mistake to think that
enormously strong springs are a sine qua non; any pressure
in excess of that required to prevent valve chatter up to the
usable maximum revs.-say,6,500 for a 500 c.c. unit, up to
7,500 for a 250 c.c. plot-simply wastes .a little more power
and loads up the valve gear unnecessanly.
The actual strength required at full lift depends upon the
cam contour and lift, as well as the revs. and weight of
valve gear, so it just is not possible to give any hard-andfast rule, but, apart from these considerations, it is usual
35
A TTENTION TO O.H. V. CYLINDER HEADS
TUNING FOR SPEED
in two-valve heads to provide 100-120 lb. pressure when

the valve is seated, and, unless the cam design is very
peculiar, 140-150 lb. should suffice at full lift (the greater
figures obviously applying to the larger-size engines).
Four-valve designs, with their much lighter individual components, require only about 80 lb. seat pressure per valve.
It is possible to make up a rig to test this pressure but
makers' genuine replacements can be accepted as giving
the correct amount provided that nothing has taken place
to alter the installed length of the spring appreciably.
Excessive seat wear, or possibly a stretched exhaust valve
~ay account for this length being
in. or so longer tlian
It should be and this is sufficient to reduce the pressure by
several pounds. For workshops which do a considerable
amount of tuning work, it is a wise plan to acquire one of
the special spring-testing machines that give an accurate
measurement of the poundage given at any particular length.
The actual installed poundage can then be determined by
first measuring the distance between the top and bottom
spring collars and then determining the pressure of the
springs at that length on the machine.
Apart from being obviously broken, coil springs often
~horten or settle an? so ~ose their eff~ctive strength; attempt~ng to re~tore the s~tuatlOn ~~ packIng up the lower seating
IS som~tImes effect\;,e, ~ut It IS then essential to verify that
the spnngs are not cOil-bound" at full lift; to do this see
that. the valv~ can be. raised a further
in. beyond the
maximum POInt to which the rocker lifts it.
Another expedient is to place the weaker set of springs on
th.e inlet valve, as insufficient strength on this component
wIll at worst only reduce the top r.p.m. a little, but severe
mechanical damage may result from lack of spring strength
on the exhaust valve.
to
n-
Hair-pin Springs
. Hair-pin ~prings were originally made with fairly small

diameter COils and long legs, but the modern variety has
36
much larger diameter coils and shorter legs. Further, the

design is. such that over the working range of the spring,
the loop In contact with the valve moves almost in a straight
line parallel to the valve stem and there is little or no
tendency, as there was with earlier designs, for the stationary
legs to move in and out on their abutments. When for
reasons of space it has been necessary to reduce the overall
width of the pair of springs, as, for instance, on parallel
twins, like the twin-cylinder A.J.S. racing machine, the ends
of each pair of springs are overlapped, and careful design
is necessary to ensure that no interference occurs between
the springs at any part of the valve movement. When
fitting these overlapping springs, particularly to an engine
for which they were not originally designed, care must be
taken to see that the end loops do clear each other at all
points from closed to full lift.
After a while, wear occurs on the straight piece of the
loops of hair-pin springs, caused by rubbing against the
spring retainer, and replacement is called for if this wear
exceeds one quarter of the wire diameter. The force exerted
by the springs in each pair should be equal to avoid tilting
~he valve, therefore it is essential to replace in pairs, not
Ju~t .one at a time. When they do break, hair-pins usually
fall Just where the top leg commences to bend round into
the coil and every care must be taken to see that no cuts
or deep scratches are made in the wire at this point when
handling such springs.
I~ is ofte~ c~nsidered advisable to convert a coil-spring
engme to halr-pms and this can be done with a little wangling
in some o.h.v. designs. On MOV and MAC Velocettes of
the original design with bolted-on rocker boxes, standard
KTT springs, and spring collars can be used, but the rockerbox extensions which normally surround the spring cups
have to be cut away, leaving the outer ends of the rockers
exposed, and the rocker-box standards on the push-rod
side must be thinned down to clear the springs which are
installed at an angle .
Removal of the50il valve spring enclosure has the effect
37
TUNING FOR SPEED
of deprivi~g the valve guides of lubrication which they

normally receive by leakage from the rocker bearings and it
is, therefore, desirable, particularly for long-distance work,
to drill lil in. holes through the head into each valve guide
to receive oil pipes which can be supplied with lubricant
either from a small external tank from the normal lubrication
system with metering jets to limit the supply, which need
only be small in quantity.
Auxiliary Return Springs
Some designers prefer to utilize return springs on the rockers
or push-rods, thereby either reducing the strength necessary
in the valve springs or, alternatively,permitting higher revs
for the same spring strength; examples are the hair-pin
springs surrounding the rocker bosses on KTT Velocettes
and the springs fitted on the lower ends of the push-rods
of certain J.A.P. motors. The effect of such springs is to
maintain the cam follower in contact with the cam cantinuously, and excessive pressure leads to local heating and
wear of the cam-follower foot unless a roller is employed.
A neat way of adding return springs to engines with hollow
rockers is to house a coil spring loosely in the bore of the
shank with one end fixed to the rocker arm and the other
attached to the rocker enclosure in such a way that the
spring can be wound up to any desired amount; this scheme
has the merit that it is easy to experiment with varying
A method oj obtaining illcreased

return-sprillg ac(ioll without crea(ing additiollal camJollower loading.
./
.'\.
.'{
38
ATTENTION TO O.H.Y. CYLINDER HEADS

MEAN DIAH.
I. OF SPRING i
W'H'C<Nm
I ' _. _._.-",!
I.
I,, EXTRA
,1" ..- GREATER DIAM
(LeJt) At maxilllum rall'e liji there II1l1st be adequate clearance between vall'e guide alld spring cap and between the
spring coils. (Right) Note how the outer diameter of. the
alloy cap is that oj the mean diameter oj the aliter spring.
amounts of pressure and thereby to discover whether

stronger valve springs are required.
Another method is to attach a tension spring to a short
vertical arm on each rocker; this idea is good because when
the exhaust valve is closing, the inlet is lifting and, therefore,
the tension in the spring is increased just when it is most
wanted, and the required effect is gained without excessive
cam-follower loading during other parts of the cycle.
In some designs the exhaust spring settles very quickly
because the coils nearest the exhaust port become overheated; a remedy can be effected by insulating the lower
spring coils by mica washers, or packing the spring seat u.p
on narrow washers so that there is the minimum of metalhc
contact and the maximum of air space between the spring
seat and the hot port metal. Generally speaking, enclosed
coil springs can be kept as cool or even cooler than open
springs by the expedient of pouring enough oil over them
through jets to carry away the heat.
Before leaving this subject of springs it may be well to
dispel a popular misconception as to their function: some
think that the cam pushes the valve open against the spring
right up to full lift and after that the spring takes charge
.and performs the closing action. In fact, the cam operates
against the spring plus the inertia of the valve for the first
portion only of the lift; after a certain point and up to full
lift, the spring is doing all the work of slowing the ~al~e
.down until it is actually stopped by the time full lIft IS
39
TUNING FOR SPEED
reached. After that, the spring accelerates the valve back

again until a certain point is reached on the closing flank
of the cam; from that point onwards the valve is being
slowed down again by pressure between the cam-follower
and cam so that it is eventually let down on to the seat
at a .very low velocity. In other words, the valve is under
cam-control for the first portion of the lift (whilst accelerating) and for tbe remainder of the motion, i.e. over the nose
of the cam, the valve is under spring control. The pressure
between the cam and follower during periods of cam-control
is equal to the spring pressure plus the inertia force or
momentum of the valve gear, whilst during the periods of
spring control, the pressure is the spring pressure minus
the valve inertia, and if at any point the inertia becomes
greater than the spring pressure the cam follower will leave
the cam contour giving rise to the condition known as
valve fioat. This may not be very noticeable, in fact if
the cams are inadequate in lift or duration, may even be
beneficial from the power production point of view, but if
the speed rises still further, the separation will become
greater and greater until the valve bits the piston or the
other valve and suffers in the process, or else very heavy
impact loads (as indicated by th.e onset of excessive noise)
will be set up somewhere in the valve gear with the certainty
of failure if the conditions are allowed to continue.
Touching on the matter of spring caps, many can be
lightened judiciously by reducing their diameter to the
The simple construction

of tire J.A.P.-type split
collet and spring cap
assembly recommended
by many racing men.
40
ATTENTION-TO O.H.V. CYLINDER HEADS
An 0 ver-Iong valve
stem causes thrust
to the righl.
mean diameter of the outer spring and filing scallops in the

edge, so that tbe spring bears only on six or eight projections. Alternatively, new caps can be turned from a
strong light alloy (preferably R.R. 77, but, failing that,
R.R. 56, or Duralumin B) to the same dimensions as the
original steel ones, except for allowmg a little extra thickness where the flange merges into the taper bore, and
around the collet boss.
If you are rejuvenating an old engine and are using nonstandard valves, the easiest way to retain the spring is by
using a spring cap, collets and wire circlip of J.A.P. design,
since all the machining called for is to turn in each stem a
narrow semi-circular groove to accommodate the circlip,
and to verify that the stems accurately fit the collets, which
are obtainable with bore sizes of -& in., Hin., and i in.
The position of the upper end of an o.h. valve-stem is
important, but is difficult to check dimensionally without
special measuring fixtures, so it will need to be done by eye.
Assemble the rocker gear on the head and observe the
behaviour of the rocker-tip relative to the valve stem as the
mechanism is operated by hand from the closed to the fulllift position. If the stem is too long, the contact point will
If the stem is too

short the thrust will
occur 011 the opposite
side to that indicated
by the above sketch.
41
TUNING FOR SPEED

.-::;.
ADJUSTER IN LINE
.04--
WITH VALVE
Just right! A valve

stem of the correct
length has the
rocker at right
angles to it, with
the adjusfer dead
central, whel/ ill the
mid-lift position.
move outwards, as shown in the diagram; if too short, the

movement will be in the opposite direction. If the outer
end of the rocker is fitted with a screwed adjuster, this should
be exactly in line with the valve stem in the half-lift position
with the contact point dead central to the stem at the same
time.
There is little likelihood that a fairly new engine using
standard components will not be correct, but numerous
causes-excessive cutting-back of the valve seats, stretched
or incorrect valves, or damaged rocker boxes-may all
contribute their quota to incorrect alignment in old power
units. Sometimes a cure can be effected by shortening the
stem or fitting hardened end-caps of the appropriate thick:gess; in otlter..cases packing up the rocker box will do the
:rick. The point may seem to be of little importance, but
It must be remembered that, however strong the springs
may be, as maximum revs. are approached they have less
and less margin of strength over that required to return
the valves to their seats; thus, an amount of friction anywhere in the whole of the valve gear, which is barely
perceptible at hand-cranking speeds, becomes a serious
item at top revs.
The clearance at full lift between the valve guide and
spring collar must also be checked. This should be a full
~ in., otherwise, if the motor is over-revved, the parts
may hit, and no valve will stand up to impact of that
nature for very long. Similarly, if coil springs are used,
they must not be "coil-bound," i.e., compressed solid
at full lift, which might occur if the bottom washers have
42
been packed up to increase the spring pressure, or if incorrect springs are installed; this point is most important.
As a general rule, it is best to defer the assembly of the
head until any necessary alternations to the barrel or piston
have been effected; the subject is dealt with in Chapter X.
Such modifications may be required to obtain the correct
compression ratio, or to ensure that the valve-clearance
"pockets" in the piston crown are deep enough and of the
correct radius.
The correct compression ratio depends upon the cylinder
capacity, type of fuel, material of the head and many other
items, such as stroke-bore ratio, effectiveness of the induction system, and so forth, but the following table gives an
approximate guide to the ratios which can be employed
with safety in o.h.v. engines under racing conditions when
using 50/50 petrol-benzole fuel, although they by no means
represent the ultimate, or necessarily the most suitable,
figures:
Cylinder capacity
250 c.c.
350c.c.
500c.c.
Cylinder-head Materia!
Alum.
Alum. or
Iron
Bronze
Bi-meta!
8.5
9.5
10
8.2
9
9t
7.8
8.5
9
Elderly engines with finning a bit on the scanty side may

not be able to use ratios quite as high as those quoted, save
Tile benefits of a
higher compression
piston can be lost if
lire resultantcombusliOll space becomes
ullduly
distorted.
Better results will
be attained with the
lower piston dome,
illustrated here by a
dotted line.
43
TUNING FOR SPEED
for very short-distance events or when using alcohol fuel.

Ratios as high as 16 to 1 can be used with fuels such as
P.M.S. 2, J.A.PAuel, or methanol, irrespective of the
cylinder size, but in " square" engines-i.e . those in which
the bore and stroke are equal, or nearly so-the piston may
need to have such a pronounced dome on it that the shape
of the combustion chamber becomes unfavourable to good
combustion.
In such cases better results may be obtained by keeping
the ratio down to, say, 12 to 1. Of course, " dope" fuels can
be used with ratios suitable for petrol-b,enzole, but the
engine will not give any more power-it will simply run
cooler and be less liable to give trouble with valves or plugs.
This subject is a wide one and has a c;hapter devoted to it
later on.
Assuming that the compression ratio and valve clearances
are correct, final-stage work on the head can be commenced
by polishing the valve stems with dead-smooth emery cloth
(rubbed along the surface, not round it) and then grindingin in the normal manner, but using a fine abrasive, finishing
off with powdered Turkey-stone or metal polish to obtain
really smooth seatings entirely free from ridges or grooves.
During this process the inlet-valve seat, which has been
in. wide through' rounding-off the edges of
reduced to
valve and port (see page 19), will increase slightly in width
to 45 or 50 thous., at which it is wide enough to run for
some time without undue hammering back. It is best to use
some flexible device rather than a rigid tool for oscillating
the valve so that it can seat itself freely; a rubber suction-cup
on a wooden handle fills the bilI, though it may be necessary
to apply a drop of Bostik to the polished valve head to
, obtain sufficient grip to turn it. Another method is to push
a piece of tight-fitting canvas-lined hose on the end of the
stem and oscillate it, either by twirling between the hands
or with a hand drill. When grinding in Vincent valves the
upper guide must be in place, for the lower one is too short
to locate the valve accurately by itself and is not intended
to do so.
'2
44
After grinding in, wash away aU traces of abrasive with

petrol or paraffin, smear a drop of clean mineral ~il on the
seats and assemble each valve in turn-first applymg some
graphited running-in compound to the stems to. mil~.imize
the risk of scoring or "picking-up" when the engme 15 first
started.
Make sure that the split collets are paired up correctly,
and after fitting the springs with the aid of the invaluable
Te~ spring compressor-an essential part of ~e shop
equipment--verify that both collets are hard agamst the
abutment on the stem and tbat the spring cap is fitting
snugly on the taper; if not, one or two light blows with a
hide mallet or the end of a hammer-handle on the spring
cap (not the valve stem) will usually jar the assembly into
position. If you are in any doubts as to the fit .of the collets
in the spring cap, it is advisable before assemblIng the valves
to lap these components together in position on ~he stems,
using ordinary valve~grinding paste as the abraSIve.
.
Wire circlips sometimes become stretched when beIng
removed and it is best to close them up before reassembly
and then to make absolutely sure, even by using a magni~
fying glass, that each is actually seated in the groove an~
nestling snugly in the recess provided in the top of the spht
collets when the full pressure of the valve spring is on them.
Occasional cases of valves dropping in with this form of
fixing are usually attributable to failure to observe this point
during assembly.
Split collets are often supplied in pairs, partially cut
through. They should always be left in this cond~tion until
required, then separated and the.s~an burr l~ft m the ~ut
cleaned up prior to fitting. MlXlng up splIt collets Indiscriminately is a fruitful source of valve failure, though
it may take some time to develop.
Finally, don't leave the head on the back ?f !he be~ch. to
get filled up with dust and swarf. Smear oIl lIghtly mSlde
the ports and block them with pieces .of clean rag, o.r,
better still, large corks, which are less lIkely to be left In
place when they should have been removed.
45

POSSIBLE GAS
CHAPTER IV
Jo,mmNe,
GASKET '/,," THICK

THE previous note on compression ratios suitable for
various engines needs amplification. In the first place,
the ratios quoted are for racing only, being far too high for
ordinary road work with silencer and "pump" fuel, one
to one and a quarter ratios less being about as high as
could be used in the latter
circumstance.
Owing to the shape of the
combustion chamber, it is impossible to obtain such high
ratios in a side-valve cylinder
without restricting the breathing ability-about 7 to 1
being as high as one can
reach-and it must always be
remembered that it is not the
actual measured compression
ratio which counts, but the
pressure of the gas at the end of
the compression stroke. This
depends as much upon the
perfection of cylinder filling
as on the ratio; therefore a
balance must be struck between the interior shape which
gives the best gas-flow and that
which gives the highest ratio.
Scallfy /illning, characteristic
of design of the 1920's, is still
satisfactory where cool-running
methanol and similar fuels are
used.
...
1
""~ __ '_'_"_"_"T ._- _ _ _ _ .r--~._. _~r.__r.--_~_ _
. _I
iiiii. . . .
PISTON CROWN LEVEL

WITH TOP OF
CYLINDER BLOCK
SLIGHT RADIUS
ON CORNER
OF BORE'
Points which must be checked when a side-valve

cylinder head has been machined 10 raise the compression ratio.
Most s.v. en.gines are now of the detachable head pattern,

with the combustion chamber concentrated over the valves,
and in this type the throat between chamber and barrel
sometimes constitutes a restriction, particularly if much
metal has been machined off the head to raise the compression. A small radius on the top edge of the barrel next the
valve chest will improve the flow lines, but must not be
carried so far down that the top ring is in danger of being
exposed. The thickness of the gasket is sufficient clearance
between the flat portioIi of the head and the piston crown,
which should, therefore, be just level with the top face of
the barrel at t.d.c. An aluminium head is oarticularly
desirable, except, perhaps, for "dope" fuel, when th~
cooling problem is not so acute.
Reverting to o.h.v. engines, the aim is to obtain the
correct ratio with the least possible area of metal exposed to
flame. However impressive it looks, an excessively domed
piston crown is not desirable, partly because its exposed
area is great, also because the "hump" hinders flame
propagation, and anything in the nature of the converging
47
TUNING FOR SPEED
pocket which is formed by the angle between dome and

head produces a likely locality in which detonation can
take place.
It is quite a good scheme to make a cast of the combustion chamber by pouring melted wax into the plug hole with
the piston at t.d.c.This shows up any pockets or restrictions, and the wax can be softened and rem"oulded with the
fingers after detaching the head, to give an improved shape
with the same volume of wax and, therefore, the same
compression. As quite a number of different makers have
used the same bore diameters-74 mm., for instance, is a
popular size for 350 c.c. power units-it is often possible
to utilize a "foreign" piston of different proportions to
replace a standard article and thereby obtain a better
shape of combustion chamber.
Unless shims are already fitted under the barrel flange, the
usual method of finally adjusting the ratio is by shortening
the barrel. A great deal of laborious setting-up and
remeasuring can be saved by making up a two-column
table showing the C.R. and equivalent compression space
and determining, as follows, the differences in barrel
length which are equal to 1 C.c.
thickness and free from wrinkles, and should be provided

in assorted thicknesses, as a large number of thin shims do
not provide a firm foundation for the cylinder. It must
be remembered that a ring-ridge may form at the upper
limit of the top-ring travel in quite a short time if the engine
is used on sand or grass tracks and broken rings or ringlands may ensue if the engine is run with one or more
plates removed after running for some time with thick
plates, unless the ridge is carefully removed. For that
reason if an engine has to be repeatedly changed over
from one compression ratio to another it is best to do so
by using the appropriate pistons, making sure that the
height of the top ring above the gudgeon pin is the same
for each piston and that the weights of each are the same,
to avoid need for re-balancing.
Bore
1 c.C.=
60mm.
7fjmm.
80mm.
90mm.
.014 in.
.0lD in.
.008 in.
.006 in.
Knowing the existing ratio, a very simple piece of arithmetic will show how much length alteration is required for
any other ratio. Skimming the barrel flange is an operation
which has to be done very accurately to ensure that the new
surface is dead square to the bore, and some prefer to turn
the top joint face where squareness is not so vital. On the
other hand, a little more than is necessary at the moment
can be turned oft the flange and thereafter adjustments to
increase or decrease the ratio can readily be made by shims.
These should be made up from good material of even
48
Determining Compression Ratios

For maximum power, the rule is to employ the highest
ratio at which the engine will run without distress, on the fuel
which is specified for the race in question. There are, of
course, many factors, both internal and external to the
engine, which have an effect on the ratio which can be used,
but the prime factor to be considered is the quality of the
fuel.
J.A.P. fuel, methanol and similar fuels composed mainly
of alcohol can be run at anything up to 18 to 1, even in
engines with iron heads and barrels and scanty finning, this
being partly because alcohol is almost knock-free and
partly because the large amount of heat required to vaporize
these fuels exercises a great internal cooling effect.
" Dope" engines for use on official dirt tracks are, however, limited by the regulations to 14 to 1 ratio. Even in
events where there is no such limitation, there is not much
to be gained by going any higher as very often the obstruction in the combustion chamber caused by the high pistoncrown required to obtain these ultra-high ratios more than
offsets the slight increase in power which might otherwise
be gained.
49
TUNING FOR SPEED
. With petrol or petrol-benzoic, the internal cooling effect

IS much lower, and detonation is liable to occur above a
certain ratio, depending upon the octane number of the
fueL That is to say, fuel which, on the average, is rated
at 72 octane, will not perform satisfactorily in engines which
have had their ratios pushed up to the limit at which they
would run on pre-war SO/50 petrol-benzole.
The precise octane rating for this latter mixture is not
available, but it is believed to be above 84 octane, or roughly
midway between" MT80" (petrol, plus 3i C.c. of tetraethyl-lead per gallon) and 87 octane aviation petrol. Benzole,
however, has an appreciably greater internal cooling effect
t?an petrol, and seemingly extraordinarily high compresSlons--over 10! to 1 in some instances-were used in the
1938 and 1?39 T.T. races, the engines concerned being the
last :v~rd III development, and possessing heavily finned
aluminIUm heads and barrels, sodium-cooled exhaust valves
and copious internal oil circulation, with jets to the piston
crowns. Similar engines had to be dropped down to 7.5
to 1 or lower when using "Pool" which is an obvious
illustration of the great effect exercised by the quality
of the fuel.
Pool spirit as supplied in the United Kingdom after the war
had a maximum octane rating of 72, but no fixed analysis;
on occasion, up to 1 C.c. of tetra-ethyl-lead per gallon was
added to low octane fuel to bring it up to the required
st~ndard. In 1950 F.I.M. regulations were altered to perIll1t the use of 80 octane fuel for international events; that
octane value may be obtained by a benzole or alcohol
additive and on this quality, compression ratios in the region
of 9 to 1 may well be employed in a well-designed 500 c.c.
engine.
The great internal cooling effect of alcohol fuel is offset
by the much higher consumption; the mileage per gallon
on pure methanol is only about one-third of that obtained
with petrol or petrol-benzole. This is naturally a serious
matter in long-distance racing, as apart from the question
of expense, it entails either a vast tank capacity or too many
50

COMPRESSION RATIO
SUITABIUTY (500 C.C.
SINGLE CYLINDER
AIR-COOLED ENGINE)
GRADE Of fUEL
Up to 8.5
9.5
" "
" 10.5
" .. 12.5
"Over 12.5
TI
X
Y
M
A
811
12.5
,INCREASED FUEL. fLOW

COMPARED WITH
STRAIGHT PETROL
Nil
35%
80%
60%
130-140%
100%
pit stops, but a compromise may be effected by using

alcohol blends such as .. 80/l0/1O" composed of 80 per
cent methanol, 10 per cent benzole and 10 per cent petrol.
Blending methanol with petrol is not quite a straightforward matter; the methanol must be of the quality known
as " anhydrous" or "dry blending" methanol, and even
then it will only mix with petrol if at least equal quantity
of benzole is added as well. Some riders for instance employ
a mixture composed of equal parts of methanol, petrol and
benzoIe, which will run very satisfactorily with a reasonable
COMPRESSION RATIO
BLOWER PRESSURE
Unblown
+ 2 lb. per sq. in. TI
TI
4
j: TI
6
II X
8
II
X
10
12
;1 M
M
14
M
16
M
18
M
20
Ii
TI
X
X
X
M
M
M
M
M
M
Types of specially-blended Australiall Shell fuel suitable for

normally-aspirated and supercharged engines ..
51

TUNING FOR SPEED
consumption on ratios around 10 or 11 to 1. In Australia

the Shell and Vacuum Oil Companies supply several
grades of racing fuel ready mixed, and the list of these
on page 51 with their approximate analysis and suitable
compression ratios gives a fairly clear picture of the
situation.
Before leaving this question of fuel, which is really too
complex to deal with at all fully in this volume, it is worth
mentioning that a great deal of nonsense is talked about
special" dopes". The amount of energy liberated in the
cylinder per cubic inch of correct mixture is very much the
same, whatever combustible substance is employed, the
theoretical quantity being 48 foot-pounds and no amount
of juggling with various fuels will alter this fact. The
reason why alcohol gives greater power is two-fold'
(a) it can be run at higher compressions and therefor~
at . greater efficiency and (b) the high latent heat of
evaporati~n of alcohol cools down the fresh charge
very consIderable so that a greater weight of charge is
induced into the cylinder: in other words more mixture
is induced and then burnt at higher efficiency, hence a
:-V0rth-w~ile increase in power, though the percentage
mcrease IS not nearly as great in practice as it should be
in theory.
An exception to the statement about dopes must be
made in the case of oxygen-carrying compounds such as
nitro-benzine, which has recently come into prominence,
first as a fuel for miniature engines for model aircraft and
then as a racing fuel. Without delving too far int~ the
chemistry of the matter, this compound contains an excess
of oxygen and when added to the regular fuel permits
much larger amounts of the latter to be burnt and thus
provides much more power. Naturally the co~sumption
lllcreases greatly and as nitro-benzine costs several pounds
per gallon, the whole business is very costly. Also the stuff
itself is very unstable, chemically and all in all it is probably
a good thing that it is not as yet permissible in international
motorcycle road-racing.
52
Finding Compression Ratios

To provide an easy method of finding out the C.R. given
by any particular combustion space volume (or vice versa)
in conjunction with cylinders of various capacities, the graph
on this page has been drawn; three curves represent the
most commonly used cylinder (not total engine) capacities,
the 125 c.c. curve is added for" Liddler" owners or those
interested in 250 twins or 500 c.c. fours. To find the ratio
given by any known combustion chamber volume, read
across from the volume figure to the curve of the cylinder
capacity concerned and then downwards to the ratio line.
12
0
II
o \
100
Vi'
I'\~ 720 CTANE

OCTANE
09 9
I
(POOL
~8 01\
l:70 \
:J
...J
060
>
t'j 50
I\.
~ 10
0(
r\ ~.T. 80
"
~
d 30
"""" r---....
20
'"
"l
f'-......
1'1'....
l""~
1\
rf~c:
-M....
I MAXIMUM RATIO
FOR DIRT TRACKS
~~
.....
1_
~sOo
\~
........
10
5
50~0 PETRO~-BE~:Z:OI
12S'C.C.
....... r0-
10 II 12 13 14
COMPRESSION RATIO
15
t-- ....
16
17
18
Rl:lationship between combustion-space volume and compression

ratio in the most commonly used cylinder capacities is readily
seen/rom this graph.
53
TUNING FOR SPEED
Conversely, to find what compression space is required

for, say, 7t to 1 in a 250 c.c. cylinder, read up from 7.5 on
the ratio line to the 250 curve, then across to the volume
line, where it will be seen that the required figure is 38 C.c.
If it is desired to raise the ratio to 8, for which the volume
should be 35-! c.c., the amount of packing (or material
which has to -be removed from the cylinder base) can be
found by reference to the second graph, which will be
mentioned in a later paragraph.
The diagonal lines drawn across the capacity curves
indicate roughly the maximum ratios which can be used
with various grades of fuel, though (except in the case of
alcohol) the indication is necessarily only approximate
owing to the aforementioned varying factors. An engille
with an iron head and barrel and poor finning cannot
employ nearly as high a ratio as one with an aluminium
head and extensive finning, unless the latter's breathing is
so poor that the cylinder does not fill properly at maximum
r.p.m., and then its actual compression pressure is very
much less than the measured value.
For the same reason, reduced cylinder filling, engines
which are to be raced at high altitudes, and consequently
low atmospheric pressures, will stand a higher ratio than
when run at sea-level, which is worth remembering in some
countries where competitions are held in mountainous
regions. A damp atmosphere also helps to suppress
detonation.
Detonation the Limiting Factor
A peculiar thing about almost all engines is that with any

particular fuel there is a limiting ratio above which one
dare not go without an almost certain risk of piston trouble,
yet the ratio at which the engine will run happily is frequently only very slightly lower; dropping from, say, 7.8
to 7.5 may make all the difference between success and
failure with but little or no diminution in performance.
Apa~t from audible" pinking," the signs that detonation
has been occurring are unmistakable when the piston is
54
90
i 80
I:
:;;
70
()
z:;
>-
~. 60
I~
"
~
~'" 50
40
'005"
'0 I O'
'015
'"
'020
~~
'025
'030
'035
CYLINDER BASE PACKING EQUIVALENT TO I C.C.

OF COMPRESSION SPACE
Here is shown the thickness of packing equal to 1 c.c. for

given bore dimensions.
examined: on the side of the crown remote from the plug

some of the- carbon film will be burnt away and the surface
of the aluminium will be locally pitted and rough, for which
condition the colloquial description is that" the mice have
been at it."
To some extent detonation due to an excessively high
ratio can be suppressed by retarding the ignition and using a
jet giving a very rich mixture, but the result will only be
diminution of power; in other words, there is nothing to
be gained and everything to be risked in attempting to use
a ratio higher than the engine is willing to accept. This
is particularly true in long-distance races in which the throttle
is wide open most of the time; for small tracks and sand
racing, it is usually possible to go up a little in the ratio,
because as the engine is not running for very long, and in
any case is being shut off fairly frequently, the internal
temperatures do not have time to become dangerously high.
55
TUNING FOR SPEED
The graph on page 55 enables the amount of packing

which must be remo\ed or added when adjusting the ratio
to be determined very easily. Referring to the previous
example, it was found that to raise a 250 c.c. cylinder from
7.5 to 8 to 1 the compression space must be reduced from
38 to 35! C.c. With this second graph, supposing the bore
of the cylinder to be 68 mm., if we read across from 68 mm.
on the vertical line to the curve, and then down to the
horizontal line, we find that .0115 in. is equivalent to I C.c.;
therefore the thickness of packing must be reduced by 2.5. X
.0115=.029 in., or just under 3~ in.
By reading in the reverse direction, i.e., upwards from
the horizontal line, then across, the difference in volume
effected by removing or adding a known thickness of
packing can be read oif directly, thus saving a lot of calculation. With the aid of this graph, a series of plates can
easily be made up so that the ratio can he altered rapidly
to any desired figure.
If the barrel is held down by studs which go through the
flange only, care must be taken to see that the latter is not
made too thin, as it is in a highly stressed locality, and
in overhead camshaft engines with the usual Oldham
couplings in the vertical shaft there is a likelihood that the
small amount of end-float normally allowed may all be
swallowed up-a state of affairs which will put excessive
end-thrust on the vertical bevel shafts. This point must,
therefore, be verified.
On the other hand, if at some future date the barrel is

stacked up again with shims by an appreciable amount, it
may be necessary to obtain a longer shaft or a pair of
coupling-discs of greater thickness. On no account must
these couplings be run with the tongues only half-engaged
in the slots, as the metal is then subjected to a form of
stress which may cause failure.
It goes without saying that the cylinder bore should be in
perfect condition. particularly at the upper end, where the
heaviest wear invariably occurs. The wear practically
ceases about an inch or so below the ridge which forms just
level with the top ring at t.d.c., but, in effect, over this
short length the barrel is tapered. In their attempts to
conform to the varying diameter, the rings have alternately
to expand and contract, a feat which may be beyond their
capacity above a certain speed.
When this occurs, gas will be able to blow past and also
oil can work its way up into the head. Worse still, if the
rate at which the rings are being forced in and out corresponds to their natural frequency of vibration, they will get
into a state known as "flutter," and after a while wil1
break, usually at a point fairly close to the gap. If the ring
design is incorrect, flutter may occur even in a perfect bore,
or may be aggravated by barrel distortion of an unsuspected
nature, which may be caused either by thermal distortion or
through the base flange or crankcase face not being dead flat.
Another snag about altering compression ratios by
omitting compression plates or reducing the base flange is
that the piston ring will inevitably strike any ring ridge
which has formed; occasionally this has even been known
to cause the piston ring land below the top ring to break
away. Even a change from one type of piston to another
may introduce the same effect due to varying ring heights,
so it is always wise to remove every vestige of ridge by very
careful work with a sharp hand scraper or a ring-ridge
remover if any possibility of fouling exists. This point
has already been noted in Chapter II but is sufficiently
important to warrant repetition.
57
c
If tire barrel is attached to
tire crankcase by sharI studs

it is inadvisable 10 remove
metal from the base. Whell

10llg bolts such as those
ilIuslrated on page 126 are
used, a small amount of
metal can be removed from
the base wilh safety.
56
TUNING FOR SPEED

Re~conditioning
Worn Barrels
Badly worn barrels may be rejuvenated by reboring or

by linering, though the first method is rarely applicable
for racing, as mentioned in Chapter I; Norton engines

of 490 c.c. are one exception as they can be bored out from
79 to 79.6 mm. before exceeding the 500 c.c. limit. Not
all barrels are thick enough to stand being linered without
danger of fracture at the flange, although. the type in which
long bolts extend from the crankcase up to the head are
much less likely to fail than those with a bolted flange.
Also it must be remembered that however well the liner
fits, its junction with the barrel constitutes an additional
heat~break, though this is of less moment with alcohol
than with petrol fuel.
Barrels can be brought back to original size by chrome~
plating, though this is a specialized process, as it is necessary for the plating to have a porous surface to provide a
multitude of oil reservoirs; the processes are known by
various trade names such as "Listard," the Van der
Horst patent administered in England by R. A. Lister and
Co., Ltd., Dursley, Glos., and" Honeychrome," a speciality of Monochrome, Ltd., Alcester. Laystall Engineering
Co., Ltd., London, S.E.!, market a range of liners under
the name of "Cromard," these being very thin mildsteel tubes with chrome-plated bores supplied finished to
size; all that has to be done is to bore the barrel to the
-correct size and press the liner into place, no subsequent boring being required. Included in the range is a
liner, specially designed for fitting to the speedway J.A.P.
engine.
This liner is slightly reduced in diameter near the lower
end and the cylinder must be machined with a corresponding step in the bore, so that there is no possibility of the
liner, which has no flange at the top, being pulled down
:into the crankcase should it come loose or if the piston
seizes badly. Such an occurrence is nothing short of a
calamity and, where possible, liners should always be pro-
58
ADJ USTING COMPRESSION RATIOS
vided with a flange at the upper end which can be gripped

between the head and barrel and thus prevent the liner
from moving in any direction.
Other points which must not be overlooked are the drilling of any oil-holes and cutting out any con-rod clearance
slots which may be provided in the barrel originally.
If the skirt of the barrel, which projects inwards into the
crankcase, is long and of fairly thin section, as it is for
instance, in MOV and MAC Velocettes, it is a wise plan
to reduce the liner diameter at the lower end so that it
is just the same size over the length of the skirt as the
bore into which it is pressed. This eliminates the danger
of the skirt developing a crack when the liner is pressed
into position.
Incidentally, liners in aluminium cylinders should be
fairly thick-not less than i in., and preferably more-to
resist distortion, but when used in an iron barrel they may
be made as thin as
in. Provided a locating flange is
present, an interference fit when cold of .001-.002 in. is
sufficient, irrespective of diameter, as the liner is bound to be
hotter than the jacket when running, and it will never
loosen. In aluminium jackets a greater interference is
required, and allowance must be made for the high expansion, on the following basis, according to the aHoy used:
nr
High silicon (L.33, .. Alpax," " Lo.-Ex"), .0009 per in. of diameter.
Y Alloy. R.R. 50 or 53, " Birmabright," .0011 per in. of diameter
Magnesium Alloys (EJektron. Magnuminium), .0016 per in. of
diameter.
After a short period of running, liners in aluminium shells

sometimes bed down a little and exhibit either tight spots
or hollows in the bore; it is therefore a wise precaution to
hone them out initially 2 or 3 thou. small and re-hone
or lap out to the correct size after a short period of runningin with an old or slightly undersized piston.
Shown overleaf is a simple method of rejuvenating a castiron barrel by boring it out and fitting a short liner, which
can be made from any good grade of cast-iron but prefer
.ably an alloyed iron containing Ii to 3% of nickel. The
59
TUNING FOR SPEED

FACED OfF
LEVEL
AFTER FITTING
WEAR SHOWN
EXAGGERATED
::
f.
END OF LINER AND FACE
OF GROOVE MUST BE FACED
OFF FLUSH TO AVOID ANY
GAP
All traces of abrasive must, of course, be finally eliminated

by several washings in clean spirit.
If this treatment is applied to side-valve barrels a hollow
will sometimes be noticed in the surface adjacent to the
exhaust-valve port; this is a sign that when hot the barrel
distorts inwards, and the" bump " s~ formed has in time
been rubbed away by the piston. When cold, of course
the former" bump" now appears as a .. hollow," and it i~
better not to continue lapping in an attempt to eliminate it.
If that is done, the bump will reappear when the engine
w~rms up and may cause trouble if full-throttle is applied
WIthout a lengthy preliminary running-in period.
Short liners may be lIsed very effectively for

barrel renovations.
important point to watch is squareness of the shoulder and

liner to avoid any semblance of gap; if properly done
lhe joint will be almost invisible and will remain so in
service.
The quickest method of truing up worn or linered bores
is by means of a cylinder hone which can be rotated and
reciprocated in the bore either under a drilling machine
with a long spindle travel or 'by means of a slow-speed
electric drill. A copious and continuous supply of kerosene or paraffin oil is vital to prevent scratching the surface, but there is no need to attain a mirror-like finish.
The ideal surface consists of a large number of very fine
criss-cross markings which retain oil and give a quick bedin of the rings.
Hand-lapping to obtain a perfect finish or to clean up a
slightly damaged bore can be done with a discarded piston,
split in halves and expanded by two springs slipped over the
gudgeon bosses. Mounted on an old rod or a piece of
wood, this device is pushed up and down the barrel by
hand and given a partial rotation at each stroke, the abrasive being flour emery mixed with oil if much metal has to
be removed, and metal polish for obtaining the final finish.
60
Higher power output callsfor greater

heat dissipation; it
is interestillg to
compare the/inning
on this 1939 Velocette T. T. engille
with the 1926 edition all page 48.
','
TUNING FOR SPEED
Enclosure of valves in large cast boxes which eliminate a

lot of fin area is bad from the cooling aspect, and in such
cases the covers should be liberally drilled and fitted with
scoops in order to stimulate a flow of air through the valve
chest; the fins sometimes cast on the covers are not really
effective, as the stagnant pocket of air in the valve chest is
the real barrier to heat dissipation.
For short-distance events with alcohol fuel-grass- and
dirt-tracking and sprint work-deep barrel fins are not
necessary (readers will remember the early dirt-track
J.A.P. power units, which might almost be called" cleanshaven "), but when road racing, complete with mudguards,
barrel cooling becomes worse as the speeds become higher,
due to the "dead air" region behind the front-wheel
assembly becoming more and more pronounced. Hence the
use of fins up to 8 in. in diameter, which would be of
prohibitive weight in iron and are, therefore, made of
aluminium or, in rare instances, magnesium alloy.
For anyone with a little skill in lathe work there is no
great difficulty in making up a bi-metal barrel out of a solid
casting, or, better still, a forged bar of light alloy, with the
fins turned from the solid. The tapered form of fin has
been found to be the most effective, although a little more
difficult to machine than the parallel variety. The bore
should be finished at as high a speed as the lathe will permit,
using for preference a tungsten carbide tool and a feed of
about .003 in. per revolution, as successful heat transference
between liner and barrel depends very largely upon the
perfection of finish of the contacting surfaces.
The actual composition of the alloy is of little moment,
except in so far as the coefficient of thermal expansion
varies. High expansion rates are not advantageous, and in
this respect the magnesium alloys and the LS aluminium
alloy commonly used for general castings are the worst of
the lot at .000026 in. per degree C. High-silicon alloys to
specification L33 are the lowest at .000019 in .. but do not
machine well. "Lo-ex," the alloy from which most motorcar pistons are made, has the same coefficient as L33 but
62
machines quite well, especially if given a low-temperature

heat-treatment of 10 hours at 1700 C.
The best compromise is afforded by Y-alloy or R.R.SO,
with a coefficient of .000022 in., while the aluminiummagnesium group-such as " Birmabright " or Hiduminium
40-are only very slightly higher and have the additional
advantage of being very resistant to corrosion, even by
sea-water.
A process which has lately come into prominence is the
" Alfin" system, in which a true metallic bond is formed
between the iron and aluminium, so eliminating a heatbreak at this locality. The process is handled by Wellworthy
Ltd., Lymington, Hants, who are prepared to cast solid
jackets on to barrels, the fins, of course, having to be subsequently turned by the customer. Racing Norton and
J.A.P. engines have been equipped for some time with
" Alfin" barrels, with fins cast to size. The main patents
covering the process are held by the Fairchild Aviation
Corporation, Farmingdale, N.Y., U.S.A., and there are
several licences operating in that and other countries. A
recent development has been the bonding-in of austenitic
cast-iron (" Ni-resist") seat inserts in cylinder heads,
thereby obtaining better conduction of heat from the
insert to the head than with shrunk-in inserts.
63
CHAPTER Y

or singly, the piston, rings and gudgeon-pin
have a number of jobs to perform. They must prev~nt
leakage of gas from, and the passing of oil into, the WOrkl?g
part of the cylinder, and also act as a cross-head t.o .resist
the side-thrust of the con.-rod and absorb the mmimum
of power in so doing. As each is dependent upon the others
to a greater or less exte~t, they can ~e considered as a
little family group in whIch the behavIOur of one. affects
the well-being of the remainder. .
..
All these components work in highly adverse condItIOns
of high and fluctuating temperatu.res,. high. pr~ssures and
rubbing speeds and scanty lubncatIOn; It IS. n?t surprising, therefore, that nearly half the. power lost m mternal
friction is absorbed by them. Luckily, as no medals are
awarded for mechanical silence in racing, the problem of
eliminating piston-slap does not arise; t~us the ~learances
which give the best results can be used ~rr:spectl~e of any
noise which may occur. Nevertheless, It lS a mIstake to
JOINTLY
RING EXPOSED
TO FLAME
CONTACT AT
EDGE ONLY
A pistoll with excessive

clearance in the barrel
tilts at t.d.c., thus exposing the top ring to
flame contact and
rounding the edges oj
all rings.
64
think that ultra-wide clearances are essential or even

beneficial; the reverse is the case in fact, for piston rings
cannot function properly in a piston with an excessive
amount of "slap." Whenever the piston is tilted, the
rings, instead of being in contact over their full facewidth, will make only edge contact over a large part
of their circumference, and so cannot seal properly; in
time the rings will wear to a barrel shape, and this will
diminish their ability to prevent the upward passage of
oil. Also, as the piston alternates in position from side to
side of the barrel the rings have a tendency to follow suit
(due to the friction between them and the ring grooves),
and even if their contact with the wall is not momentarily
broken, the action is very prone to initiate ring-flutter.
These effects are governed by skirt clearances, i.e., the
diameters measured at right angles to the gudgeon just below
the lowest ring and at the bottom edge of the skirt. Since
there is, or should be, very little oil present above the rings,
the lands between them cannot be expected to carry any
load at all, and should run just clear of the barrel. As there
is a sharp rise in temperature as the crown is approached,
the lands must be progressively smaller, but again it is
undesirable to make the top land too small, since this would
expose the top ring to the fully fury of the combustion,
whereas a close-fitting land will shield it to some extent.
Deliberately exaggerated, this illustration

gives an idea oj how
ring lands CUll be
checked on a surJace
plate Jor clearance and
concentricity.
O~
65
TUNING FOR SPEED
A correctly fitted piston, after running at racing speeds for

some time, should show an even bearing over the thr~st
faces of the skirt and just the faintest witness of rubbmg
contact on the ring lands. To obtain this result it is imperative that the relationship between skirt and lands is correctthat is to say, in addition to being the right relative sizes,
the lands must be concentric with the skirt, which cannot be
determined simply by direct measurement witha micrometer.
It is, of course, possible for a piston which is too small
all over to exhibit the appearance just described, but this
can be detected by the presence of thick carbon deposits
in the plane of the gudgeon pin, or by the skirt showing areas
of heavy bearing at the upper and lower extremities and light
bearing in the centre, which appearance is caused by the
piston running tip-tilted, in turn permitting the ring lands
to make contact. Easing down these areas of heavy bearing
under the erroneous impression that they are .. high
spots"
only make matters worse.
Perhaps at this juncture it might be as well to deal with
this matter of easing down the high spots, which is frequently misunderstood. Obviously the piston must touch
the barrel somewhere, but if every time the motor is
down all the rub marks are carefully eased off with a file,
the piston will simply become smaller and smaller until it is
useless. The thing is to distinguish between the appearance
of normal and abnormally hear), contact areas. The latter are
potential sources of seizure, and the surface may be either
torn and rough (in which event momentarily seizure has
actually occurred) or possess a smooth polished appearance
easily visible by viewing tangentially across the curved
surface.
Areas of normal pressure have a dull-grey matt surface, possibly intersected by machine marks of trifling
depth, and should be left undisturbed, but the polished
areas, or tfue " high-spots," should be eased down with a
dead~smooth single~cut, or Swiss file used very sparingly,
for it is a simple matter to convert a high spot into a hollow,
and, once done, the fault cannot be rectified-it can only
will
66
A piston typical of
modem racing practice
employing two narrow
compression rings and
an oil scraper ring.
be disguised by still further filing. The file must, of course

be used with a circular motion, so that the cut washes ou~
smoothly to zero at the perimeter of the area being worked
on.
As regards the actual clearances, conditions vary so
greatly that it is virtually impossible to give any hard~and~
fast rules and in case of doubt the manufacturer's advice
must be sought. Generally speaking, the temperatures in
r~cing engines, even with 50/50 fuel, are not a great deal
lugher than those sometimes encountered in touring engines,
thanks to the use of open exhaust systems and the fact that
the racer is almost always making a pretty fair breeze of its
own. As a rule, therefore, an increase of .001 in. to .003 in.,
according to diameter, will suffice in most cases.
The thermal expansion of the piston and barrel both
exert an influence, and as materials such as Y-aUoy and
R.~. ~3 B are preferable ~o the high-silicon, low-expansion
vanetles on ~ccount of their greater strength at high temperatures and slIghtly less coefficient of friction, allowance must
be made for their greater thermal expansion if a change~over
i~ made. As an approximate guide, the following table
gLves the clearances per inch of diameter which will be found
67
TUNING FOR SPEED
suitable for Y-alloy or R.R. 53 B pistons running in iron

barrels:
.0065 in.
.0055 in.
.005 in.
.0036 in.
.0027 in.
Top land
Second land
Third land
Top of skirt
Bottom skirt
Bi-metal barrels with shrunk-in liners have a slightly

higher expansion rate than pl~in iron, a~d the cl~arances
in such cases can be about .001m. less per mch of dlameter.
These clearances apply to pistons without T-slots (which
are not commonly employed in racing), and are measured
at right angles to the gudgeon.
.
All pistons distort to some extent under .the )~flue!lce
of heat and load, and pistons of the plam cylIndncal
pattern (sometimes referred to as the "pot" type) are
oval-turned to the extent of .006 in. or .010 in. to give
greater clearances on the sides, although the ring ~ands are
frequently circular; when cold, therefore, the skirt clearances progressively increase from the central plane outwards.
Actually, quite a small area is sufficient to carry the con.rod side-thrust, and this fact led to the development of the
slipper piston, which is both stronger and causes less
friction loss through oil shear than the "pot" variety,
but to some extent the latter can be improved by heavily
relieving the sides so that they run well clear of the walls.
The shaded portion

suggests how the contact faces of a slipper
pistoll call be relieved.
68
\
And here is indicated
how the depth of relief
should merge smoothly
into the contact face
when work 011 the slipper type of piston is
completed.
.003.00S SMALLER
THAN OIAM. AT A
Owing to the heat-flow down the side ribs and their consequent expansion, the thrust faces of slipper pistons depart
from their cold shape quite a lot when hot, and the clearances should be increased perceptibly towards the lateral
edges, particularly just below the top ring, as indicated in,
the diagram on the left.
There are two ways of measuring clearances: (a) by'
measuring the bore with an internal micrometer or a dial.
gauge such as the Mercer, and subtracting the various.
piston diameters as measured by "mike" or vernier; and.
(b) by direct measurement, using various feeler gauges.
The latter is the means which most private owners will have
to adopt, and is quite accurate if carefully performed.
For the wide clearances round the top lands it is best to
use a number of thin feelers rather than one or two thick
ones, as the latter give a false reading through being too
stiff to conform with the barrel curvature.
As mentioned, the ring lands must be concentric with
the skirt, and this is easily proved by placing the piston on
a surface plate and measuring by feelers the gap between
plate and lands, which should, of course, be equal for
corresponding localities on each side of the piston.
69
TUNING FOR SPEED
It is sometimes necessary to reduce an oversize piston

to fit a standard bore, or to finish to size a semi-finished
piston, i.e. one with all machining completed except the
final sizing of the ring-lands and skirt. This job can be
done in a lathe using a large 60 cone centre in the open
end with a loose pin or blade bearing on the gudgeon
bosses to provide the drive, and a regular centre in the
tail-stock. Alternatively a soft spigot can be turned in the
chuck to fit the register bored in the open end of the piston
which is then pulled back on the spigot by a loose pin
and bolt passing through the hollow head-stock. In either
case the ring-grooves must be checked with a dial-gauge
to see that they run free from wobble; with any reputable
make of piston, this is an indication that the piston as a
whole is running true and finish-turning can commence.
Camming, or relieving, the sides as just described, is best
done in one of the piston-finishing machines used by
engine re-conditioners. Of the range of cams developed
and widely adopted for these machines, type "C" is usually
most suitable. It gives .009 in. total ovality and .006 in.
reduction in diameter measured at the 45 points. Failing
access to one of these devices, camming can be done either
by hand or in the lathe by. offsetting the piston .010 in.
in the direction of the gudgeon and relieving the high side
to a depth of .005 in. and then repeating the process on the
other side. This will leave about half the skirt on each
thrust face still circular and the sides of these areas must
finally be merged into the relieved portions by hand
filing.
A precaution which must always be taken, even with
new factory-made pistons, is to see that the corners of the
ring-grooves are slightly chamfered. If not, a fine square
file run round the grooves will do the trick. The purpose
of these chamfers is partly to provide a small channel for
oil distribution, but mainly to prevent the rings being
locked into the grooves by metal being pulled over at the
corners should the lands happen to run in contact with
the cylinder. Locked rings permit local blow-by to occur
70
and destruction of the piston in that area may then occur

with startling rapidity.
Another worthwhile idea is to chamfer the open end at
45 almost to a knife-edge. This has the effect of skimming
excess oil off the bore and directing it up towards the piston
crown, whereas a square shoulder just rolls the oil up
ahead of itself, creating oil-drag and increasing the work
to be done by the oil-control ring.
The skirt clearances of two-stroke pistons are of great
importance because the piston has to act as the pumping
element for the induction system. Pressure of up to 6 lb.
per sq. in. are involved in the crankcase breathing cycle,
and as there is very little oil present and no rings at all on
the skirt, leakage can only be prevented by the latter being
a close fit all round its circumference. Extreme ovality
leads to loss through the transfer ports, round the piston
sides and out through the exhaust. Consequently many
such pistons are ground circular or with not more than
.003 in. camming, and that concentrated in the region of
the gudgeon bosses, leaving the last ! in., or so, of the
skirt parallel and with a clearance of .0015 to .002 in. per
inch of diameter.
The bores of air-cooled split-singles distort quite considerably on account of the difficulty of cooling the metal
between the bores. It pays to make these pistons a fairly
close fit, then, after a warm-up period, to give the motor
a short burst at full-throttle, shutting off immediately at
the first indication of tightening-up. Now dismantle the
engine, ease down the areas of high pressure and repeat.
This process will give much better results than any amount
of running-in at light load.
The most satisfactory ring equipment is two narrow
pressure rings, and one scraper, usually of the slotted type,
which is so effective that a very copious supply of oil can
be fed to the barrel. The pressure rings should have
.0015 in. to .002 in. side clearance in the grooves, but
scrapers perform their work better with about twice that
amount. Oil wiped off the bore is thereby enabled to pass
71
CHECKING .PISTON AND RING WEAR
TUNING FOR SPEED
through to: the back of the ring and out through the drainholes. Sometimes a row of drain holes is provided below
the oil-ring; these should be horizontal and not drilled
downwards at an angie, as the rapid acceleration of the
piston away from t.d.c. then has the effect ~f driving any
oil in, or near, the holes back through the pIston.
On the other hand, the primary job of the upper rings
is to maintain gas-tightness, but this is done not so much
by the natural pressure of the rings against the walls as
by the action of the high-pressure gas passing through the
clearance between ring and groove into the clearance space
behind, and thus forcing the ring outwards against the
cylinder. If the side clearance is insufficient the gas cannot
get through sufficiently fast to build up this vital pressure
in the very short time available. But if it is too great, then
there is likelihood of the grooves being hammered out wider
still as there will not be much oil present to provide a
cushion. To assist in building up pressure, the upper ring
grooves should be only .005 to .010 in. deeper than the
radial thickness of the rings. The small clearance also
helping to suppress ring-flutter.
The pressure-ring grooves should be only .005 in. to .010 in.
deeper than the radial thickness of the rings (the small
HIG<H_F'REl;SUI\E GAS
The effect 0/ COIIIpressed or expanding

gas as a means 0/ increasing the radial
pressure a/piston rings
against the cylinder
walls.
72
___~J
clearance at the back of the ring helping to prevent ringflutter), and in all cases the sides of rings and grooves must
be free from waves or ridges. These defects do not often
occur in the grooves unless the piston has seen a great deal
of service, but are sometimes noticeable on rings; the
remedy is to lap the sides on a flat iron plate with flour
emery and finally metal polish, using only light finger
pressure. Incidentally, it is better to have dead-flat rings
running at slightly excessive clearances than wavy ones
with the correct amount of side-slack.
....
DIAMETIiR-D---
RADIAL
THICKNESS.
D.26 FOR
RACING
SIDES TO BE
DEAD FLAT.
PARALLEL
AND PREFERABLY
LAPPED TO A
MIRROR FINISH
Important ring dimensions

which
WIDTH
must be taken in~o
FREE GAP
account when III
) TO " TIMES RADIAL DEPTH search 0/ the last
CONTACT ~ACE MUST BE FI>.EE FROM
ounce 0/ per/o rm BROWN PATCHES INDICATING BLOW-BY
ance.
The grooves in pistons which have been run for some

time in severe conditions often wear unevenly. When new
rings are fitted in such grooves they may feel as if the side
clearance is correct because, being of correct radial depth,
they are located in the unworn portion of the grooves.
However, under working conditions, the top ring in particular is subjected to a severe twisting action, due to lack
of support at the outer edge, and it will almost certainly
. fail in short order. The only cure is to skim the grooves
out parallel and, if wider rings are not available, to fit two
rings per groove by grinding-down standard rings by the
requisite amount. There is, of course, a limit to how much
can be taken from the grooves because the lands eventually
become too thin and may break under load. Definite
figures are hard to give but, as a general rule, i.\ in. would
be about the safe minimum land width and then only if
the piston is of strong material such as Y-alloy or R.R. 53.
73
. _______.______.__ _
TUNING FOR SPEED
" I
Effective gas ~ealing is achieved mostly by the top rin~,

the second acting as a standby should the top one fall,
and also as an oil controller. For sprint work, or in very
small engines when there is little room for rings, it is possible to omit the second ring and use a single groove, with
two narrow side-by-side rings therein as just described.
By placing the gaps on opposite sides minute leakages
through this point will be almost entirely eliminated.
A recent development is the Dykes ring, which is of
"L" section, the short leg providing for location in the
piston, while a comparatively large clearance at the top of
the ring allows a clear path by which high-pressure gas can
flow rapidly into the space behind the vertical portion, and
thus expand it against the bore. Only one such ring per
piston was employed in the 1954 A.J.S. r.acing twin, though
the usual slotted ring was also fitted for 011 control purposes.
There is little to choose between the various well-known
brands of ring, but each maker has perfected types specially
suitable for racing, notably the Wellworthy "Thermochrom" and "Limalloy" or Hepworth and Grandage
"Phormicast." Whatever the make, it is advisable to obtain
rings with lapped sides, since they maintain correct side
clearances for long periods, and there is much less friction
between the ring and groove when subjected to heavy gas
pressure.
The usual radial depth of English or Continental piston
A piston rillg reclaim method which n~ay be

used ill cases where spares are llot avO/lable.
74
rings is iij-D is the barrel diameter-but lately high-pressure rings of greater radial depth, ~, have come into use.
The latter type are less prone to flutter at high speeds, but if
they are used care must be taken to see that the ring-grooves
are deep enough to accommodate them. If they are fitted in
grooves of normal depth there is a likelihood that they will
stand proud of the ring lands even when forced down to the
bottom of the grooves, and, in service, the ring-faces will
then be forced to carry the side thrust which should be
borne by the piston skirt. This condition, although easy to
overlook. is very bad indeed, and must be rectified by
deepening the grooves. Another point about these rings
is that, being stiffer, they will not conform to a worn, or
oval, barrel quite as freely as will the lighter variety, and it
is advisable to lap them in very carefully to ensure gastightness if the barrel is not so perfect as it might be.
In side-valve engines and most two-strokes some barrel
distortion is bound to occur due to unequal distribution of
metal and varying temperatures around the cylinder and in
these motors it is a good plan to peg the rings so that they
cannot rotate in the grooves. They then bed-in closely to
the shape which the cylinder attains when running and will
maintain good gas-tightness and oil control even when the
cylinder has worn appreciably.
Two-stroke rings are also pegged for another reason-to
prevent the ends springing into the ports and becoming
broken, and they are also generally much wider in proportion
to their depth than four-stroke rings of equivalent diameter
to enable them to traverse the ports with less shock. When
fitting new rings care must be taken to clean out the corners
adjacent to the pegs very thoroughly, otherwise the rings
will be propped up above the lands and either they or the
barrel will suffer as soon as the engine is started up. The
tiny lips on the ends of the Puch rings are very frail and can
easily be broken off during assembly if held proud of the
piston by the pegs or by carbon.
Wide ring gaps are advisable, particularly for engines
SUbjected to rapid cold starts; .010 in. per inch of diameter
75
TUNING FOR SPEED
is a good figure to work to. To make a thorough job,

after rough gapping, lap the rings in to the barrel with
metal polish until they are bedding uniformly all the way
round, and finally finish the gaps to the correct width.
Regarding gudgeon pins, it is a great aid to assembly
to make these an easy hand-push fit in the piston when cold.
Some makers prefer to fit them a little tighter, but this
entails driving the pin out with a drift, or warming up the
piston with hot water or a bit of meta-fuel in a cocoa-tin
lid; as in any case a i-in. pin will have about .002-in.
clearance at running temperature, another .0005-in. will
make very little difference. The best way to take a thou. or
so out 'of gudgeon pin holes is with a hone of the Sunnen or
Delapena type, using kerosene or paraffin liberally. It is
not easy to take out small amounts with an ordinary reamer
without leaving small chatter-marks in the bore: indeed,
sometimes a better job can be made with a triangular hand
scraper.
The main attributes of a gudgeon pin are stiffness and
high surface hardness; pins which have become .. blued"
If'.. ENDS OF CIRCllP SHOULO BE

r'! iN APART WHeN CORRECTLY
___
~O':"'il.
t:
I
\ :,
\
FITTED
.,:
)
I
-..... .............. _......._----.

-_._
~:
~-
IJ a special pair oj
pliers is 1I0t available
the poillfS oj an old
pair oj scissors call be
ground dOWII alld used
to remove eire/ips oj
the Seeger type.
76
CHECKING PISTON AND R1NG WEAR

ENDS CHAMfERED 45"
Chamfered outer
gudgeoll pin edges act
as eirciip retainers,
whilst the tapered bore,
left soft, reduces
weight alld is polished
so that flaws may be
detected..
are softened and should be replaced. Insufficient pin

rigidity is a fruitful source of cracked piston bosses, and the
most satisfactory designs have a central bore not greater
than .6 of their diameter. Although this can be tapered out
for one-third the length of the pin at both ends, the bore
must be left soft and is preferably polished to remove surface blemishes from which cracks may start. A nickelchrome steel sllch as K.E. 287, case-hardened, .025 in.
-.035 in. deep, is an excellent material for the purpose.
The best form of end location is by circiips. either of the
spririg-wire type (without bent-in ears, which sometimes
break off) or the Seeger pattern, although for fitting the
latter variety you must have the right sort of pliers, available quite cheaply from the makers of the circlips. The
wire type is perfectly satisfactory provided the ends of the
gudgeon pin are chamfered, and the overall pin length is
such that the final circlip can just be sprung into place
with the pin thrust hard against the opposite one. The
chamfers then act as retainers, and there is DO likelihood of the circlips ever managing to work out of the
grooves. Seeger circlips require flat~bottomed grooves
turned to accurate depths in accordance with the table
which covers the range of sizes most likely to be met. This
type of circlip must never be fitted in the half-round grooves
used with wire circlips because, although it may seem to fit
correctly, in fact it would only be seating on the corners
77
TUNING FOR SPEED
and in a short period of running the corners would dig in

and the circ1ip would become loose.
Wire circlips should always be stretched out before refitting to make sure that they will fit tightly in place.
Occasionally an engine will exhibit a tendency to hammer
the circlips loose, more so on one side than the other;
in such a case examine the alignment of the small end bush
for both squareness and absence of" twist"; the latter is
difficult to detect but may be the unsuspected cause of the
trouble.
If a spare is not available at all, or the standard article is
known to be unreliable, a new piston can be made from a
solid bar with only a lathe, a drill, and the usual hand tools.
The process is rather too long for inclusion here but as a
guide it will help to note that the best material is aluminium
alloy R.R. 59, which is specially made for pistons and
retains strength very well at high temperatures. The
American material to SAE260 specification, and known
commercially as 14 S also is very suitable. Alloys such as
R.R. 77, which are even stronger at atmospheric temperatures, contain zinc and, therefore, are not so good at
higher temperatures.
CHAPTER VI
THE function of the flywheel assembly is to convert the

straight-line motion of the piston into rotary motion at
the engine sprocket. The crankcase also forms a vital
link in this mechanical chain by providing a rigid location
for the cylinder and mainshaft bearings, whilst, in addition,
housing timing gears, oil pumps and so forth. Referring
back to Chapter 1. it was pointed out that any areas where
undue wear has occurred are open to suspicion and that
the cause of the wear (which may be malalignment, lack of
lubrication or faulty material) must be investigated with an
eye to its cure. This advice should come in useful at this
stage, for the main job of work to be tackled in the" basement " will consist of the elimination of friction usually
due to malalignment.
The Velocette slow-taper crankpin:!fywheel assembly calls

for the use of a specially modified" Pickavant" hydraulic
tool when dismantling (left) or reassembling (right).
78
79
TUNING FOR SPEED

.020" CHAMFER
0"0"
'020 RADIUS ON
ICtJjI~:'
A worn big-end assembly can be renovated by

grinding down and fitting a press-fit ~a~dened
sleeve to the crank-pin. Where the pili lS not
drilled centrally for lubrication, the sleeve must be
provided with annular grooves and a transverse
passage.
Talking of big-ends, in that chapter was mentioned the

desirability of having the oil fed to the centre of the rollertrack and not simply to one side, a scheme which is now
practically extinct. Unfortunately, big-end components are
in rather short supply from time to time and final assembly
of the flywheels may have to be postponed if the existing
parts are too badly worn to be worth retaining ..
The type of big-end in which the central portIOn of the
crankpin is considerably larger than the two ends can be
repaired very satisfactorily by grinding down the centre
portion and fitting thereto a hardened sleeve. If the outer
race in the con.-rod is also worn, this should first be ground
or lapped out true-preferably without removing it from the
rod-after which operation the pin-sleeves can be ground to
the particular oversize necessary to suit th~ finished outer
race diameter and whatever rollers are avallable.
The pin-sleeve should preferably be madeJrom ball-race
END CLEARANCE TO AVOID
ROLLERS BEING NIPPED IF
FLANGES ARE DISTORTED
With the shouldered

type of crankpill there
mllst be end-float between rollers and
sholliders to avoid
dallger of the former
becoming nipped when
the pin is tightened ill
the flywheels.
80
steel (1% carbon, 1% chromium), which hardens right

through by quenching in oil after soaking at 8200 C. for
half an hour. After quenching, temper for half an hour at
200 C. and cool in air; this procedure relieves internal
stresses without materially affecting the hardness, which
should be Rockwell C. 60-64 or diamond hardness number
750-800.
If the crankpin is already drilled centrally for oil, a mating
hole must also be drilled in the sleeve before hardening.
Where the pin is not drilled, an annular groove should be
formed at one or both ends of the sleeve to collect lubricant
from the flywheel oil-hole, a shallow transverse groove with
central hole finally delivering it to the rollers as shown
in an accompanying sketch. After hardening, the sleeve
must be ground internally to a diameter .00 I in. smaller
than the pin, then forced into position under a press. Finally,
it must be ground externally and on the two end-faces;
squareness of the latter is particularly important, as the
slightest error will be greatly magnified at the mainshafts.
If carbon-chrome steel is not available, a case-hardening
variety can be used, in which event the bore can be left
soft if desired, thus avoiding the necessity for internal
grinding. The scheme is first to rough-turn the blank to
grinding sizes and then to carburize for six hours to obtain
a case-depth at least .045 in. After cooling, rough-bore
then recess the end faces to t in. larger diameter than the
bore to a depth of /6 in., drill the oil hole and finally heattreat. After hardening, the sleeve can be finished internally either by boring or grinding, fitted to the pin and
then finish-ground exactly as previously described.
Mild steel is not up to the arduous nature of this work, the
case being liable to flake away from the soft core; 3%
nickel to specification 815 or EN 33, or single-quench 3%
or 3!% nickel will be satisfactory for moderately heavy
duty, but better varieties are 882, 890, EN 34, 38 and 39,
the last three being wartim.e specifications. The heattreatment for all these grades is practically the same, viz.:
carburize at 900-920 C. Refine by quenching in oil from
81
0
TUNING FOR SPEED

0
850-860 C. Harden by quenching in oil from 760 C.

Temper by cooling in air from 150-170 C., except for the
3% and 3i% nickel single-quench grades, in which the refining and hardening can be effected simultaneously by
quenching in oil or water from 760-780 C.
Equivalent American specifications are S.A.E. 51100 for
carbon-chrome ball-race steel, S.A.E. 2317 for 3t% nickel
case-hardening steel and S.A.E. 3316 for 3!% nickel-chrome
C.H. steel. If it is necessary to make up a completely new
pin, either of the two case-hardening steels mentioned will
do. The plain nickel steel, though having less core strength,
gives a slightly harder case than the nickel-chrome steel.
Steels which harden right through, such as the ball-race
steel mentioned or medium-carbon direct-hardening steels
lUlIst not on any account be used for complete pins, because,
when hardened sufficiently on the roller track, such steels
are far too brittle in the core and are certain to break in
service. In any case the threaded ends must always be left
soft and this can be achieved by copper-plating prior to
carburizing and hardening. Another method is to roughtum the threaded portions .100 in. oversize, carburize the
whole pin, turn off the carburized skin locally and screw-cut
the threads and finally harden. Incidentally, great care must
be taken to ensure that the threads are square to the axis,
otherwise it may be difficult to maintain tightness of the
nuts.
on no account may they be convex. even to the slightest

extent. Unfortunately, with normal grinding equipment,
it is difficult to attain this finish, but somewhat the same
effect can be gained by turning a shallow undercut in the
faces which also provides a little extra room for the vital
fillet.
Sometimes the oil-hole in the pin is drilled inwards from
the corner of the shoulder, causing a local weakness from
which a fatigue crack will start. It is best to modify the
drilling to avoid this danger; it may also be necessary to
drill another hole in the flywheel to coincide with the new
position of the hole in the pin,
De/ails of
SHOULDER
crallkpill de-
sign which are

likely to cause
weaklless
under highdilly condi-
FAULTS TO BE
LOOKED FOR
IN CRANKPIN$
tions.
Breakage of crank-pins of the shouldered type almost

invariably occurs in the corner of the shoulder, and it is
desirable to have a definite fillet in this corner, even at the
expense of reducing the length of bearing in the flywheels.
The faces of the shoulders should be slightly concave;
When reassembling the big-end new rollers should be

fitted for preference, but old ones can be used, provided
they are absolutely free from flats or other surface blemishes (which can be detected by examining them through a
magnifying glass under a fairly strong, oblique light) and
are all of the same diameter to within one or two tenths of
a thou. One or two oversize rollers will rapidly ruin any
roller bearing, this being the reason why new and old
rollers should never be run together.
Caged rollers are maintained in parallel by the spacing
bars, and, if the pin is of the shouldered type. the rollers
can have several thou. end clearance-in fact, this is advisable, otherwise if any deflection takes place the shoulders
may close in a trifle and nip the rollers endways. If the
pin is without shoulders, the bars of the cage should be
82
83
.,...,..,...,..,.....~GROUNa SHOULDER
This method oj mach il/....
HA
lIIg a groove alld proLF I!.OUNO CROOVE vidi~lgagrol(1/d slwlllcjer
obViates
pO:iSlbJe
weakness wilhOllt re-
the ejJective
length of spigot.
ducillg
TUNING FOR SPEED
BRONZE
PLUG
RADIUS OR FILLET
IN CORNERS ~
Details of shoulder
convexity shown ill
exaggerated profile for clarity.
blL HOLE DRILLED HALF WAY

ALONG SPIGOT DIAMETER
-I~CONVEX SHOULDERS (GREATLY EXAGGERATED)

ABOUT .001" CONVEXITY IS SUFFICIENT
relieved on the internal diameTer so that the cage bears on

the pin only outside the roller path, for although the cage
is soft it will wear away the pin surface much faster than
will the rollers. In engines running at 7,000 r.p.m. or more,
this form of wear can become very serious, and it may be
advisable to hard-chrome-plate the rubbing areas; .001
to .002 in. thickness will be ample, but the deposit must be
finally ground to a dead-smooth finish and must on no
account be allowed to encroach on the roller tracks. Useful
as it is in other ways, hard-chrome cannot stand up to the
action of heavily loaded rollers and will inevitably flake off
if applied to the tracks.
What makes the life of a big-end so strenuous is not so
much the speed or the loading but the fact that the roller
assembly does not rotate round the pin at a steady speed,
as it does in a main bearing. Instead, it has alternately to
speed up and slow down because at or near top dead centre
the swing of the con.-rod is in opposition to the motion of
the pin, whereas at bottom dead centre the two motions are
in the same direction though of course not at the same
speed. Like all objects possessing weight, the roller assembly
objects to this process and relatively heavy loads are created
between the rollers and the bars of the cage and a certain
amount of sliding motion must take place.
The lighter the assembly, the less its inertia, and for that
reason racing cages are made of one of the strong wrought
aluminium alloys such as R.R. 56 or 24 S. Some cages are
made from cast aluminium which, though light, is relatively
84
weak and for serious work should be replaced by parts

made of bar or tube. Factory-made cages have the slots
formed by broaching, a method which is not available to
everyone; an alternative process is to end-mill the slots
and then machine the ends out square. A cage with relatively long slots to accommodate long rollers such as in.
by r\ in. (a very useful size) or three short rollers placed
in each slot can easily be made by milling the slots with a
standard Woodruff key-seat cutter, and finishing the ends
out by hand using a square file ground to width so that it
does not damage the machined sides. The blank should be
turned with a shoulder or groove at each end to mark the
length of the slots, which is not vitally important. What is
important, however, is that the slots must be dead parallel
to the axis, otherwise the whole roller assembly will run
over to one side and, therefore, accurate setting-up of the
milling machine and dividing head is essential.
Large uncaged rollers are not good for high-speed work
because of the heavy pressure between them due to centrifugal force. This force can be reduced to an acceptable
figure by using rollers which are small in diameter and.
consequently light. The Vincent big-end uses three rows
of rollers 3 mm. diameter by 5 mm. long per rod, separated
SLOTS FILED TO
LENGTH Of ROlLERS
--=t
~
~
.tZ2Z
I
~
~
......:...--0
1"'- /1
EQUAL TO LENGTH
OF ROLLER
MILLING FIRST SLOT

CUnEI'. 15 JUST
BREAKING THROUGH
WITNESS OF cunEI'.
AT ENDS OF SLOT
IS IMMATERIAL
TURNED BLANK
Three stages in the fabrication of a big-end roller bearing cage,

using bar or tube material.
85
TUNING FOR SPEED
by parallel rings which prevent the rollers skewing; ~e

total end clearance must be held to limits ?etween .00.4 m.
and .010 in. if skewing or binding endways IS to be aYOl ded .
The 250 Guzzi also uses small rollers of 3 mm. dl~met~r
by approximately i .in. l?n~ with ;to cage, but m t~lS
instance the big-end IS split, Just as If the rod had a plam
bearing. Plain big-ends have been successfully used for
some time now on high-speed engines though they u\e
necessarily smaller and more, heavily load~~ ~han theIr
counterparts in the automobl!e world. .Rlgldlty of the
metal surrounding the big-end IS of more Importance than
1ightness, therefore no attempts at lightening ?f altering
the design of plain rods should ~e attemp~ed WIthout consultation with the makers, who m some mstances supply
rods of different design for racing. The correct clearance for
plain big-ends is .001 in. to .0015 in. per i~ch wh~n :he rods
are cold though a much greater clearance IS permissIble provided the oil pressure is maintained above the allowable
minimum figure as quoted by the makers.
In any design of roHer bea:ing there i.s bound to b,? a
certain amount of rubbing gOIng on, WhiCh,. at ultra-hl~
speeds wiB generate sufficient heat to cause seizure unless It
is carried away by the lubricant. For reasons of strength,
the oil-h(lles are never very large, and care must be tak.en
to check that they are clear by feeding oil thr?ugh then;, w~th
a force*feed oil can. Incidentally, when domg up a vmtage" model, it is often possible to obtain a more up-:odate big-end which will fit the old pattern wheels With
1ittle or no modification.
. .
After assembling the rollers with oil (not grease,. WhlC~ IS
likely to block the oil-holes) the con.-rod should Just sh~e
into place. If it has to be pushed o~ scr~wed on, the fit IS
too tight: if there is more than .002-m. dlametr~l clearance
it is too loose. The best way to get an exact fit IS b~ means
of several sets of rollers graduated in steps of .0001 m. But
if these are unobtainable the outer race can be lapped out
with flour emery and a mild-steel lap, but on no acc01;lnt
must any attempt be made to ease a tight big-end by rotatmg
86
it with abrasive applied to the rollers; such a procedure will

almost certainly lead to rapid breakdown in service. A
piece of hard wood turned to size makes quite an effective
lap for emergency use, but when many rods of the same
diameter have to be handled, it pays to make up an expanding lap. This tool can either be held in a vice and the rod
turned by hand or it can be used in a drill; in either case the
rod should be allowed to "float" on the lap and be frequently removed and reversed to avoid tapering the bore.
When spun round with the pin held horizontally, the
rod should not show the slightest sign of working over to
-one side; if it does, either the pin -or the outer race is tapered
-or the slots in the cage are at a minute angle. Reversing
the rod may effect a cure, in which event mark the correct
position of pin, rod and cage; if not, the cause should be
located and rectified. Sometimes deep grooves are formed
in the flywheels through rods running over, and if these
are present they should be skimmed out to accommodate
phosphor-bronze or hard-steel washers of suitable thickness.
The correct amount of side-clearance varies according to
the design, but is rarely of great importance, .010 in. to
.020 in. being about right in most cases.
If not already done, it is a good idea to polish the con.rod all over, since the removal of the rough outer skin
greatly diminishes the chance of fatigue cracks developing.
Although rare, cracks of this nature usually occur at the
junction of the shank and small-end. They can be detected
CAST IRON SLEEVE
I-----ABOUT 6 - - -.....
Details of a lapping tool of the split-adjustable type suitable for

the cleaning up of the big-elld eye.
87

TUNING FOR SPEED
by immersing the rod in hot paraffin for some time, and,

after drying, dusting the surface with french chalk; if any
cracks are present they will be indicated by the chalk adhering to the paraffin retained in them. Cracks may also be
detected by the Magnaflux, or similar methods, which are
commonly employed in aircraft manufacturing or maintenance establishments and it is a good idea to avail oneself of such facilities if they are easily accessible.
Shot-peening the surface is of great value in preventing
fatigue-cracks-the system is to bombard the surface with
steel shot, projected at high velocity by compressed air
through a nozzle. This process must not be confused with
shot-blasting, which uses sharp-edged grains and removes a
certain amount of metal. The smooth shot used in shotpeening removes no metal at all but, through its compacting
action, places the surface skin of the metal in compression.
Obviously, shot-peening must be done after polishing, as
any work done on the surface after peening will remove the
One method of checking for bend in the rod.

In place of the adjustable distallce piece, a
plain bar call be filed
to a push {it between
the mandrels. Any
discrepallcy 011 the
other side can be
checked witlt feelers.
compressed layer and destroy the effectiveness of the

process.
Experiments in lightening rods extensively are not advisable, unless the maker's advice has been obtained. There is
sometimes a fair amount of excess metal round the smallend due to taper in the forging dies and manufacturing
tolerances and much of this can safely be removed by filing.
As the stress concentration falls mainly on the extremities
of the ~mall-end; ~hese .should be left untouched. It may
be possIble to obtam a lIghter rod made specially for racing
from higher-grade steel; as one example, the early KIT
::elocet~~ rod can be used in the push-rod MOV engine.
M.anx Norton rods can also be used in "International"
engmes.
A guide to the
amount of material
which may be removed from the
small-end eye for
Iighteningpurposes.
Rod.s must be tested carefully for alignment in all planes.

Checking off the sides of the big-end, the small-end should
be central within .005 in. to .008 in. The big-end and smallend bores must be ~ead parallel; although it is possible to
c~rry out an appro~Imate check for this by using the crank
pm and. gudgeon pm as mandrels, a better way is to make
up a paIr of bars, at least 6 in. long, which are a tight push
fit in the respective bores. For preference, they should be
hardened and grou~d, bu.t soft ones will suffice if carefully
handled. Co~merc!al bnght-~rawn bflr is usually .003 in.
~ess than nommal SIze and qUIte effective mandrels for use
m small-ends can be made from such material hardchrome plated up to finished size.
'
With these mandrels in place, the distance between them
D
88
89
TUNING FOR SPEED
A "works" method of
testing for twist in the
rod. Any error is shown
as a clearance between
one mandrel and the
knife-edge jig.
measured at the extremities should not differ by more than

.002 in. In the absence of the necessary measuring equipment, a steel bar can be filed so that it will just fit between
the mandrels at the tightest end; the gap at the larger end
can then be measured by feeler gauges. The bearings must
also be free from" wind" or twist, i.e., the mandrels should
be parallel when viewed along the length. of th.e rods.
Accuracy in this respect can be checked qUlte eaSIly on a
surface-plate by the use of blocks and a dial gauge or feeler
gauges.
.
It is quite likely that at least one of these errors wIll be
found, possibly all three. To avoid pulling the rod about
unnecessarily, the three checks should first be made and the
situation carefully weighed up, otherwise one may cure one
error by making another worse. As a general ru~e, first
eliminate any .. twist" and then set about gettI~g .t~e
centrality and parallelism of the small-end correct. .oJ It IS
central but inclined, bend the rod close to the small-end;
if off-centre and inclined in the same direction, bend near
the big-end. The rod should never be gripped by the bigend sleeve when being "set"; this action might just cant
90
the sleeve in the bore and in service it would revert to its

original position. Setting should preferably be done in a
straightening press, but failing that a strong vice and three
bolts can be used, two being placed on one side of the rod
and one between them at the point where the bend is
required; the bolts should be of large diameter and soft to
avoid local d~lmage to the rod. Alternatively, narrow
strips of thick brass can be used with much less possibility
of damage.
If these errors are in opposite directions, the rod will
have to be set at both ends, but do not use the checking
mandrels for the purpose and bend these as well. Since the
side clearance between small-end and piston bosses can
be fairly generous, it is permissible to face off one side of the
small-end if all the errors cannot be corrected otherwise.
Considerable patience may be needed, but it is vitally
important that the con.-rod be absolutely true; even the
slightest out-of-truth factor can have a big effect at, say,
6,000 r.p.m.
Rods which have been set only slightly often show a
tendency to revert to their original position, so the best
plan is to set the rod too far and then correct the amount
of overset to bring the alignment right. Rods treated in this
manner are very unlikely to lose alignment again.
BLOCK
SURF,6,CE PLATE
A third method of checking for bend in the rod.
91
TUNING FOR SPEED
The small-end bush for racing must not have less than
.001 in. clearance, but more can be allowed without detriment except for the noise caused. Nortons can be permitted
, up to .005 in. before replacing the bush. Phosphor bronze
is the usual material but any of the light alloys, such as
R.R. 56 or R.R. 59 or Duralumin H, which retain strength
at high temperatures, make very good bearings and save
perhaps an ounce of reciprocating weight.
It occasionally becomes advisable to make up a completely new connecting-rod, and at the same time to redesign
for greater rigidity. In this connection, the best design is
the simplest; the flanges of the I-section should be tangential to the small-end boss, and taper outwards to a width
equal to the bore of the big-end eye. Circumferential
rigidity of the big-end is most important, and to obtain this
without excessive weight there should be either one or two
deep ribs running right around the eye. For material,
" Vibrac "V 30 is extremely good, and for best results the
rod should be rough machined in the soft state, then heat
treated to 85 tons tensile. All polishing of the exterior
should then be completed and the big- and small-ends finally
ground to size. Rods which are not likely to be run at
ultra-high revs, do not require a steel of this nature, and
carbon-manganese steel to Spec EN 16, heat-treated to 65
tons tensile, gives a very good result; it is a standard steel
used in many English machines built after 1945. It can be
machined in the fully heat-treated state and, therefore, does
not require to be ground in the bores after hardening, which
makes the manufacture somewhat simpler.
Where extreme lightness is sought, the weight of the
big-end sleeve can be saved by making the rod from 5%
nickel or 3!% nickel-chrome case-hardening steel, casehardened to a minimum of .045 in. in the big-end eye only,
the rest of the rod being kept soft by copper plating before
carburizing. Many two-stroke rods are made on this
~ principle and it was standard practice in Rudges for many
years.
Although it seems a rather revolutionary suggestion, the
92

5th welD
=-==
,---.:..~_ _ _ _
2nd WELD
=-----_0
c::=
6th WELD
The autho,r's idea of a cOllnec/ing rod buill up of hightenSile sheet steel (lnd fabricated by welding.
writ~r sees no reason why it should not be possible to

fabncate a rod from high-tensile alloy steel sheet united
by. welding. If the small and large ends are composed of
stnps cut along the gra~n of the steel and bent to shape, and
t~e flanges of the I-sectIon are carried round as shown in the
~Iagram, the grain flow in every part of the rod would be
Ideal, which is not always the case with one-off hand
forgings. The joints could be made with filler rod of the
same analysis as the components and the whole assembly
heat-treated before finally machining the rod to size.
93
CHAPTER VII

maximum of rigidity in the flywheel and crankpin
assembly is essential to avoid power-wastage thr?~gh
internal vibration. This is obtained partly by provldmg
heavy sections of metal in the regions of maximum, stress,
and partly by the manner in which the assembly 1S held
THE
l':' I
together.
'
Some years ago it was common practice, to gnnd laIr y
steep tapers on the pin, and to pull these mto e wheels
by fine-thread nuts. This method of constructlOn makes
a good job, provided that th~ tapers ~t really. accuratel~;
thus, in doing-up such an engme (particularly If a new.pm
is being used) the fit should be ~hec~ed by m~ans of prussla~
blue smeared on the pin, whlch IS then hghtly ro:ated m
the holes. If contact is not made over the whole mter~al
surface of each hole the parts can be lapped together, usmg
a fine-grit abrasive, but great care must be taken to see that
the holes are not lapped out-of-square, in w~ich event t~e
second state will be worse than the first. ThIs process will
cause the wheels to come a little closer together and the
side-float of the con.-rod will be reduced by a like amount,
thus the clearance must be checked carefully on final
assembly. The tapered crank-pins used in J.A.P. Speed~ay
engines are supplied in tbree lengths, stanc,lard, plus ~\ 1~.,
and plus fa in., to allow for any e~largement that may, III
time, occur in the holes. By selectmg the c?rre~t leng~h ?f
pin, the con.-rod side-c1e:arance can .be mamtamed wlthm
.
the correct limits of .015 m. to .030 m.
It is now more usual to pull the wheels up agamst substantial shoulders on the pin, rigidity being gained ffi:ore by
this action than by the actual fit of the reduced portIOns of
:n
94
~e pin ~n the w?eels. If parallel, these are usually made a

lIght. dnve fit (I.e., about .001 in. larger than the hole),
and if for any reason the fit is much looser than this the
pin can be hard-nickel-plated and ground to correct 'size.
Velocette crankpins are ground with an almost imperceptible
taper of .001 in. per inch length, and should push in by
hand for half the depth of the flywheel hole; the nuts will
then do the rest, but if the pin will not go in so far as that,
the shoulders may not be pulled up into hard contact with
the flywheel as intended by the makers and not only will
the assembly lack rigidity but there will be a grave danger
that the crankpin will break in service.
Whatever the form and fit of the pin, the whole assembly
event~ally depends upon the nuts, which should be
examtned for possible damage or distortion of the threads
and ~crapped jf there is any doubt about them. The
abuttmg faces must be square to the threads and can be
c~ecked and, if necessary, rectified by turning up a mandrel
WIth a thre.ad tightly fitting the nuts, which are then screwed
on and skImmed up dead flat. If it is necessary to make
new nuts because sI;ares are unobtainable, mild steel is not
good enough: nothmg less than 45-ton tensile steel should
be used, .and alloy steel of 55-65 tons such as K.E. 805 is
better st~l1 for racing. Most English crankpin nuts are
tapped either 20 or 26 threads per inch, Whitworth form
the c~m:nonest sizes .being ! in., t in. and 1 in. and, there.:
fore, . It IS often pOSSIble to utilize nuts of another make if
genume spares are not available.
.Recently there has been a growing te~dency to do away
With nuts altogether and rely simply upon a press fit. This
system has been used for a long time on two-stroke Villiers
and Velocettes, and also for built-up two-throw crank~
sh~fts such as the D.K.W. To be successful, the surface
fi?ISh and accuracy of both shaft and hole must be of a very
hIgh order and great care must be taken during assembly to
avoid broaching metal from the holes.
An excellent method of eliminating nuts which is worth
consideration by experimenters is the S.K.F. Rystem of using
95

TUNING FOR SPEED
CRANKPIN
LIGHT DRIVCFIT
IN WHEELS
.0" TO .OQ.(W
~~~~~~
INTERFERENCEJ2~i!=====B;W~
'EXPANDER PLUGS. ~~~~~~t-:~

.002" TO ,003" TIGHT ~
Two types of ex, .. " pallderplug jiltillg

I..-r,...,..~-' 10 supersede a COIl-
l'ellfiollal crallkpill.
.:-::"
"',
D~ THIS WAY
I--~~;:;'~"",'~~~~'~.ONEPIECE WAISTED PLUG
THIS END SLIGHTLY

~~~"""'"~':iW'7;1.LARGE~ IN .D!At1ETER
~ . THAN O:rH~R END
TO FACILITATE ASSEMBLY
expander plugs in a hollow pin c:r shaft. The schem~ i~ .to

make the shaft a tight push fit 1U the hole so that InItial
assembly is easy with no danger of scoring. ~he~ a hard
steel expander plug is forced into the bore, WhICh IS about
one third to one half the diameter of the shaft, and the latter
is thereby expanded and locked firmly in the flywheels.
Diagrams show two methods by which this sc~eme could
be applied. A somewhat similar idea was used 10 the 1954
A.J .S. racing twin though, in this instance, the expander
plugs had a fairly steep taper.
....
Regarding the flywheels, there is .very .httle P~l~t 10 gO.1Og
in for a big course of weight reductlOn; In fact, It IS possIble
to reduce rather than increase the performance by such
a proceeding, except perhaps for speedway work. Some
96
riders find they get better results with light flywheels on

short tracks where acceleration out of the corners is of
paramount importance, but for road work, where full use
can be made of the gearbox and cl utch, there is not much in
it. However, a general cleaning-up of all surfaces will
not be a waste of time and is of decided benefit on the
rim surfaces, as oil-drag is perceptibly reduced by a high
polish.
Another method of rc.ducing oil-drag is to chamfer the
outside diameter of the rims, leaving only a narrow land
next the inner faces running in close proximity to the edge of
the crankcase scraper. As the majority of the oil to be
skimmed off travels down the inner faces of the flywheels,
the scraping action is not unduly impaired but remember
that, during assembly, the wheels cannot be roughly aligned
by a straight-edge across the rims if this modification is
performed.
The correct fit of the main-bearing inner races on the
shafts varies according to type; in no circumstances should
there be any actual slack present, but if the races are locked
up endways in some manner, they can be quite an easy fit.
If not so retained, there must be sufficient interference
between bore and shaft to prevent" creep" which, once it
commences, will cause the shafts to wear, particularly if
SPLIT MUFf ENGAGING WITH ROLLERS
YWHEEL ASSEMBLY
LOOSE SLEEVE
Split puller for removing inner mab;bearing races.
97
. ,,
TUNING FOR SPEED
they are not case-hardened; a light drive fit is the ideal at

which to aim.
One snag about these tight-fitting races is the difficulty
of removal without damage; applying two levers to the
roller-cage will almost inevitably distort it, thus nipping one
or two rollers endwise, and this defect should always be
looked for in old engines. The best scheme is to make up
. a split puller which cannot harm the cage, but, failing that,
two sharp-edged chisels ground to a narrow taper may be
driven in from opposite sides between the race and flywheel
until there is a gap wide enough to insert a pair of levers.
Of course, before. reassembling, any burrs raised on the
faces must be filed off flush.
As on the crankpins, Velocette mainshafts have an almost
imperceptible taper of .001 in. per inch on the bearing seats
also' the inner bearing races are ground to the same taper
and 'to a diameter such that they have to be lightly driven
on for the last quarter-inch, or so. If looseness develops,
the inner race may appear to fit correctly if it is reversed on
the shaft but this must not be done on any account, as
the race is then only in contact at one end and will rock
about under load.
If slackness is found, do not adopt the barbarous process
of centre-punching or chiselling the shaft, which" however
good it may seem at the time, ceases to be effe,ctlVe afte~ a
very short while. Building up the shaft by, mckel-plat~ng
or metal-spraying with zinc are both effective and lastmg
repair methods and are necessary for these and similar shafts
which are forced into the flywheels and permanently locked
by grub-screws. Though the course is not advisable, Vincent
shafts can be pressed out after driving out the Mills' pin
fitted through at an angle. The correct interference fit is
.003 in. and this must be verified before pressing~in the new
shafts. On the drive side, press in the new shaft hard up to .
the shoulder with the angle-hole about 60 away from the
original position, then drill back through the shaft at 45
with a ~ in. drill, and fit a new Mills pin.
Where the main bearing rollers run direct on the shafts, as
98
in J.A.P, engines, renewal is the only course open if wear

shown by pitting of,the roller tracks is present, As previously
noted, chrome-platmg cannot be used as a reclaiming process
~or roller races. Take particular care to line up the oil-holes
m shaft, and wheel accurately on the timing side, and, as a
precautIOn, check that the axial distances of the hole from the
face of the wheel and from the shoulder of the shaft are the
~ame, o~herwise the oil flow may be restricted. If any error
IS fo~nd, ~ull ~ow-a~ea can be obtained by grooving in the
reqUIred directIOn WIth a half-round chisel.
It is essential after new shafts have been fitted, and advisable even if not, to check the truth of the wheels before final
assembly commences. One method is to apply a dial indicator
to the outer side of the wheel near the rim with the shaft
supported between lathe centres; another is to rotate the
assembly with the shaft running in an accurately-bored bush
or even in its own bearings, If there is more than .003 in:
r~n-o~t, !he s~aft wil! have to be set in the requisite directIon; It WIll be ImpOSSIble to line both shafts up accurately. if
one, .or both, are not dead true.
Strictly speaking, any balancing which is thOUght to be
necessary should be done at this stage so that each wheel
can be dealt with individually; it is, however, easier to work
on the wheel~ after they are assembled, so unless two odd
wh~els are bemg made into a pair the job can safely be left
untIl later, on the assumption that they were individually
balanced by the makers in the first place.
After all work on the components has been completed
assembly is a straight-forward job, although it is easy t~
overlo,ok the obvious precaution of ensuring that the oil~ol~s 10 the, crank pin, and. timing side wheel are accurately
m, line. ThIS p~ecautIOn IS not required when fitting pins
WIth annular oil grooves, which should be placed with
the ~ole leading to the rollers at the 3 o'clock position,
10~kmg towar~~ the drive-side with the pin at the top, this
bemg the pOSItIOn of least load, Cleanliness is essential
not only for the sake of the bearing but also to ensur~
that no particles of foreign matter are trapped between the
99
TUNING FOR SPEED
A simple method of
rough-aliglling the flywheels with the aid of
two wooden chocks
and a straight-edge
before filial checking
alld tightenillg. ...
abutting faces to the detriment, of mains,haft alignment.

If the pin is of the solid type, It can be Inserted ~nd the
nut fully tightened on one wheel only before assembling the
rest but if it is of the two-piece variety with a pressed-on
slee~e, it is not advisable to tighten either ~ut fullyynless
the other wheel and nut are in place, otherwIse the pm may
be pulled over too far to one side, leaving insufficient thread
for the nut on the other end. Grea~ care m~st be. taken
with tight parallel-fitting pins to avoId damagmg the hole
when fitting the second wheel, which is, liable to cant over
due to its overhung weight. The temptatIOn to p~ll the wheel
on with the nut as soon as a couple of threads project through
must be resisted, as so doing will overload, and very probably damage, the thread inside the n u t . .
.
After semi-tightening, the wheels can be hned-up fa~r1y
closely by means of a straight-edge laid across the nms
and, when reasonably accurate, the nuts are pulled up nearly
as tightly as they will go. As easy a way as an~ to move the
wheels relative to each other is to bump the nm of one on
a lead block or the end-grain of a hardwood post; only a
100
Where a pair of centres is 1101 available, the flywheel assembly

may be supported on a pair of bearings in a simple WOOdell
jig. A steel base plate, or sheet of thick glass, on which to
mount the jig, makes for added accuracy,
light blow is necessary, as the inertia of the wheels themselves does the trick, and no damage at all is done to the
rim surface.
For final tightening a properly fitting box-spanner with
a strong tommy-bar at least 2 f1. long is essential; the normal
type of tubular spanner is not really up to the job (unless
reinforced by a brazed-on ring turned to clear the crankpin
nut counterbore), but if made from chrome-molybdenum
steel tube, this type will do. The best spanners, of course,
are those made from solid steel with integral tommy-bars,
which can be purchased from any good small-tool factor.
Real'strength must be put into the final tightening, something of the order of 400 foot-pounds being necessary.
The final check must be made on the main shafts, using
an indicator reading to thousandths of an inch. If a lathe
or a pair of centres is available, the assembly should be
mounted thereon, taking care that the wheels are not
101
TUNING FOR SPEED
some very careful work can be done with a scraper applied

t~ th~ flywhe~l fa~s, b?t m.ost people adopt the easy way of
Dlppmg the mns ID a vIce. 10 the plane in which both shafts
are .. high." This process is all right if the wheels are of
steel and the pin is robust, but is not to be recommended
if the pin is of light section or the wheels are ca."t iron
There is a third error which is rarely met with uniess
the w~ee~s were not originally a pair, and that is when one
sh~ft ~s high and the .other low in the plane of the crankpin.
ThIS IS almost certalOiy due to the radius of the pin-hole
Whell checkillg a splilled mail/shaft for truth with a scriber
or dial gauge, all externally ground sleeve, which is a push
fit 011 the shaft, will prove of great assistance in making
accurate measurements.
deflected inwards by excessive pressure applied to the centres.

Next apply the indicator to each shaft as near as possible to
the ends, as all too frequently the centres in these shafts
get knocked about; any defect will immediately be shown
by the indicator, and must be rectified by scraping, lapping
or re-turning the centres until concentricity is obtained.
A number of engines have one or both shafts splined, and
in such cases a tightly fitting; accurately ground concentric
sleeve fitted over each spline will greatly assist matters.
The 'next move is to check the shafts close up to the
wheels; this will probably show one shaft to be " high "
and the other " low" in a direction at right angles to the
crank pin. If this is the only error present the situation is
good, because bumping the wheels in the appropriate
direction will (albeit after quite a number of shots) eliminate
it entirely, but if in spite of all your efforts both shafts are
.. high ". in the same direction, the whole assembly is in
effect" bent" in the centre.
This may be due to swarf between wheels and pin, lack
of parallelism between the outer faces of the pin shoulders,
or similar causes. If not due to dirt, the trouble can
sometimes be cured by refitting the pin in a different position should the oil-hole arrangements permit. Failing that,
102
These deliberately exaggerated

diagrams
depict three faults which
may be discovered upon
assembling the wheels.
Their causes and the
appropriate
remedial
actioll are outlined in
the text.
(Top right) Wheels out

(Above)
Wheels out of alignment, an easily cor. rected fault. (Right)
Odd wheels, necessitating a difficult machining
operation.
0/ parallel.
103
"HIGH
TUNING FOR SPEED
in one wheel differing slightly from that in the other and

can only be remedied by a machine shop operation.
If a pair of centres is not available, or the shaft-centres are
too badly damaged to be usable, another method of checking
is to support the assembly by its own or a s~milar ,Pair of
main-bearings resting in V-blocks. Alternatlvely, If there
are only two bearings the crankcase itself can be utilized,
although it is rather laborious having to separate the whole
.issue in order to bump the wheels whilst carrying out the
truing process-needless to say, it is a bad plan to attempt
to shift them while supported in their own bearings. If there
are two fairly widely spaced bearings on either side of the
. case the job is a little easier, as then one shaft can be run in
these bearings, leaving the other wheel and shaft completely
exposed for checking. These last two m~t.h~ds are, howe~er,
only suggested as makeshifts where facIlitles for employmg
one of the others are unavailable. Another workshop
method is to bore out a block to a close running fit on the
shafts and rotate the assembly with one shaft fitting in the
hole and the other in the air, a dial gauge fitted to an arm
attached to the block indicating the truth or otherwise of
the free shaft. One shaft is first tried, and then the other,
and this method is just as good as doing the job between
centres except that one cannot check both shafts at the same
setting.
.
The question now arises-how true should the shafts be?
Well, the correct amount of error is zero, but this is rarely
attainable, and if the sum of the errors indicated on both
shafts comes to less than .002 in. there is nothing much to
grumble about. Two-bearing assemblies can run with a
greater error than those with three or four, particularly
where the latter are housed in a very stiff crankcase. Having
made the shafts as true as they can be made, the nuts must be
given a last nip-up, and the shafts finally rechecked. !f.aU
i.s well, a couple of squirts of oil up through the feed dnllmg
in flywheel and pin to clear out any possible grit will finish
the job off, unless for any reason, such as a big change in
piston weight, you wish to rebalance the engine.
104
The Balance Factor

~he lowe~ half of the ~od can be considered as rotating
weight, and It, together With the crankpin, can be completely
balanc~d by an equal and opposite counter-weight, or by
removm m~tal from t~e crankpin side. But it is not possible
to do this WIth the recIprocating weight in a single-cylinder
parallel-twin engine for, if an equal and opposite weight
were added, the balance would be good in the direction
of t?e cylinder axis, but j.ust as bad as before along a line
at nght angles to the cylinder. Hence a compromise must
be .effecte? by ~dding ~nly a certain percentage of the
reclprocatmg weIght, thiS figure being termed the balance
factor.
It is quite useless to postulate any particular balance
factor as being the ideal; so many considerations enter
int~ the matter that it varies with almost every design of
eng me, .or even the type of frame in which engines of the
~ame kmd are mounted. That being so, do not be misled
m.to rebala?cing ~our engine just because one of your pals
With an entlrely different machine thinks he has some magic
formula of his own.
The only s~u.rce of ;eliable information is the parent
factory but, fallmg adVIce from that direction, a factor of
.66 of the reciprocating weight usually gives reasonable
results. Strangely enough, some makers use a smaller
factor for racing than for touring, while others do the
rev~rse; the idea in all cases, however, is to get an engine
~hlch runs most smoothly in the speed~range at which it is
mtended to operate for the majority of its life. It does not
~atter much if the engine feels" rough" at 4,000 r.p.m. if it
IS t? be raced and feels smooth at 6,000, whereas such an
engme would be ~ery. undesirable for fast touring, where
much of the runnm~ IS done at the lower r.p.m. Altering
the balance factor wlll usually succeed in moving the rough
period, if any, well away from the most-used speed.
The aforesaid balance-factor applies of course to the
reciprocating parts only-Le., piston, ;ings, gudg~on pin
105
TUNING FOR SPEED
and the top half of the connecting rod. A lot of weighing

and measuring can be eliminated by dealing with the complete flywheel assembly only, thus-wise. Arrange it on a
level surface so that the con.-rod (minus piston) is lying
horizontally, with the small end resting on one side of a pair
of scales, or supported by an accurate light-spring balance.
If the balance pan is other than flat the rod must be supported in the plane of the gudgeon pin. This is be~t done
by passing a short bar through the bush and restmg the
former on the edges of the pan. First, a few nuts and
washers can be placed in the other pan to counterbalance
the short bar and then there is no likelihood of getting confused with the other weights used to find the weight of the
small end, which we will suppose is found to be 6 oz. Next
weigh the complete piston assembly, which comes to, say,
15 oz. Thus the total reciprocating weight is 15
6=
21 .oz. and employing a factor of .66 the amount to be
balanced is, therefore, 14 oz.
The wheels are already bound to be balanced to. some
percentage, and if they are placed with the. sha~ts resting
on a pair of accurately horizontal metal straight-edges
they will eventually come to rest with the crankpin vertical~y
upwards, although they may.show a tendency for the pm
always to be slightly to one side of the dead centre-hne.
This indicates that the bob-weights are off-centre, a fault
which must first be rectified by drilling equal-sized holes
into the side of each rim, in a position at right angles to .the
pin and, of course, on the "heavy" side of the vertical
centre line.
When symmetrical balance has finally been obtained,
attach to the small end a weight equal to the amount to be
balanced minus that of the small end; using the figures
quoted above, this would be 14 - 6. 8.oz. The manner of
making up this weight or of attaching It to the small end
is purelv a matter of choice-it can be, for instance, a bag
full of 'oddments, or a bolt with the requisite weight of
washers. If the balance-factor does actually correspqnd to
the figure desired, the wheels will roll freely along the
106
A jig for checking balance call be made

from two lengths of angle iron bolted to
the bench and protruding sufficiently to
allow the assembly to revolve freely.
If, with a weight to give the desired
balance, attached to the small-end, the
crankpin comes to rest in the uppermost
position, the c,?unterweights must be drilled as shown on the left.
If the crankpm stops at the bottom the wheels must be drilled
adjacent to the pin.
stra.i~t-ed.ges and show. no tendency to settle in anylone

position; If not, the pm will go to the top or bottom
according to whether the bobweights are too heavy or too
light.
Correction is usually made by drilling the rims in the
appropriate positions, being careful to take equal amounts
out ?f each wheel and on each side of the centre plane,
bu.t ~t can equally we!l be done by tapping and plugging
eXlstmg holes. If a bit of experimenting to find the best
balau.ce is part of the tuning programme, it is a good idea
to dnll or tap a few holes of, say, i-in. or i-in. diameter,
into which plugs of the required weight can subsequently be
fitted or removed. In some engines it is possible to drill a
hole somewhere in each crankcase wall below the mainbearings, and tap it! in. gas, (or more correctly, ! in. B.S.P.)
thread which is the size commonly employed for drain
107
TUNING FOR SPEED
plugs, and drill several holes in the flywheels at the same

radius, tapping these out iin. B.S.P. Plugs of various lengths
in steel or bronze can then be inserted or removed from the
flywheels through the crankcase holes in a couple of minutes
without disturbing the engine and experiments to find the
most satisfactory balance can be very rapidly conducted.
This is a very good scheme to employ when adapting engines
for use in 500 C.C. racing cars, because the different method
of mounting as compared to a motorcycle yery often leads
to trouble in obtaining smooth running.
In closing, a word or two about those "horizontal
straight-edges." For permanent workshop use it is nice to
make up a proper jig, but there is no necessity to go to such
lengths. There are plenty of other ways: for instance, two
lengths of angle iron can be bolted to the bench-top overhanging the latter by about a foot. The only vital qualification is that the top edges must be flat, smooth and absolutely
horizontal and level with each other when the wheels are
resting on them. A good spirit level is the best aid to checking this point, but failing that a dead-round bar such as a
piece of silver steel will prove the point, since it will
obviously tend to run towards whichever is the lower end.
Most machine-shops equipped with cylindrical grinders
possess a pair of straight edges on which the grinding wheels
are balanced and, with a little blandishment exercised in the
right quarters, it is often possible to obtain the use of such
equipment.
CHAPTER VIII

THE primary duty of the crankcase is to provide a rigid
mounting for the main bearings and a foundation for
the cylinder sufficiently solid as to maintain that com~onent square to the mainshaft axis at all times, irrespectlve of temperature changes or cyclic load variations. The
rigidi.ty of the bearings is settled by the original design;
thus III the search for speed little can be done other than to
see that the outer races are a good fit in the case when the
latter is hot. If they are not, the races will show signs of
" creep" indicated by a polished appearance of the outer
surface. If there is any doubt about the matter, each side
of the case, complete with bearings, can be immersed in
boiling water, after which the bearing rings should not be
free enough to turn by hand.
HolY lYorn mainbearing housings

may be renovated
by the insertion ofa
shrunk-ill bronze
liner. (Inset) How
the edge of the
liner is chamfered
to permit of peening the aluminium
surround. J
.030" CHAMFER ON EDGE OF

SLEEVE ALUMINUM PEENED
INTO THIS CHAMFER AFTER
DROPPING SLEEVE INTO PLACE
BRONZE SLEEVE .002" PER
INCH OF DIAMETER LARGER
THAN HOUSING BORE
108
109
- TUNING FOR SPEED
I
I
Several methods are available for curing loose races,

of which the easiest-tinning the surface-is not to be
recommended, as it is almost impossible to get an even
coating, and in any case it inevitably gives away after a
while. Chromium or nickel-plating the races is satisfactory,
provided that the finished outer surface is absolutely circular and true to the bore; if only about .001 in. thickness of
deposit is required to restore the fit it will not need to be
finished to size if the job is done by a competent plating
shop, but deposits of greater thickness will need to be
finally sized by grinding owing to the tendency of these
metals to build up more thickly at the edges than at the
centre. Copper-plating on the other hand does not show
this tendency and can easily be cleaned up to size with a
file whilst the race is rotated in a lathe chuck. Owing
to its softness, copper plating does not always last very
long but is such a convenient method that it is often
utilized.
Probably the best method of effecting a permanent cure
is to bring the housing back to its original size by boring it
out and fitting a bronze sleeve aboutnr in. thick and .002 in.
per inch of diameter larger than the housing. Having heated
up the crankcase to 200 0 C. (400 0 F.)-which can be done
with the aid of a domestic oven-the sleeve can be dropped
into place, and after peening the aluminium over the edge as
shown in the diagram, the sleeve is then finish-bored to
.001 in. per inch smaller than the race. A job such as this
requires to be done extremely accurately; the crankcase
must be set up in the lathe with the join face and spigot
diameter both running dead true. In some instances, one
or more of the outer races are clamped endways, either by
plates or ring nuts, in which event they cannot turn or work
out sideways, and, provided there is no actual looseness
perceptible, the diametral fit need not be so tight. "Manx"
Norton bearing sleeves, besides being retained in this
manner are machined with grooves on the sides, into which
the metal of the retaining plate is punched to form keys to
prevent rotation. When refitting these sleeves they must be
110
placed with the grooves visible and the retaining-plate

punched in to them in the prescribed manner.
Main Bearing Clearances
If any bearings need replacement through excessive
wear or pitting of the race tracks, remember that the outer
races close in by approximately half the amount of interference between them and the case. Consequently, the
bearings should have a slight amount of slackness in them
before fitting, otherwise the rollers or balls may be subjected
to a damaging overload, even though the inner races may
still seem to turn quite easily. Some makers supply bearings
with various amounts of diametrical clearance to allow for
the tightness of shaft or housing, and mark the races
accordingly. Two small polished circles on the edge of a
. Hoffmann outer race, for example, indicate that it has the
right clearance if the outer race is tight, but the inner a push
fit; three circles are visible if both races are intended to be
tight. Apart from this indication, the inner assembly of a
roller bearing should just slide into place by hand if the fit
is correct, and ball-races, when free of oil, should spin for
several turns. Most ball-bearings, rotated dry, emit slight
noise but, if the noise sounds irregular or rough, the
condition of the bearing is open to suspicion and its replacement is the wisest course.
Another important matter is the alignment of the bearings, particularly if there are two in either or both halves
of the case. The only check of any real value is to turn or
grind a mandrel to a size (or sizes) which will just push
through the inner bearings without shake, and then insert
this through the whole set of bearings with all the crankcase
bolts tightened up. The mandrel should push through
without effort and turn freely when in place; if not, the
error must be tracked down and eliminated.
.
Sometimes in old engines the spigot on one half-case is
a poor fit in the register in the other half. and relative movement which would alter the alignment can take place. This
111
TUNING FOR SPEED

THIS FACE TO BE DEAD FLAT
DIMENSIONS A AND B MUST
BE EQUAL
The studs should be

removed, the case
placed mouth dowlIwards 011 a surface
plate and the distances A and B thell
checked for equality
with a mal/drel. The
end diameters of the
mandrel must be of
equal size.
. can be prevented by reaming out two existing bolt-holes

to accommodate special oversize bolts acting as dowels;
before reaming, the cases must be tapped lightly this way
and that until the best position is attained, as indicated by
the freeness of the test-mandrel.
Fortunately, since it is a difficult fault to eradicate,
incorrect bearing alignment is a rare disease, but it should
be looked for just the same.
If plain bearings are included in the make-up and have
been renewed, it is best to leave the bores a few thous.
undersize and finally line-ream with the reamer piloted in
the bearing on the opposite side. When a bearing of this
type is used to transfer the main oil supply to the crankshaft
the fit is particularly important and the maker's instructions
regarding clearances must be rigorously followed. Naturally
any communicating oil-holes must be accurately lined-up
and checked by injecting oil through them from a pressure
can.
The test-mandrel also comes in useful for checking the
cylinder-base, which must be parallel to the main-shaft
centre line in addition to being flat and free from any step
112
at the joint. Presupposing that it is accurate as to flatness,

parallelism can be checked by inverting the case on a surface-plate and measuring the distance on each side between
the plate and mandrel. Should the bearings be of differing
bores on each side, a short length of the mandrel at each
end can be turned to exactly equal diameters-which avoids
having to make allowance for varying sizes. If the face is
not fiat it must be re-machined or trued up with file and
scraper until it is correct in all respects. Lack of flatness
is bad; it permits oil leaks. allows the barrel to rock about
under load, and can be an unsuspected cause of barrel
distortion or cracked base-flanges.
At this stage it is as well to cast a critical eye over the
cylinder studs-particularly if the compression ratio is
being raised considerably, because the gas pressures will
be greatly increased thereby; the accompanying graph
shows the actual rise of compression pressure with increase
of ratio, and as explosion pressures are some 3t to 4 times
700
600
;I'
500
:r
u
ALCOHOL
/-
~"oo
'"ffi 300
",.
0..
'"a
PETRf5("
5o 200
//
0..
100
;/"'Ir~OMPRESSION
PRESSURES
FOR VARIOUS RAT/OS
(Assuming 14 lb. per sq. in.
preuure in cylinder at end or inlet.)
8 9 10 11 12 13 1'1 15 16 17 18 19 20
COMPRESSION RATIOS
A study of this graph will reveal the reason why cylinder

studs should be examined when the compression ratio is
raised.
113
TUNING-FOR SPEED
TAPPING
START OF LARGE THI\EAO
1/16t h BELOW fACE
SIZE
If cylinder studs are
loose, internally and

threaded
externally
sieeves will effect a
good repair. Aftenwave/y, double-diameter
studs call be made up
STUD LOCKED BY THREADS

TIGHTENING AT BOTTOM OF
HOLE
the compression pressures, it will be appreciated that a good

deal depends upon the cylinder fixing in an ultra-high
compression motor.
Almost invariably the studs are tapped direct into the
aluminium. As the latter metal is relatively weak, the length
of thread in engagement should preferably be twice-and
certainly at least one and a half times-the stud diameter.
In addition, the threads should tighten at the bottom and not
at the top of the hole; in the latter case the run-out of the
stud thread exerts a powerful wedging action on the surrounding boss, which may be the cause of a crack making
an unwelcome appearance either at the first onset or .later
on when the engine is put to work.
Should the crankcase threads appear to be damaged or
partially stripped, there is usually enough metal present to
~ permit drilling and tapping to a greater depth. If not, two
courses are available. One is to re-tap to a size .f.tr in.
larger and fit a double-diameter, or "bull-headed", stud;
for example, a stripped hole previously -& in.-Whitworth
can be re-tapped to i in.-Whitworth, having first opened it
out with a letter N drill. In other cases it may be preferable
to retain the existing size of stud by tapping out the hole to
take a bronze or steel bush, screwed inside and out to the
. appropriate sizes, this being particularly advisable if there
114
2 B.A.
l/4-in. B.S.F.
1/4-in. Whit.
5/16-in. B.S.F.
5/16-in. Whit.
3/S-in. B.S.F.
3/8-in. Whit.
1/8-in. B.S.P.
l/4-in. B.S.P.
3/8-in. B.S.P.
7/16-in. B.S.F.
7/16-in. Whit.
1/2-in. B.S.F.
1/2-in. Whit.
9116-io. C.E.I.
SIZES
POR VARIOUS THREADS
PITCH
THREADS
PER INCH
CORRECT
DRILL
SIZE
No.23
No. 5
No. 8
LetterG
LetterF
Letter DorP
LetterN
11/32 in.
29/64 in.
31/64 in.
3/8 in.
23/64 in.
21164 in.
13/32 in.
31.4
26
20
22
18
20
16
28
19
19
18
14
16
12
NEAREST
FRACfIONAL
DRILL SIZE
5/32 in.
13/64 in.
13/64 in.
17{64 in.
1/4 in.
21/64 in.
5/16 in.
is any future likelihood of the studs being removed and

replaced at intervals.
For the external threads on these sleeves, either rin. or
i-in. gas threads-both having 19 threads per inch-are very
suitable. The table on this page, which may come in useful
for other jobs, gives the tapping sizes and drill designations
for a number of pitches.
Still another scheme is to instal Armstrong "Helicoil"
inserts, which somewhat resemble springs made from steel
wire of diamond section, and are screwed into an oversized
tapped hole and finally locked in by punching or "staking"
the end helice.
Timing-side Mechanism
It is now time to give a little attention to the mechanism
contained in the timing-side case, to wit, the timing-gear
and oil-pump. Despite their relatively small size and heavy
115
TUNING FOR SPEED
loading, the camshaft bearings in push-rod engines give

very little trouble, and unless they are in a very bad state
it is best not to disturb them. Provided the surfaces are not
torn or worn into ridges, clearances up to .005 in. are permissible, and ifnew bushes are needed they should be reamed
to .002 in. clearance, except in the rare instances where they
form part of the lubrication system, when slightly less
clearance is necessary in order to prevent the escape of too
much oil.
The modern tendency is to provide fixed spindles upon
which the cam-wheels rotate, the spindles being drilled for
lubrication and it is an easy matter to make sure that all oilways are ~lear and that the wheels spin fre~ly. D~si~ns
in which the outer cam-bearings are located III the tlmIllg
cover are not so easy to check or to rectify if incorrect, the
chief difficulty being to get both pairs of bearings in line
simultaneously.
If the cover is attached only by set-screws its location
may not be sufficiently positive as to ensure that it remains
in the correct position indefinitely when under the influence
of irregular load and severe vibration. This contingency
can be entirely prevented by' fitting a couple of dowels,
which need only be quite small-l-in. diameter will do if
there is no room to fit a larger size-the holes being drilled
and reamed after the cover has been worked into a position
where the cam-wheels are at maximum freeness with all
screws tight; with journals and bushes free from oil, either
wheel should spin for several turns if given a sharp flick
with the fingers.
The dowels should be a light drive fit in one component
and a good push fit in the other, otherwise it is difficult to get
the cover off subsequently. If there is no room to fit dowels,
two of the existing holes can be modified to accept screws
with close-fitting plain shanks, which will serve the same
purpose.
Replacement bushes are almost invariably bored slightly
undersize to allow for reaming. This operation must be
done with a piloted reamer to ensure alignment, and
116
occasionally a special jig-plate is really required to do the job

properly. Failing this device, which is only worth having if
it is extremely accurately made, the bushes, after being fitted
with the crankcase or cover heated, can be reamed a little on
the small side, and then eased out with a scraper, using
prussian blue on the shafts to show up the tight spots. Do
not attempt to lap the shafts in with emery powder, as this
will embed itself into the soft bearing metal and subsequently
cause rapid wear of the shafts. In the matter of end-play,
it is better to err on the generous side rather than have too
little, except when the gears have helical teeth, in which event
only .001 in. to .003 in. should be allowed. Excess play in
the timing gears of Velocette engines is the frequent cause
of harsh, noisy running.
Several modern engines, the A.l.S., Matchless and Vincent
for example, employ porous-bronze bushes for camshafts
and also at other localities. These bushes should never be
reamed if at all possible; the correct method is to size them,
after fitting, with a planishing broach-i.e. one with no
cutting edges-or else bore them with a sharp single-point
tool. The ordinary shop reamer almost invariably smears
the surface instead of cutting and thus blocks up the pores
in the metal, though a very keen reamer will not do much
harm if the amount of metal to be removed is very slight.
Increased Lubrication
Lubrication of the timing-gear is not always well carried
out, and can frequently be improved by chiselling extra
V-grooves leading to oil-holes in the bushes, or by arranging
an extra oil-feed to the timing-chest. In dry-sump engines
an easy way to accomplish this is to take a lead off the pipe
which returns oil to the tank and couple this up to a union
and internal jet in the timing-case placed where it will do
the most good-preferably so that the oil drops directly on
to the cams. Naturally there is a limit to how much oil
can be re-circulated in this way, and the flow needs to be
restricted, otherwise the scavenge pump will be overloaded.
117
TUNING FOR SPEED
WORK ON TIlE CRANKCASE
A fixed jet with an orifice about .030 in. in diameter should

suffice to start with, or alternatively an Enots adjustable
drip-feed can be used. Small jet orifices are rather prone
to blockage when return-oil is being fed through them, but
this can be largely overcome by using a larger hole into
which a loose wire or split pin is fitted. This scheme is used
by Vincents in the rocker-feed jets, but, for racing, the
restrict or wires should be removed from the exhaust valve
feeds and also from the inlet-valve feeds if the valve guides
are in good condition.
If the lubrication is of the total-loss type, it may be worth
while fitting an additional sight-feed pump, or if the
existing pump is of the single-delivery type, to replace it by
a duplex pattern, the aim being to let one pump feed the
big-end while the second takes care of the timing gear.
Each supply can then be adjusted to suit its own department
without risk of swamping or starving other parts of the
motor. This of course is the system used on J.A.P. Speedway and 8-80 engines which are intended for short-distance
work, but full dry-sump systems are used on these engines
for long-distance racing and Formula cars.
. The ~il-pump in dry-sump engines is usually housed

~hr~t1y m the ~rankcase, and in most o.h-camshaft designs
OIL
FROM_
OIL TANK
our
Y~~~a_ SUPPLY
ENGINE
N~~a...:~ TANK
OIL IN
The efficiency 0/ a gear-type oil-pump depends upon the minimum possible side clearance between wheels alld case. The
case/aces can be rubbed down when clearal/ce becomes excessive.
118
IS tIghtly fitted mto a recess bored in the back of the timing-
ches~. ~o .remove, it is necessary to heat up the case and
t~p It, tlmmg-chest down, on the bench-top. It may sometimes be necessary to run a tap into two of the retaining
scr~w holes, so that a couple of screws can be inserted by
WhIC~ the pump can be pulled out. The internal bore of the
housmg lI!ust neve: ~e enlarged to make refitting of the
pump eaSler, as thIS IS almost certain to lead to leakage
between the pressure and scavenge side of the pump. As a
rule, gear-pumps give very little trouble, unless a piece of
metal has found its way in and damaged the teeth or the
gear-p~ckets, but in time side wear will develop and reduce
the delIvery rate.
It is sometimes thOUght that backlash between the teeth
affects the pumping, but it makes no difference at all-in
fact, for smooth functioning at least .005 in. backlash should
be p:esent. ~he side .clearance, however, should not exceed
.~O I m. (partIcularly If the gears are thin in relation to their
dIameter), and if more is present the faces of the pump
body should be very carefully dressed to the required
a.mount. Should a paper packing be a standard fitting, the
sI?lplest course.is to lap the pump-face on a surface-plate
WIth the gears ~n place until the whole surface, including
that of gears, IS . fiat and level; the paper packing will
subsequently proVIde the requisite clearance. Pre-war pushrod Velo.cettes had double-gear pumps, somewhat similar
to those In the o.h. camshaft models, but driven at a much
sl?wer spee? by a sin.gl~-start bronze worm-wheel engaging
wIth. a stral.ght-cut pInIOn on the pump spindle, the pump
housmg bemg bored at a small angle to give correct
engagement between worm and pinion teeth. On later
models the pu.mp speed ~~s dOUbled by using a two-start
worm and h~hcally-cut pInIOn; and it is possible to utilize
the .later desIgn of geanng to increase the oil-flow of earlier
engmes.
In those pumps which have a plunger which rotates and
119
TUNING FOR SPEED
. ~ ECCENTRIC DRIVE
~ ON
SHAfT NUT
TRUNNION
As in the case of the

double-gear wheel pump,
the plunger type employs
a scavenge outlet with a
capacity larger thall that
of the feed pump.
0&
~7
TO OIL TANK
i) SPRING
NON-
LOADED
mu,," " " "
reciprocates, the points to note are the cond~tion of ~he

gear-teeth, the fit of the plunger portions in their r~s'pectIVe
housings at each end, and the state of the small dnvlI~g peg
which is screwed into the crankcase and engages WIth an
inclined groove machined in the plunger. As these pumps
turn at a relatively low speed, wear is very. slow as a ~ener~l
rule; it is a wise plan, nevertheless, to venfy that. alliS as ~t
should be for even a momentary shortage of 011 at maxImum revs: is fraught with the most expensive, and sometimes
physically painful, consequences. The drivi,ng. wor.m, if
renewable, should be replaced with0,ut heSItation If the
teeth have worn thin, and, when the design has bee~ changed
from bronze to hardened steel, it is desirable to mstal the
new design in any event.
. .
Reciprocating plunger pumps are not deSIgned for
working against pressures of more than a couple of pounds
or so to the sguare inch, and if any blockage should occur
120
in the outlet passage or pipes from either the pressure or

sca~enge ends, ~ither the worm. drive or the driving peg
which engages With the cam-slot m the plunger will fail in a
fairly short time. Therefore this point must be carefully
watched both inside the eng me and in all the external
connections.
Apart from that eventuality, this form of pump lasts for
a long time and has the merit that it does not allow oil to
syphon back through it whilst standing, which is an annoying defect of the gear-pump.
When plain big-end bearings are used, it is essential to
~aintain a high oil pressure, consequently they are lubricated
eIther by a gear-pump or one with purely reciprocating
plungers as used on Triumph and some Ariel engines.
These also give little trouble but the trunnion block should
be examined for we~r and care taken to see that the plungers,
ball-valves and sprIngs are all operating correctly.
. Should an engine ever suffer the misfortune of having a
pIston break up badly, or have a hole burnt in the crown
the resulting mass of aluminium particles penetrate int~
~very nook and crevice with astonishing rapidity, and are
hkely to congregate in odd corners or pockets such as exist
where two drilled oil passages meet, or where unions are
screwe~ into bosses. Flushing-out cannot really be guaranteed
to get nd of them all and the only safe way is to remove all
plugs and unions and make a thorough job of the cleansing
pr:ocess. When replacing these parts, remember that the
slightest trace of an air-leak on the suction side of the
scave?ge pump. is certain to ~ead to poor scavenging and,
even If the engme does not 011 plugs as a result it will be
down in speed due to oil-drag. Consequently,' too much
care cannot .be. taken to see that every joint, or plug, is
completely aIr-tight and locked up so that it will remain
so indefinitely, this remark applying also the driving-pin of
a rotary pump whic~, after being fully tightened, should
have metal punched mto the screw slot as an insurance
against loosening.
E
121
CHAPTER IX

WITH the exception of the cam-fol!O\yers and .overh~ad
gear, practically all the components ~Ithi~ the engIne whIch
demand attention or possible modlficatl~n ?ave by ~ow
been dealt with sufficiently thoroughly to IndIcate the hnes
upon which to proceed.
.
.
If everything has been done to your ow~ sa!lsfactlonand the more self-critical you are the more likelIhood there
is of your beating other and more easily satisfied rivalsa spot of assembly work can now be put in hand prior to
getting down to the vitally important matter of valve and
ignition timing.
.
It is probably unnecessary to me?tion the point agaIn, but,
to be on the safe side, let me remmd you that before commencing this work the bench and its surroun.dings ~ust be
swept clean of all foreign matter-dust, filmgs, cigarette
butts-and the crankcase halves, flywheels, bearings, ~nd
so forth laid out, preferably on sheets of paper, after havmg
been thoroughly cleaned.
.
The first step is a trial assembly of the flywheels In the
crankcase in order to verify the centralization of the con.-rod
and the amount of end-float, if any, on the mainshafts.
Lateral location of the flywheel assembly is normally ~ffected
by one of two main methods: either one of the mams?~fts
is clamped endways in its bearings (~sual1y on the ?flvmg
side), or, alternatively, the main bearIngs on both SIdes. of
the wheels share the duty between them, each preventmg
.
movement towards its own side of the engine:
In the first-mentioned category the questton 0: get~lll:g
the amount of end-float correct does not arise, smc~ It IS
fixed and limited to a very small amount by the design of
the location bearing. Care must be taken, however, to see
122
that any distance-pieces or washers. either next to the wheels

or between the various bearings, are in place in their correct
order; otherwise, there is a risk that some bearing not
intended to take end-thrust may be subjected to loading of
this nature, which, in addition to causing stiffness, will
rapidly destroy the bearing involved.
In the G3L Matchless, for instance, two ball bearings,
with a pair of spacing washers separating their inner races,
are fitted on the drive side, and should these spacers be
inadvertently omitted, the pressure of the shock-absorber
spring will put very heavy thrust-loads on the bearings, and
may ultimately cause the whole flywheel assembly to move
out of position in the direction of the chain. In this design.
as in quite a number of others, the end-thrust of the shockabsorber spring is ultilized to clamp the drive-side mainshaft
in its bearing or bearings; thus, when verifying the flywheel
location, the spring should be lightly tightened with the
sprocket and face cam (or an equivalent distance piece) in
place.
In this condition there should be no measurable side play
It is all easy matter to

mislay ar misplace essential washers ar packing
pieces which vary cansiderably with individual
makes.
This draWing
shaws the correct methadof
assembling drive-side bearings IIsed for many years
for the Matchless G3L
series.
..,J..-.,--":-'--.--.J BALL RACE
TO BE INSIiRTEO
'-;;~;;;;;;;j;;:, WASHERS
L
123
TUNING FOR SPEED
present, but if the spring is then removed, there should be

a perceptible amount; if not, the bearing on the timing side
is too close in and appropriate measures must be taken to
obtain, say, .005-in. float and thus avoid the possibility of
undesirable end-thrust being placed on the bearing.
In other designs such as the post-war Vincent singles
and twins, a distance-piece is used between the inner and
outer main bearings. If this is inadvertently omitted, the
shaft will be locked up solid when the main shaft nut is
tightened up, an operation which may not take place until
the motor is almost completely assembled. It is therefore a
golden rule always to assemble all relevant components
on the mainshaft and tighten the nut fully early in the piece
just to make sure that all is in order. There is never very
much clearance to spare in this locality and it may be that a
new crankpin is a fraction longer than the original and fouls
the crankcase, though this might otherwise have escaped
notice. In some engines, the crankpin nut runs so close to
the bearing housing that a portion has to be cleared away
with a chisel and this naturally must be done before
proceeding further.
In those designs in which the bearings on both sides of the
wheels assist in their location, it is not uncommon to allow
up to .OlO-in. end-float, except in the case of o.h.c. Velocettes,
in which the mesh of the bevel timing gears is partly controlled by the position of the mainshaft. End-play to the
extent of .010 in. would be very undesirable; indeed, it is
preferable for the wheels to be slightly tight sideways to the
extent of two or three thou. when cold. The thermal expansion of the crankcase, which will be four to five thou.
greater than that of the wheels, is then sufficient to ensure
freedom of movement at running temperatures.
In o.h.c. Nortons, the assembly is located by the timingside ball-race in order to limit the amount by which thermal
expansion can affect the bevel mesh. Centrality adjustment
is made by fitting shims between bearing and flywheel and
should be verified with all timing side components filled and
pulled up tight.
124
All.other method, employed ill the Ariel factor .

POlflter on the drive-side mail/shaft just touch;';" IS to fix a
en e whe"ls are pushed/ully over to the~f!~lIc;~;~~-
Tt: :1
clearall~: be~~e:1I e~a::~~~eiJ~:j~~f!:t~~it:h::k:Ji~':i/hIl1

feeler gauge.
li;lit\t~e :heels placed correctly in the case and the latter

y 0 te up, the small end of the connecting-rod oUght
~o ble exactly central in the cylinder-barrel register but as
III
a most every d
'
eSlgn
l'
~ In.
,
or more side c1earan
~lIowed.betw~en the small-end bUsh and the piston bOs~:s 1=
lIttle
is permissible . The re IS
. a Iways' a
ce
t .latItude In centrality
.
h r ~n amount of ~Ide play of the rod itself if the big-end
:s ~eD correct~y fi~ted up, and so the simplest method of
c ec ~g centrality IS by callipering the distance between
~h~Jeglster ~nd each side <?fthe.small end, with the rod lightly
e over In the OppOSIte dIrection. The two measure
mb enthit~ should not differ by more than .010 in. when checked
y s means.
to Ift~he difference is pe~ceptibly greater, it is not a good idea
a ~mpt to correct It by bending the rod over such a
proc~ ure would simply undo the value of the work
preVIOusly carried out in truing up .the rod. Instead
125
'
TUNING FOR SPEED
The overhead-camshaft engine diflers fro/ll the o.h.v. alld

side-vail'e types in that elld-float ill the jfywhee? assembly
must be avoided, otherwise the bevel gears wIll tend ~o
mesh illcorrectly, thereby permitting backlash. Fill! details
of the correct treatment are given in the accompanYlllg text.
correction must be effected by altering the lateral location of

the flywheels-hence the reason for the trial assembly be~ore
finally adjusting the end-float. Should the latter be exceSSive,
obviously the side on which it is taken up should be the one
which will tend to centralize the small end.
There are various methods of adjusting end-float, but
most of them entail the use of shims, which, if not available
from the makers, can be cut from shim-steel stock or alternatively, use can be made of the shim-washers commonly
126
employed for adjusting the end-float in the steering knuckles

of cars. The location of the shims varies, but the golden rule
is never to put them between two parts which rotate relatively
to one another, but only between faces which are relatively
stationary. Thus, they can be placed on the mainshafts
between the inner races and flywheel bosses, or between the
bearing outer rings and the inside of their housings.
The latter entails rather more work in fitting. but in some
designs is the only suitable place. For example, Velocette
mainshafts are ground with a very narrow taper, and the
races are designed to fit tightly thereon when hard up
against the flywheels; obviously, if shims were interposed
the fit of the inner races would be slack, so the shims must
be placed behind the outer races.
Again, in many J.A.P. models the rollers run direct on
the hardened shafts and end play is fixed by the clearance
between the faces of the outer rings and the flywheel bosses.
If this is too great, these rings must be shimmed up from
behind. At the same time, it is wise to verify that the rollers
have sufficient end clearance between the flywheels and the
hardened washer placed at the other side of the outer race
otherwise they cannot rotate freely and will rapidly destro;
the shaft surface. In the racing twins of this make, phosphorbronze thrust washers are interposed between the flywheels
and bearing sleeves and side-float can be regulated by using
shims of varying thickness.
W~en building up an old engine with new parts, it will
occaslOnally be found that there is no end play; there is,
so to speak, negative clearance. In that event, first heat the
case, drop out the main bearings, remove any shims there
may be, and then try again. If this treatment is of no avail
the only thing to do is skim the requisite amount of metai
off the flywheel bosses-the amount being determined by
measuring the distance between the main bearings with
the case bolted up and subracting this from the width over
the flyv.'heel bosses. For this job a large pair of external
and internal caHipers (or an internal micrometer) are
required.
127
TUNING FOR SPEED
REASSBMBLING CRANKCASE COMPONENTS
The use of iI/side alld

outside micrvmelers
will determine the
amount of end-float;
the mlilor reading is
subtracted from the
major olle. This is
a good method for
camshaft
engines
wldch employ "nil ..
clearal/ce, bill the
amateur who has I/O
such elaborate equipmelll will /ind II pair
of iI/side and oil/side
campers a reasol/ably
satisfactory
substitute.
Having finally obtained a state of perfection in which the

rod is central and the end-float correct, the flywheels should
spin with the greatest ease as the rod is pumped up ~nd
down with the fingers, and there should be no perceptible
dilference in .. feel" with or without the sprocket and shock
absorber in place. If there is, the cause must be tr~cked
down and eliminated forthwith and not glossed over III the
hope that it wm run off shortly in service. If all is well, the
crankcase can be joint-coated with the preferred f~rm of
jointing compound and reassembled .. The author s preference is for one of the non-hardenmg cements such as
" Gasket-goo", which does not dry quickly so that there
. need be no undue haste to slap the case together before ~he
compound hardens. Make absolutely sure that. nothing
has been omitted which should be filted at thIS stag~,
particularly in unit-construction engines in which certam
gearbox components play have to be in place before the
case goes together. Even though some may h~ve to be
refitted subsequently, it is preferable to fit and tl?h!en all
the crankcase bolts at this stage, but at all events It l~ very
bad practice to put in a few in the region of the cylmder,
leaving the bottom ones out because they are not easy to
get at just at the moment. Subsequent joint leakage may be
traced to this cause.
128
Some crankcase bolts are drilled for cylinder-base oilfeeds and care must be taken not to overstress them. As
part of the checking routine, inject oil with the pressure can
through every oil passage there may happen to be to make
absolutely cert~in all are. clean and to wash out any foreign
matter. Sometimes an oil hole may be blocked with excess
gasket cement and it is beter to clear it at this stage than after
some mechanical damage has been wrought later on.
The next step is to lit the piston, minus rings as a pre.
!iminary to checking its" lie " in the barrel. If'the piston
IS new or undamaged, the bottom edge of the skirt should
1
f
All error ill workmanship may necessitate

setting ofthe connecting
rod; this call safely be
attempted if a block of
metal is first placed hetween crankcase and
connecting rod, wilh
tlte latter at boltom
dead ceflfre.
be par~llel to the crankcase face in a direction parallel to

the mamshafts, but. owing to the various amounts of side
play present and the shape of the bottom edge, which is
frequent.ly by no means fiat, this is not an easy matter to
check WIth real accuracy. In any case, it does not matter a
great deal; what does matter is whether the piston lies centrally in, and paraIlel to, the barrel.
. T~is is che<:ked by lightly bolting the cylinder in place and
VleWlllg the pIston from the open end of the barrel under a
good ~ght. It. is. then quite easy to see just where the piston
does he, and If It shows any undesirable tendency to work
over to one side when the flywheels are rotated.
If ~t ?~es so, it is a sign that an error-or, more likely, a
multipliCIty of small errors all tending the same way-is
129
I
I
TUNING FOR SPEED
causing trouble. Sometimes reversing the piston will effect

a cure, although this method, of course, cannot be adopted
(except as a temporary method of checking) with a piston
which has unequal-sized valve clearance pockets in the
crown, or with the split-skirt pistons fitted to a few sports
models which must not be run with the split towards the rear,
or thrust, side.
Even if the top lands appear to be central, this provides
no guarantee that the piston as a whole is lying parallel.
This can best be verified by turning the engine over several
times after coating the cylinder bore with a very light (almost
imperceptible) smear of prussian blue. Both thrust faces
of the piston should then be symmetrically marked, but if
differences of only a minor nature exist, a little careful work
with a dead-smooth file will usually rectify matters. Severe
cases call for a little more drastic treatment, and it may be
necessary to set the small end slightly. although this will
be a reflection on the quality of the previous work. To
perform this operation, position the rod at bottom dead
centre and select a block of metal which will just jam into the
space between the crankcase mouth and the shank of the rod
about half an inch below the little end. Insert a mandrel
(not your checking mandrei!) into the small-end bush and
bend in the requisite direction, remembering that only a
very smaUamount is required to make a big difference to
the position of the piston, owing to the latter's relatively
greater length.
Once the piston is really true, it ought to be possible to
make it reciprocate (without rings) at least haIfa dozen times
by giving it a smart push with the thumb downwards
from t.d.c. and it should always come to rest at b.d.c.,
because the flywheel counterbalance weight is less than that
of the piston and rod.
Before finally completing the assembly, temporarily fit
the barrel and head and pull the nuts down to correct
tension to verify that no barrel distortion is thereby brought
about. If distortion is present, it will be shown up by the
engine not being quite so free to turn as it was before. Lack
130
of flatness or ~quareness at the cylinder-base flange or

IS the most probable cause, and this can
sometimes be detected by observing the effect of loosenin
thedbolts one or two at a time. Final fitting is dealt with at th~
en of Chapter X.
The} ~olden rules during assembly are always to check
e~'efyt llngyoudo a
.
.
s you go a long, Glut not to assume allY thin
IS. corre~t WII/lOut verifying it. Many hours of hurried mil
~Ight toll woul? have been avoided time and again had these
SImple precautIons been observed.
crank~ase face
131
ORIGINAL CROWN CONTOUR
.l,.
,.
,,
CHAPTER X
The masking effect of

deep recesses may be offset by milling down the
original crowlI confour
and forming flats.
,"
I
....
....
"
I
WE have now proceeded to the point where the barrel is
ready for final bolting-down, provided that there is no

further work to be done either upon it or the piston.
If, however, larger valves or a high-domed piston have
. .
been fitted, or the barrel shortened to any extent, It IS
essential to verify that the valve pockets in the piston crown
are sufficiently deep, and if not, to deepen them, which will
mean that the barrel must come off again in order to carry
out the re-machining or to get rid of swarf if the pockets are
deepened with the piston in place. Another matter which
can be attended to at this stage is the head joint, a vital
factor which will be touched upon later.
Reverting to the question of valve clearances, in themselves piston recesses are ~o earthly use-in fact, they are
detrimental, inasmuch as they break-up the crown contour
and create pockets of gas in which detonation may take
place. In extreme cases they may even render a considerable
portion of the valve-opening area inoperative for a few
degrees before and after t.d.c., so .reducing the effectiveness
of high-overlap valve timing at an exceedingly important
period of the cycle.
Although this is only likely to be serious if ultra-high
compression ratios are being sought after, the tendency to
masking of the valve heads can be reduced by milling flats
on the crown-to, say, half the maximum depth of the pocket
-and counterboring into the flats to make up the rest of the
clearance required.
Another point to be watched is the thickness of metal
between the corner of the pocket and the top ring groove;
this can be quite thin on the inlet side, down perhaps to
132
in., but on the exhaust side as much metal as possible

must ?e left. Insufficient thickness at this point allows an'
excessive amount of h~at t? pass through to the top ring;
moreover, the upstandmg nm of metal will become very hot
and consequently weak and liable to crumble away if for
any r~aso~ the carburetter is delivering a thin mixture.
This bnngs up the question of how deep the pockets really
need to be, to w~ich the answer depends to a large extent on
the type of ~ngme and the conditions of racing. In o.h.
camshaft designs (provided the springs are of the correct
stren?th) t~e valves follow the cams very closely up to a
cert~m designed speed. Provided, therefore, the revs. are
studIOusly kept below the danger mark by the use of a rev.
co~nter, the clearance need. be only a little greater-say,
lD.-than that necessary Just to clear the valve-heads at
the ~oi?t at w~ich the valves and piston are in the closest
proximity. This. is no~ necessarily at t.d.c. but may be a
few degrees on.el~her ~Ide, depending entirely upon the cam
contour and t.lmmg m relatIOn to the piston movement.
The danger pomts can be determined by temporarily setting
up the head with light springs on the valves and checking
the amount of clearance under each head at 10 intervals of
crank rotation, measured by a degree-plate on the mainshaft.
.Another method is to fill the pockets of an old piston
WIth deep recesses with plasticine and rotate the engine
so that the valves form impressions in the soft material
These will be the minimum depths, and thermal expansio~
133
TUNING FOR SPEED
"TOP-HALF" RBASSBMBL Y
Recesses in the
piston crown should
be radiused off. If
they are too deep
the top ring land
will tend to o)'erheat.
of the piston, any rocking of this component in the barrel

and take-up of bearing clearances, all tend to reduce the
safety factor.
Owing to the greater possibility of whip or spring in pushrod mechanisms, particularly if the rockers are not too well
supported, a little more clearance is desirable in o.h.v.
units, even for road racing work where the rev. counter can
be kept under observation. In other varieties of racing,
particularly scrambles or sprint events where, in the he~t c:f
battle, engines are liable to .be grossly over-revved, It IS
best to play safe and to cut the pockets sufficiently deep to
clear both valves at their fuIllift.
Failing the use of a miiling machine, the common
method of deepening the pockets is to file cutting-teeth on
the heads of an old pair of valves and, after bolting the
cylinder head on with the piston locked at t.d.c., to rotate
those valves with the aid of a tap-wrench or similar tool
damped on the stems. This scheme has the merit that the
actual clearance can be measured as the work proceeds; but,
of course, the radius of the pocket thus formed is exactly
equal to the valve head. It is obvious that in running conditions the rim of the valve head will almost certainly scrape
the pocket sides, therefore it will be necessary either to use
valves (or, better still, a proper cutter, which can easily be
made from a piece of hardened tool-steel) about -h in. larger
in diameter than the real valves, or else to remove the
necessary amount of metal by scraping, a laborious and not
very easy job to do nicely.
134
~f you are intend~g to experiment with compression

ratlOS and have provided for one or two packing plates
beneath the cylinder, it is advisable to arrange the valve
clearances to suit the highest obtainable ratio.
As a final touch the edges of the pockets should be
rounded ?ff and the piston crown polished, preferably on a
power-drIven cloth mop with Tripoli compound' small mops
or " bobs" which can be fitted in a drill-ch~ck, or to a
Unduly deep recesses tend to decrease the effective area

of the inlet vah'e.
flexible ~h~ft as used for port-grinding, can be purchased

for a shllImg or two, and this form of polishing gives a
s~rface superior to anything which can be obtained by hand
wIth the aId of metal-polish.
. The value of Cr?WD polishing, li~e that of port polishing,
IS a debatable pomt, as some engmes do not give of their
best until a perceptible film of carbon has formed. The
sequence of events is actually this; a bright, highly-polished
surface absorbs less heat than a dull black one and as less
hea~ is extracted by it from the burning mixt~re. more is
avaIlable for conversion into mechanical work. This is the
135
TUNING FOR SPEED
state of affairs when the engine is first run, but shortly after,
if using fuel other than alcohol, a film of carbon forms
which greatly increases heat absorption, thereby reducing
power and increasing the piston temperature. Carbon,
however, is a very bad conductor of heat; therefore, as the
thickness increases the heat able to pass through the
deposit to the metallic surface becomes less, and so the
power returns almost to its former figure.
If using alcohol a polished crown will remain bright for a
long time, and, whatever the fuel used, is much easier to
decarbonize. Generally speaking, a high degree of polish is
preferable.
When an extremely high dome is used in an endeavour to
obtain a very high compression ratio, there is a danger that
the plug will be badly masked. The close proximity of the
crown to the plug points blocks the rapid spreading of the
flame-front which is vital to good power production at high
speed, and this masking may commence quite a long way
before t.d.c. measured in crankshaft degrees, particularly
if the shape of the head approximates a hemisphere rather
than being flat. The result will be that engine power
will fall off badly at high r.p.m. unless the ignition is
advanced an excessive amount, and even this expedient is
no real cure as it is merely trying to eliminate one error by
putting in another.
Sectioning a wax cast of the combustion space as recommended in Chapter IV will clearly show the true state
of affairs which can be improved by forming a local depression in the crown near the plug, or even making the
whole crown unsymmetrical. In any event, it is better to
use a lower compression and obtain good combustion
than to aim only at maximum compression)atio.
Cylinder-h~d
Joints
Next,attention can be given to the head joint, which can

be made gas-tight in a variety of ways. The" doubleground" variety in which simultaneous contact is made
136
"rOP-HALF" REASSEMBLY
bet~een t,:,o areas (but .at different pressures) is widely used,

and IS, particularly ~ood If the head is not very robust and therefore I,lable ,to be distorted by the tension of the head-bolts.
ThiS desIg~ takes the form of a recess in the head which
fits over. a spIgot on the barrel, recess and spigot being of
substantially equal depths. To form a gas-tight joint the
two components ar: ground together, with fine grinding
paste between the spigot faces and coarse paste between the
broa~ outer faces through which the bolts pass. The result
of thIS mefhod is to leave a minute gap hetween the broad
faces when the ~pigot faces are in actual contact; thus when
the bolts are tIghtened heavy pressure is applied to the
latter to form the gas-seal, but distortion is prevented by
the broad faces coming into contact.
During grinding the coarse paste should be continuously
renewed, but as the spigot faces approach perfection the
fine paste can be gradually reduced in quantity by wiping
the barr~l spigot ~lean and adding a drop or two of oil,
af~er which, attentIOn the process is continued until the
SPl~ot a~d Its recess acquire a bright, smooth finish over
their entire ~rea. The broad faces will have (when the
coar~e paste IS cleaned off) a matt surface which oUght to be
cont1OUO~S, although ~ne or two small patches of indifferent
contact WIll not be sertously detrimental.
. In the.tu:alasse~bly no jOin;ing compound is necessary10 fa~t. It IS ~des1rable. BeSIdes being gas-tight, this type
of JOInt perrmts a good flow of heat from the underside of
th~ exh~ust valve seat to the barrel, and so assists in keeping
thIS ,regIOn of the cylinder head cool. On the other hand
f~edl~g.he~t back locally into the barrel IS not good practice:
SInce It IS ha ble to cause distortion therein.
, This sn.a~ is avoided i~ some of the later designs (employIng alummlU:n heads With thick basic sections and particularl~ effective finning) by permitting the two components
to be In contact ?nly o~er a s~all area, preferably at the top
of the barrel regIster, SInce thIS method eliminates the smal1
annular gap which would be present if contact took place
on the outer faces.
137
TUNING FOR SPEED
The principle ofthe

"double-ground"
head joint, to be
found ill several
successful racing
designs.
UP UNDER BOLT
PRESSURE
10 top rake, a slow cutting speed and a comparatively

heavy feed. On no account must the tool be allowed to rub
or skid on the work; if this happens the surface will immediately become hard and impossible to cut unless the
tool is run back and then dug in under the skin.
The coefficient of expansion of the steels mentioned is
.000017-18 in. per degree C., but there is another typethe Firth-Brown product called N.M.C.-which has .a
coefficient of .000022 in. per degree C., nearly equal to that
of many aluminium alloys. Owing to space limitations the'
he~d studs on KIT Velocettes are only
in. in diameter and,
bemg tucked away under the rocker boxes necessitate a
special short-headed T-spanner, which mu~t be handled
with care to avoid overstressing the studs, which if replaced
must be of ~igh-tensi1e steel. Early-pattern Vincent engines
are fitted WIth large hollow bolts, through which pass the
smaller bolts that carry the frame stresses. The large bolts
have large nuts but must not be tightened to more than 30
pound-feet torque, otherwise they may break when the
cylinder expands; the later-pattern solid bolts are not subjected to this restriction.
In some engines it is necessary to fit the head before
assembUngthe rocker gear, but in others it is more convenient
to reverse the procedure. In view of work described in
Ch~pter XI, dealing with valve mechanism as a separate
subject, let us suppose your engine to be in the first category and that it is, therefore, now ready to have the barrel
and head bolted on.
The procedure is quite straightforward. First, just make
sure that the gudgeon circlips are all in place, then, with
clean fingers, smear the rings and skirt of the piston with oil
of the sa~e grade you intend to use for racing, making sure
t~at the nng grooves are full, for this will provide a store of
od to cover the starting-up period. Smear the whole surfac~ of the. bar~el likewise, .then slip it over the piston,
easmg the rIngs mto place With their gaps staggered round
the circumference; do not omit the base gasket or packing
washers if any are to be used.
For this type of joint laminated copper gaskets ,are

excellent, but they should always be ~nnealed. every ~nne
they are fitted. Annealing of copper 1S done 10 precIsely
the reverse way to steel-i.e., the gasket is supported on a
flat metal surface, heated to redness, and then dropped,
edge-on, into cold water, from which it emerges in a deadsoft condition.
In cases where long bolts extending from head to crankcase are used, and particularly if the he~d and b~rrel a:e of
light alloy, the bolt tension with the engme hot IS consIderably greater than when cold,. possibly to such an extent that
unsuspected distortion may take place, or the bolts may
actually yield and become permanently elongated. These
conditions should be suspected (a) if the engine does n~t
hold its compression as well when hot as when cold; (b) If
the head-bolts frequently have to be re-tightened.
c.
If the engine on which you are working is known to
possess tbese faults, it may be as well to make up a new set
of bolts in high-expansion austenitic steel ~uch ~s K.E. 965
or Jessop's G2-which conform to SpecliicatlOn D.T.D.
49B-or the" 18/8" grade of stainless steel to Specification
D.T.D. 176A.
.
Each of these steels has a tensile strength of 37 tons or
more but all are a little difficult to machine due to their
work~hardening propensity. A tip to remem~er ~hen
\Working on them is to use keen-edged tools With 8 to
138
139
TUNING FOR SPEED
On some engines it is not easy to handle the rings as they

enter the barrel, due perhaps to the presence of a forest of
studs and none of the conventional ring compressors are
much use either. The simplest and handiest tool is a strip
of thin sheet metal, about 1t in. wide, bent to the circumference of the piston and pulled up wit~ a small ?olt t~rough
the ends. The top-edge is lightly cnmped WIth pliers to
prevent jamming in the chamfer of th~ barrel, and after the
piston has entered half-way, the s.tnp c~n be pulled off
. simply by removing the bolt. A.ll pIston rmgs, and slotted
oil rings in particular, are fragile, but as one cannot tell
whether they are broken or not once the barrel has gone
on, the operation cannot be carried out too carefully. Afte~
wards rotate the engine a few times to spread the all
evenly and wipe excess from the piston crown.
Next, after making sure the contacting faces of ?ead ~nd
barrel are clean, bolt the head down by carefully tlghtClll~g
each bolt a fraction of a tum in rotation in order to obtam
an equal amount of tension on each.
Finally, check your compression rati? exactly and enter
it up, together with any relevant detaIls suc~ as type of
piston, number of packing plates and so forth, In your notecompleted.
book , and one more stage will have been
.
140
CHAPTER XI

of any air of finality about the reassembly
work comprising the last part of Chapter X, if cams
giving a modified timing are to be used it may pay to
make a trial assembly with only a single spring on each
valve, or to put in only one rocker at a time when
checking the opening and closing points. Otherwise
it may be difficult to measure the top overlap with exactitude, because near t.d.c. the exhaust valve, as it closes,
tries to tum the camshaft forwards, thus taking up all the
backlash in the drive in the wrong direction.
The best sequence of operations for any particular engine
depends entirely upon the design, but in any case the points
to watch closely in connection with the rocker gear are
similar. In addition to operating correctly in relation to the
valve stems, all the moving components must be as friction
free as possible and all sources oflost motion, such as flexure
in the rockers or rocker supports, must be ~liminated.
Friction in valve gear comes into a different category
from that of friction in the rest of the engine. The latter,
while being very undesirable, simply subtracts from the
amount of power the engine is potentially capable of
developing at any speed. The former, even though seemingly
insignificant in amount, can have the much more serious
effect of limiting the maximum revs. by lowering the speed
at which the valve gear can be returned from fuHlift by the
springs, which have to overcome any friction present before
the valves commence to close.
At low speeds the inertia of the valve gear is small, and
there is plenty of excess spring-force available to overcome
friction, but as the speed rises towards the peak the excess
of spring-force over inertia becomes progressively less. The
IRRESPECTIVE
141
LIGHTENING THE VALVE MECHANJSM
TUNING FOR SPEED
Wear may be occasioned by crossloading on the

rocker mechanism
of semi - radialvalved Rudge and
Python engines.
ill-effect of friction thus becomes proportionately greater

and either or both valves cease to follow the cams on the
closing side at a much lower speed than one would think
without giving the matter some consideration.
Rockers and cam-followers must, therefore, be absolutely
free on their pins and have enough end-float to ensure that
they cannot possibly bind at running temperature. Occasionally spring washers are fitted to take up end-play and
eliminate a possible source of rattle, but for racing it is
better to replace these by distance pieces of the appropriate
thickness. Felt washers are sometimes used to retain oil,
and these can be an unsuspected cause of binding if they
are too thick or made of insufficiently resilient material.
Radial-valve Rudge and Python engines have an unusual
form of rocker layout which puts a certain amount of crossloading on the bearings of those rockers which act at rightangles to each other. After a while wear occurs on the
sides of the central bosses and the hardened washers against
142
which they bear. In addition to causing noise, this wear

diminishes the valve-lift and should be taken up by renewing
the washers and shortening the hardened central rollerrace until the side clearance is reduced to .001-.003 in.
The length of the needle rollers must also be reduced if
necessary so that there is no chance of their being nipped
endways when the rocker pins are fully tightened.
. Lightness is. a very desirable quality in valve gear, but it
IS fatal to achieve it at the expense of rigidity. Where the
latter is lacking at high speeds, when the loads are heaviest
the motion of the valves will be quite different from th;
A typical overhead
rocker with hexagonheaded adjuster and
lock nut in standard
form.
UNDERNEATH VIEW
AS SEEN IN DIRECTION
OF ARROW
the~retical figures; in fact, !he rockers in some racing

engmes are actually made heaVier than those in their touring
coun~erparts t~ ensure maintenance of power at high r.p.m.
In thIS connection one of the main advantages of the double
o.h. camshaft layout is the elimination of places between
cams and valves where flexure or lost motion might occur.
Br?adly speakin, the arms of a rocker are subjected to
bendmg loads which are heaviest close to the point of
support and fade out to zero at the points of contact at each
extremity. The central part of the Z-shaped rocker usually
fitted to push-rod engines is subjected mainly to torsion or
twisting of constant value from end to end.
The only safe places to remove metal are those at which
143
TUNING FOR SPEED

HEXAGON HEAD REMOVED AND
TWO FLATS ADDED
/
POLISH THESE SURFACES
.~
CONTACT FACE
SHAPED TO MATCH
TOP OF VALVE STEM
Judicious removal of
metal saves weight;
polishing minimizes
the risk of fatigue.
stress is clearly low or non-existent,paying most attention

to the ends of the arms, since these portions move at the
highest speeds. It is often possible to grind a sizeable amount
off the faces which contact the valves, or to reduce the
weight of the screwed adjusters sometimes fitted. Although
reducing the section of the arms is not to be recommended,
it is a good plan to polish the whole rocker in order to
minimize the chance of fatigue cracking.
Flexure may occur in the rocker mounting unless this is
bolted very firmly to the head. Anyone tuning one of the
older engines with the rockers held between sideplates
should verify that the attachment bolts are a really good fit
NEEDLE ROLLERS MUST
BE SHORTER THAN
CENTRAL SLEEVE
ROCKER
The U-shaped rockerbrackets of Rudge and

Python radial- valve
gear can seldom be
kept tight with the
An
standard bolt.
effective cure is to weld
them to the head.
in the ma~g ~oles. Some editions of the Python engine

were poor. III thiS respect, the cross-rockers being held in a
steel pressmg attached to the head by rather inaccessible setscrews which were both difficult to tighten and inclined to
work .loose. The sovereign remedy is to have the pressings
electncally arc-welded to the head, this process being preferable to acetylene welding because oftbe absence of distortion
risks.
Push rods are normally made either from Duralumin or
high-tensile steel tube. It is important to check that they are'
absolutely straight (by rolling them along a surface plate)
and. also that they have not been badly bent and restraJghtened, because a rod which has been so treated is
~lways liable to go again. If spares are not obtainable,
lIght-alloy ~ube ~o spe~ification 4 T 4 is the correct grade,
the usual size bemg i-m. diameter by 16 S.W.G. This is a
han~y gauge because the bore is slightly under t in. and
can Just be reamed out to that size to take the end-fittings.
Wh~ther these a;e fitted internally or externally, they must
be .tlght, ?therwlse the end.s of the tube will hammer away
qUIte r~pldly. If the tube IS steel instead of light alloy the
end-fittmgs should be securely sweated in place with tinman's solder, completely penetrating the whole joint.
In. some designs the push rods are of considerable length,
and It n:ay pay to experiment with tubes of larger diameter
o~ heaVier gauge; unsuspected buckling may take place at
high speeds, and this will have the same bad effect as
flexure in the rockers.
Tappets or push-rods ~re occasionally on the heavy side,
due to the presence of adjusters. If you are well acquainted
with the engine's habits when cold and hot with regard to
~orrect valve clearances, something is to be gained by makmg up a set of non-adjustable components, finally setting
the clear~nces either by adjustments to the push-rod length
or by usmg hardened caps of the appropriate thickness
fitted to the ends of the valve stems. These components are
very often fitted as standard to minimize stem wear but for
making them up individually K.E. 805 steel harden~d in oil
145
TUNING FOR SPEED
at 830 0 C. and tempered by quenching in oil from 550'" C.

cannot be bettered.
An additional advantage of using fixed tappets for long
distance racing is that there is no possibility of the adjusters
coming loose when there is only a lap or two to do before
the finish! The scheme is very easy to apply to sidevalve
engines with their fiattopped tappets; the method is to use
the existing adjustable pair, set to the correct length, as
. gauges from which to obtain exactly the length of the non
adjustable pair, which can be made up out of silver steel,
hardened on the ends by quenching in water from a bright
red heat.
Sometimes shims of the correct thickness for placing under
the valve stem cap are not available when making adjust
ments to o.h. camshaft engines which employ this system.
Satisfactory substitutes can, however, be made by cutting
discs from a set of feeler gauges, this material being suffi
ciently hard to withstand the rigorous duty entailed.
(1) Pinion and crown wheel ill pel!ecl mesh, with slight back/ash.
(2) If excessive backlash, raise pillion. If no backlash, move

crown wheel to left. (3) If too much backlash, move crown
wheel to right. I{lIo backlash, lower pillion.
Even in o.h. camshaft engines, motion can be lost in the

valve gear unless all is as it should be. If bevels are used
they must be in rcorrect mesh, which means the minimum
of backlash consistent with absence of whine. Fortunately,
neither these gears nor their bearings wear very rapidly,
but if the adjustment is faulty it may not be possible to
rectify it by moving only one of a mating pair of gears.
As an approximate guide, the inner ends of the teeth on each
pair of gears should be just level with each other as they
pass through the point of intersection.
If this is not the case, one or both gears must be moved in
the appropriate direction, either by adding shims or skim
ming a small amount off the face of the vertical shaft bush
housing, or the bush itself. Since the camshaft, and hence its
crown wheel, is usually located by a ballrace some distance
from the gears, thermal expansion has the effect of increasing
the backlash by a small amount, and it will often be found
that a slight whine, noticeable when the engine is cold, will

cease as running temperature is attained.
Another point to notice is the fit of the tongues of the
Oldham couplings in the bevel-shaft slots; they should
be just free enough to slide, but have no trace of clearance
sideways. If any shake has developed, new couplings are
required, or, alternatively, the originals can be built up
with hard nickel plating as in the" Fescol " .process.
Conversely, a small amount of endclearance must be
present in the vertical shaft assembly. If this is not provided. severe thrust load will be applied to the bevel shafts
when the head is bolted down, and this point must be closely
watched if the barrel has been shortened appreciably. If
no clearance exists it is possible to gain a little by grinding a
small amount off the ends of the bevel shafts and vertical
shaft. This process, however, must not be overdone or the
area of contact of the coupling tongues will be seriously
reduced. If at a future date the barrel is packed up again on
compression plates, it will be necessary to obtain either a
longer shaft or some couplings with thicker centres, otherwise the uppermost tongue will not run fully in engagement
with the top bevel-slot. This condition is likely to lead to
failure as the metal in the tongue will be subjected to a
146
147
Camshaft Drives
TUNING FOR SPEED
TOP BEVEL SHAFT
'HACC
\
WHEN <''''ANe<
HEAD IS
BOLTED DOWN
8.
The essential'
small clearance above
the.J. upper universal
joint. (Right) A common fault which can
be. cured by a longer,::;~
shaft or a thicker
universal joint as
shown in the centre.
(Left)
TONGUE NOT
fULLY ENGAGED
G) C
.
destructiv~ combination of shear and bending loads instead

of pure shear as intended by the designer.
There is a small opportunity for saving weight in the
vertical tube and its retaining ring-nuts by making these in
light alloy, using Duralumin or R.R. 56 for the tube and
any reasonably strong aUoy, such as L 1 or D.T.D. 423, for
the nuts and possibly the bevel-shaft housings.
Attention to Cam Followers

Coming back to push-rod and S.v. motors, freeness of
the cam followers is important. The procedure with lever
followers is similar to that for the overhead rockers; in
addition, if any of the rocker bearings or pins show patches
which are unduly rubbed or worn, indicating lack of
lubrication, an additional oil hole or a nick ground into the
end face of the bush may improve matters.
A little weight may be saved by judicious reduction here
and there, but it is a mistake to remove too much from the
region of the follower-foot, as this may lead to local overheating with consequent softening and rapid wear. Depending on the shape of the cam and the foot, there is a certain
minimum area over which rubbing contact takes place, and
it is an advantage for the foot to be appreciably longer than
this, as the excess portion acts as a lead-in for oil and thus
improves lubrication at a vital spot.
148
Fol;owe;s.which are only slightly :vorn can be reshaped

to their ongmal curvature by hand With an oil-stone, but if
,,;,ear to a depth.~eater than .030 in. is present, there is very
httle of the ongmal hard case left and new rockers are
required-unless the originals can be re-casehardened and
ground, a process which some service departments are
willing to carry out.
If it is desired to embark on the project of making a
new set of followers, 3i% case-hardening nickel steel is
an excellent material; the hardening should be .045 in.
to .055 in. deep on the contact faces, but the remainder of
the surface should be kept soft by copper-plating before
h~rdening.. An even better j.ob c~n be made by using 5%
atr-hardenmg steel to SpecificatlOn S 28 with a coating
of "Stellite" welded to the contact faces.
Following the application of the "Stellite" the component is heat-treated by cooling in air from 830 0 c., after
:-;hich. t~~ . faces are finally ground to shape. Welded
Stelhte IS not very hard but will withstand very arduous
treatment without appreciable wear as, unlike case-hardened
steel, it does not soften under the influence of heat
.. Stellited" faces must be finished by polishing to a dead~
smooth surface, otherwise the cams will suffer undue wear.
Points to Bear in Mind
When reconditioning or making new followers (or o.h.
with curved feet two things must be
borne m mmd. One is that the radius' of curvature does
not affect the timing at aU, but that it does affect the rate of
acceleration; in other words, if you substitute a radius of
, in. for one ?f i in. t.he opening and closing points measured
m degrees wlll remam the same, but the initial acceleration
will be gn~ater with tbe larger radius. The second point is
that the dIstance between the rocker pin centre and the line
?f contact with the. cam does affect the timing. A difference
m length of .004 m. makes one degree difference with a
cam base circle of 1 in. diameter; obviously, therefore, the
camsh~ft r~ckers)
149

TUNING FOR SPEED
~resence of fiat due to wear at this point will lengthen

"he period of opening in proportion to its width. This
matter will be gone into in greater detail later, when
'Valve timing comes up for discussion.
Varying Types of Follower
A few designs employ lever followers with fiat feet and
-with these the rate of wear is usually lower owing to the
,greater area over which the rubbing action is spread. Eventu.ally grooves will appear near the middle of the faces
"but except in extreme cases the parts can be reconditioned
-simply by smoothing on a fiat oil-stone, taking care
1:hat the faces are maintained parallel with the rocker-pin
:ilioles.
)
The only other common variety of lever follower is fitted
-with a roller in place of a solid foot, and as the bearing in the
Jroller is necessarily very small, heavily loaded and not too
-well lubricated, wear develops after a period. Replacing the
pins 'usually effects a cure, but while the parts are dis. mantled it is advisable to check the rotler holes for
.ovality. If they are out of round they can be lapped out
, with a mild steel bar and emery paste, after which oversize
pins will be required of a size. giving .002-in. clearance in the
"lapped-out holes.
The holes in the follower itself must be enlarged to a light
. drive fit on the new pins, which can be made from silver
,steel hardened and tempered to a straw colour in the centre
but blued at the ends to enable them to be riveted over after
_assembly. Some racing J.A.P. engines employed rollers
running on needle-roller bearings which were quite an
. improvement over the plain variety.
.
If some serious defect, such as extensive fiats or pitting of
_the circumference of the rollers, is in evidence it is best to
replace them. If spares ara unobtainable the 1% carbon-l%
. chrome steel known as ballrace steel (Spec. EN 31) is about
,. the best material from which to make new ones; heat-treat
0
'; by quenching in oil after heating to 820 C. for one hour,
150
'L
t~en
temper at 200 0 C. for half an hour and allow to cool in
aIr.
. "Dire~t impact" tappets bearing on the cams without the
mterventI?n of followers are used in many s.v. and a few
o.h.v. deSIgns. Th~y are usually of the mushroom type offset
fro~ the cams ~ h~tle .so t?at a slow rotation is continually
takmg place, thIS dlstnbutmg wear over the entire surface of
the ~ead, which in consequence remains flat almost indefimtely. Non-rotatable patterns wear, of course. in just
the sa~~ ,~ay as lever followers, and require the same
recondltIOnmg treatment.
Working loads on the stems and tappet guides are heaviest
at the ends an~ become zero in the central portion; thus
a ~m~Il reductIOn ?f weight. can be effected by heavily
relIevmg the tappet In that region. This will also reduce the
ef!'ort necessary to shear the oil film between the surfaces and
WIll have no adverse effect except, perhaps, to increase oil
leakag~ from the top of the guides, which will not matter
much 10 competition work.
Watch the Exhaust Lifter
The exhaust lifter is not required for racing. In the case of
A.J.S. and Matchless the control lever in the
tlmmg case can be disposed of. and after extracting the guide
and tappet from the crankcase, the valve-lifted collar can be
removed through the slot cut in the side of the guide; a small
s~a~dard
151
TUNING FOR SPEED
point, but one which illustrates further how the odd fraction
of an ounce of unwanted weight can be saved.
In any case, it is wise to verify that the exhaust lifter is not
holding the follower off the base-circle or fouling the cam
nose if any change on cam form or type is made. As an
instance .. Lightning" cams when fitted to a Vincent
" Rapide" or" Shadow", foul the exhaust lifter levers: the
latter .must therefore be filed away to clear the cam-nose,
though it is probably better for most racing to remove the
entire lifter mechanism and block up the hole through which
the lifter rod passes with a tin. B.S.P. plug, the same as that
used in the crankcase drain hole.
I
CHAPTER XII

THIS and the next chapter are closely related and may well
be studied in a single reading. My aim is first to deal with
standard valve settings and, in Chapter XIII, to indicate
adjustments which may result in improved performance.
Newcomers to speed work would be well advised to stick
to the maker's timing for a start, and it is also a wise plan
for more experienced riders to follow the same course if
major modifications have been made elsewhere to the engine.
As a general rule, it is a bad policy in experimental work to
alter more than one thing at a time, otherwise it is difficult to
assess the value of any single modification.
Engines usually have their timing gears clearly marked by
dots on the mating teeth or lines inscribed on the rims of the
gears. When taking notes on the engine before commencing
the initial dismantling, a diagram of these marks should have
been included, together with any information relating to the
position of the small timing pinion on the mainshaft since it
'
IS a. common practIce to use a number of keyways, anyone of
which may be selected, in order to provide a fine variation of
timing at this point. If this is the case, it is usual to mark the
"standard" keyway in some manner, though this may not
be visible when the pinion nut is in place.
By no means do all engines employ a keyed-on pinion. In
Rudge and Python power-units, for example, the pinion fits
on a plain taper, and thus can be set readily in any position
desired.
. .Marks or no marks, before finally fitting the timing cover
It IS best to play safe and to check the setting with a timing
disc and pointer set to zero with the piston at top centre. It is
not too easy to detect exactly where this occurs to within a
degree or two just by feeling the movement of the piston
F
153
. .
152
TUNING FOR SPEED
An excellent timillg stick for ellgilles

fitted with a COlltrol plug call be made
up from all old plug body, a steel rule
marked ill. illches or millimetres, a
piece of rod, a bicycle brake draw
bolt alld a simple poillter.
practice and specify closer clearances for setting than for

running, giving the appearance of greatly increased overlap
compared to some other makes, whilst a special system is
adopted on Yincents because the design is such that it is
impossible to measure the clearance accurately. This does
not matter for normal adjustment, because then the cle,arances are simply set to zero when the motor is cold. For
timing, the figures are taken with the valves lifted .005 in.
as measured by a dial gauge in contact with the valve.
crown through the plug hole, but it

can be determined quite accurately
by means of a depth gauge of such
a length that it contacts the crown a
few degrees-the precise number is
immaterial-before top centre. If the
pointer is correctly set to zero, the
readings at which the crown touches
the gauge on the upstroke or downstroke on each side of top centre
will be equal. Another method
is to apply a film of thick soap
solution over the plug-hole, or better
still over a small hole in a dummy
plug body; the point at which the
film balloons out to its maximum
extent as the piston is rocked over
centre will clearly be t.d.c.
With this important matter settled,
the timing can be determined and
compared with the maker's figures, if
this knowledge is available, not forgetting that many makers
specify valve clearances for checking which are wider than
those at which the tappets are eventually set for running.
This is done because with some cam contours the rate of
lift over the first few degrees is so gradual that it is very
difficult to determine the precise instant at which the valves
open or close. As was pointed out in Chapter I, this must
be borne in mind when making comparisons between the
timing diagrams of various makes. Nortons reverse this
Whatever clearances are used, it is very rarely that the

figures obtained will agree absolutely with those quoted by
the maker, so that a little common sense has to be used to
obtain the closest approximation. Generally speaking, the
least illlPortant point is' the exhaust valve opening, and the
most important is the relationship between inlet opening
and exhaust closing with reference to t.d.c. Sometimes it
may be found that the timing is several degrees out on one or
more points at the specific clearance but the error is reduced
if the clearances are set a trifle wider; whilst not being a hardand-fast rule, this at least gives an indication that the error
is not so serious as it appeared and may even be disregarded.
It is often difficult to get an equal timing diagram for both
cylinders of a twin unless some allowances of this nature are
made, particularly if some wear has taken place in the cams
or cam followers. It is as well, while the timing disc is fitted,
to obtain some more figures for future reference, and as an
154
155
All alternative timlllg stick

for inclined plug apertures
is made by drillillg a
sparkillg plug outer body ill
this mallller.
TUNING FOR SPEED
Some simple form of handle to

screw on to or grip the mainshaft will prove invaluable when
carrying alit experiments with
the timing gear. If screwed on,
a lock 1IIIt will be required.
aid to getting a clearer picture of the valve-lift diagram,

without which any subsequent experimenting will be simply
like taking leaps in the dark.
The more obvious additional things to measure are the
amount each valve is open at top and bottom dead centres;
the actual amount of lift (which is not necessarily the same
as the height of the cam on account of leverage-ratios wh~ch
may exist in the rocker gear); the crankshaft ang!es ~t WhlC~
the valves reach full lift, and the dwell at full 11ft, If any IS
present-" dwell" representing the number of degrees of
movement during which the valve is on full lift.
The best way to measure lifts is by means ofa dial-gauge
rigged up in contact with the valve-spring collar. With this
simple apparatus a really painstaking individual can draw a
complete lift-curve for both valves by taking readings at
intervals of, say, two or five degrees. These results are plotted
on squared paper as vertical measurements, cranks~aft
degrees being measured horizontally. The graph so obtamed
can be made to give quite a lot of information, as will be
shown later on.
There has been such an enormous number of engines
manufactured in various types and editions that it is not
possible to quote all their timing figures. even if they were
available, but I have compiled a list of quite a few, not all
of which are modern. Included are a number of figures for
both sports and racing engines of the same make, the object
156
being to show how the timings vary according to the type of

performance required.
The checking and running clearances are also given-it
being understood that .. running" means "as set for
running" with the engine cold, although at operating
temperatures the clearances will probably be very different.
As a general rule when hot, the clearances of both valves in
push-rod engines tend to widen, though more on the inlet
than the exhaust. On camshaft engines both clearances close
up, though more on the exhaust than the inlet, and in sidevalve engines the inlet widens, whereas the exhaust closes up
-the latter point in particular must be watched carefully if .
a K.E.965 valve has been fitted in place of one of silchrome
or some other steel with a much lower expansion coefficient.
It is best when measuring valve lifts, to have the clearances set to the "running" figures, otherwise serious
errors will be introduced over the early portions of the liftcurve when the valves are rising only a few thou. per degree.
Each valve should be measured separately, with the other
one rendered inoperative by leaving out its pushrod or
rocker so that the backlash in the drive is not taken up in the
reverse direction by the cam not under consideration.
The rockers on MatchlessjA.J.S. twins and KSSjKTT
Velocette units with fully enclosed valves are mounted on
eccentric spindles, as a means of adjusting the clearances, and
either rocker can be removed, even if the valves are fully
assembled, simply by unscrewing its spindle, the rocker in
A dial gauge, contacting

with the valve-spring and
cap, read in conju1lction
with the timing disc, will
enable the tuner to prepare
an accurate valve-opening
diagram.
157
TUNING FOR SPEED

INLET
EXHAUST
some cases the inlet angle is a few degrees less than the
exhaust. The reason for this is that better low-speed
perform~ce and fuel economy are thereby obtained, and, of
course, lU standard production machines these aspects,
although perhaps not so important as flat-out performance
have to be given a fair degree of attention.

Engines with "short" timings will respond to some
extent to the fitting of an open exhaust system and an inlet
pipe of the correct length, but they cannot be expected to
perform as well as those with racing timings in conjunction
INLET OPENS
'3S. BEFORE 9 m.m.
Bya series of hole-and-peg locations, Norton" Manx" cams

can have vernier setting ill five different positiolls.
the process partially lifting the valve at each revolution.

When finally setting the clearances, care must be taken to
see that the plane of eccentricity lies in the right direction,
otherwise the timing will be badly out and the rockers will
not contact the valves properly. Arrows are stamped on the
rocker spindle ends to indicate their correct positions, but
any doubts can be settled by checking that the rocker-ends
contact the valve steins centrally; if not, the error will be
quite noticeable and can be cured by moving the pins half a
turn in either direction.
A glance at my timing data shows that there is fairly
close agreement between the timings used by various makers
for fast main-road work or for racing, and an engine which
does not figure in the list is bound to perform reasonably well
if timed in a similar fashion. A closer study shows that, on
the whole, not only is the overlap on engines intended for
use with a silencer much less than on those made mainly or
wholly for racing, but it is also split up differently.
In the latter case, the crank angle at which the inlet valve
opens before t.d.c. is usually several degrees greater than
the angle at which the exhaust valve closes, whereas in
sports engines the angles are more nearly equal; indeed, in
158
HolY to convert
degrees to linear
measurement. First
describe a circle the
diameter of which
equals the length of
piston stroke. With
a compass set to the
length of the connecting rod between
celltres, strike arcs
on the vertical line
from top and bottom points on the
Circle: Next strike
arcs from valve
opening alld closillg
points and measure
distances betweell
these arcs and top
and bottom-centre
arcs respectively.
T.O.C
I EXHAUST CLOSES
1 m.m. 30 AFTER T.O.C
" I\[
T.O.C
&0
I
88
1
-m~m
"l.m.
.,,
I
'H:
60 M.m ..
STROKE',
~
If/
-IEIi lJlI'-d+
'
I I; : ~. EXHAUST OPENS
INLET CLoses,
60 AFTER a.D.C '"
_l....
! \ : E -65'
, ,-1:
.1.
BEFORE B.O.C
B.O.C.
II '.':
~ ~
, ,I
II
/ I
iI
.i
I.
i!
/1
ji
ii
.~
~.
!l
~\\
'3
-.~.
""1.<)
'\...v:~
;:
.... -.
-Jo" . .
l I
"
,.1
, .
:' I
,I
; '.
~, \
/""
.r-
'>"'- <::----....-----::
~ ! -;,.//
"'V'
GOo ...... _":.~-:'~.""::.'
159
TUNING FOR SPEED
with the correct pipe lengths. Conver-sely, if maximum

performance with a road silencer fitted be the aim, there is
nothing to be gained-in fact, a good deal may be lost in
speed, economy and acceleration from low speeds-by
attempting to utilize cams of the T.T. variety. If it were
otherwise, manufacturers would not go to the trouble of
developing different cams for various types of engine.
In road-racing engine speeds can almost always be kept
high by proper use of a mUlti-speed gear-box; good torque
at low speeds is not essential though it isrvery desirable on
short circuits where time lost in gear-changing may become
too high a proportion of the actual running time. In
scrambles and trials in which a speed test is incorporated,
conditions are very different because one needs an engine
which is fast but can put up with large variations of speed
because of the fairly wide ratio gear-boxes which have to be
used. On speedways it is also essential to have an engine
with good pulling power from 3,000 r.p.m. upwards, and
for aU this sort of work overlaps are not generally greater
than 80", and may be much less for competition models
where flexibility is as important as sheer power.
Engines with separate inlet and exhaust cams, as for instance the Triumph twin and the double o.h.c. Norton, lend
themselves well to variations in the respective cam settings.
In fact the Norton is specifically made so that the inlet cam
can be set to the best position for the exhaust system in use
by means of a vernier device; as shown in the table, the inlet
cam is set to open 10 earlier (and close earlier by the same
amount) when a megaphone is fitted than it is with a straightthrough pipe.
Actually. the timing figures given do not convey any
real information as to the efficacy of the opening diagramwhich varies, of course, with the valve lift and the steepness
of the cam contours-but assuming a reasonable degree of
similarity in the height of lift in relation to the capacity of
the cylinder between various designs, they do indicate the
lines along which to proceed in the search for speed.
It is frequently possible to utilize cams from one type of
160
~ I--~~~ ---~.---------------------
i I g~:;,:g
.---~---.-----.----------.---
gg
~:;':!2!2~
g!2!2
..-- ex
~ g~g!~
~==i'--===~~-~~=~~:~
~~I
ot"'lt'-"d--
\C~\O"
00
I't-.
--+--------...
~I
Vlco\nOM
Vl\O\Ot"--\O
~.----
V)V)V\
,....\0\1)
r...
OV)f1")""'"N
QO(;()r-...~\O
oW')
V\\D
V)
'"
C1
:::_:::::
.
161
5l!2 !2
. . . . . . . . . ----
~.
........
.....
....". ..
~
SPI!Cl!dEN VIJ..V/I. SE'ITINaS
WTI'H Cot.iPLEMENTAIlY DATJ.-..cQn~d
_ _ _ _ _ _ _:--:_ _ _ _ _ _ _ _ _ 1 OPENS
10
J A.P. contd.
,....
1.100 o.ll.v. W/c 60 twin

8/75 h.p. touring
175 and 2SO D.h.v. Standard

and Racing..
..
1,000 o.h.v. and 1935/37 Racing

Standard
27
1.000 o.h.v. Racing (1932/34) ..

3SO. 500 and 8/75 o.h.v. Racing
500 Dirt Track ..
..
..
15
45
45
65
65
. 25
60
Levis
350 (high.lift cams)

500 (high-lift cams)
Guxz;
Albatross
Matchless
3SO G3L (early)

350 03L (late)
080
..
04S (9.5 C.R.)
Newlmp<!rlal
Norton
15
2.$
~NaS
II----t---I----t---f~--------I
65
65
65
65
2.$
2.$
.002
.002
.002
.002
.002
.002
.002
.002
38
67
67
27
.002
.002
.002
.002
4SO
66
65
63
70
70
2.3
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
.002
45
45
38'
3S
60
66
2.$
30
.002
.002
.002
.002
I"
35
.008
.006
.010
8Omm.
.016
.016
.016
.008
nil
nil
nil. nil
nil
nil
.004 .008
7/16 in.
7/16 in.
lin.
6()
78
30
0\
tv
~I ~usr
I~~I ~
00Br
MODEL
MAKl!
23
35
35
.. ,122mm.] 170.mm.11 170mm.1 80 IDJD.II .008
2.8
016
.016
.016
.004
45'
20
32
32
55
67
63
63
68
18
65
65
2SO and 3SO OrandJ'rix
2.8
62
60
30
nil
nil
nil
nil
14mm.
35O'al1<1'soil International
SOO Modd 18 and ES2 and

6'00 Modd 19..
..
350 Modds SO and S5
30M, 1M straight pipe
30M. 1M megaphone
Double o.h.c. 30 M. I M.
30M short stroke
1M short stroke
471
70
8S
42;
.004
.004
.010
.020
42io
2.$f30
30/35
43148
60165
nil
nil
42/47
48'/SO'
SS
8S
85
72
94
2.$/30
35/1
42i
42i
4S
64
70
nil
nil
10
.024
34.371"
34"
39"
47!
51-%
60
67t
6()
94
74
95
82
30
30
44
70
.007
.002
.002
.019
.002
.002
.028
.028
45
>-l
c::
z
52
o
"1
en
"d
I:I:!
I:I:!
t)
Millimetre dimensiona are aJ measured at rim of flywheel.
SPECIMEN VALVE SETTINOS WITII CoMPLMENTARY
EXHAusr
INLET
IdAK.t
Paotber
SOO and 6'00
Roc!p'
Ulster
3SO Replica.
500 Replica
25
Model 9() (Wolverhampton)
30
Triampb
250 and 3SO o.h.v. s:lngle

SOO o.h.v.ling1e..
Speed Twin and Tiger 100

Grand Prix
TIOO (Racing Kit)
36
Velocette
VIIIOOIt
KTT.. 1931 ."

KSS. (iron head)
KTT. Mark IV and V .

KTT. Mark VI
KSS. (alwninium head)

MOV., MAC. .
.
MSS. early
MSS.,late
10
70
70
E
75;
61
47
43
68
39
51
SO
30
Rap/de, Black, Shadow and

Comet
Ligbtning. Flash
60
52
55
34
50
42
S6
J>istanc:cs
RUNNING
OPENS I
69
51
65
47
60
60
60
60
71
75
64
70
70
60
.001
.002
.001
.002
Un.
.020
.008
.010
.020
.008
.010
nil
nU
nil
.003
.003
.003
ISmm.
16mm.
18mm.
30
.002
.012
.002
.012
44'
36
.001
.001
.001
.020
.001
.001
.001
.001
.001
.001
iio.
nil
.001
.001
.001
.020
nil
.002
.004
42'
.012
.012
.020
.020
.025
.010
.oIS
.02.$
.012
.020
.020
.020
.025
.01.5
.020
.025
.012
.012
.01S
.01S
.012
.003
.003
.OOS
.022
.020
.025
.025
.012
.006
.006
.010
oil
nil
nil
nil
30
~
32
SO
48
I
43
45
29
I
I
30
68
68
~ 00
IGNITION
AOVANCE
EX.
10mm. t3mm. 16mm.110mm.

9mm. 14.4mm. 14.4mm. 9 mm.
9mm. 14,1/,mm. 14.4mm. 9 mm.
Suohellm
Cl.EARANCIlS
CHECKING
MODEL
OPENS
.....
0\
DATA-continued
Valves lifted
.OOS off seats
stroke of piston
~"'in.
t.u.
:7"
42'
42
35
32,"
>t)
....
c::
til
>-l
e::I:I:!
Z
...,
til
...,
<:
>t"'
<:
I:'ll
10
otll
I'
I'
::0
38'
39"
38'
>-
TUNING FOR SPEED
engine in another type of ppwer unit from the same stable

and thus obtain either a longer timing or higher lifts, or
both. As an example, the "45, 65, 70, 35" cams used
in the 350 c.c. and upwards racing J.A.P.s can be fitted
to the 250 c.c. model. Again, some of the push-rod,
Velocette cams are interchangeable, although the helical
teeth gears, unique to this make, have been made with
two or three different angles of helix and may not, therefore, be interchangeable. However, the camwheels are a
press fit on the cams and can therefore be pressed off and
changed over; if each pair is marked before being separated
by scribing a line radially inwards from the marked tooth,
and then refitted with the marks in line, the timing should
be correct. Vincent camwheels are also pressed on, but
in this instance the holes are parallel and not tapered as in
the Ve1ocette. Slight variations in shaft or bore diameters
may lead to insufficient tightness if shafts and wheels
are interchanged; the interference fit must be at least
.001 in. requiring 1 to 1t tons fitting pressure. If less than
this the shaft should be nickel or chrome-plated up to
.the requisite size, otherwise the wheel may move on
the shaft when running and damage to the valve gear
will result. Rudge cams ate yet another example of interchangeability, in this instance for the 500 c.c. and 350 c.c.
capacities.
The flexibility and low-speed pulling of side-valve engines
is due in large measure to the "slow" timing usually
employed, but there is no reason why quite an appreciable
amount overlap should not be used. Here, again, it is often
possible to substitute the original cams by a pair from a
corresponding o.h.v. engine. Many years ago, I remember
fitting the cams from a " big-port" A.J.S. into a 350 sidevalve of this make with really remarkable results.
Owing to the comparatively poor breathing and lower
compression ratio inseparable from any side-valve design,
its volumetric efficiency is low at high revs, and it pays to
hold the inlet valve open a few degrees longer than on an
. o.h.v. unit, since it is no use closing this valve until true
164
compression actually begins. That does not occur, if the

cylinder filling is not too good, until the piston is well up
on the compression stroke. This subject of cam design
is very complex and will be dwelt upon at greater length
in the next chapter.
If the timing has been altered from standard, or there is
any reason to doubt the markings, set the engine by means
of the degree plate with the crankpin in the position corresponding to the "inlet opening" point. The tappet clearances having previously been set to the amounts specified
for use when checking the timing, the cams (or cam, if two
separate shafts are used) are rotated until the clearance is
just taken up, and the half-time pinion is then fitted to the
main shaft in the appropriate position by means which vary
according to the detail design. On J.A.P. engines there
a.re five k~~ays in the pin~on and the most favourable posi~lOn for It IS found by tnal and error, using each keyway
mtum.
(\ ~imilar sch~me is used on Vincent engines, the key in
thiS Instance bemg parallel so that it can be tapped into
place after the best position for the pinion has been found .
Normally the key is fitted in the keyway marked with a
punch-dot.
B.S.A. cams are listed, paired up by part numbers, with
the appropriate pinion to give the correct timing, but
pinions are also supplied which advance the whole timing
by 10.
I~ t~e ~esi~n. allows of but one position for the pinion,
vanatlOn m tImmg can be accomplished only in steps of one
tooth at a time, which is usually far too great. An exception,
however, is found in some Velocette push-rod models'
very fine-pitch helical teeth, of which there are 48 on th~
half-time pinion, giving a variation of 7t 0 per tooth.
One method of surmounting this difficulty is to make up
some stepped keys, remembering that at 1 in. radius 1
equals .017 in.; from this data the amount of "step"
to be filed on the key for any particular diameter of shaft
and angular variation required can be worked out by direct
165
0
TUNING FOR SPEED
proportion. Keys so filed must, of course, fit the keyways

in both shaft and gear closely, otherwise looseness may
eventually develop.
If the engine has separate camshaft for inlet and exhaust,
next turn the flywheels until the crankpin is at the" exhaust
opening" position and insert the exhaust camshaft so that
the valve is just about to lift (or at the nearest tooth to this
position), and slip the timing cover into place to steady the
outer ends of the camshafts-a precaution which is not
necessary if the shafts run in the modern manner on fixed
studs though it is advisable to fit the steady plate temporarily to eliminate spring.
.
The complete timing figures can now be checked. It will
probably be found that they do not come out exactly as
anticipated owing to small differences here and there in the
mechanism, or because an error has been made in positioning the half-time pinion. Usually a compromise has to be
effected, bearing in mind that the overlap points are more
importanf than the others. In this connection it is better,
in most instances, to have the inlet valve opening earlier
rather than later than intended.
Several attempts may have to be made to obtain the most
favourable figures, particularly if the cams have been
modified. The stepped-key dodge may be very helpful in
such cases. For future reference, make a careful record in
the note book of the timing eventually obtained, together
with any remarks, such as clearances used at the time of
checking, the valve lifts, and so forth.
Then oil all bearing surfaces copiously and fit the timing
cover, after making dead certain that everything is correctly
in place. At this juncture it is all too easy to omit perfectly
obvious things, such as distance pieces or thrust washers on
the rocker spindles, any rubber washers for the transfer of
lubricant, or perhaps the small spring-loaded jet which
feeds oil to the big-end of certain J.A.P. models, or even to
overlook the elementary precaution of fully tightening the
half-time pinion nut and locking it by whatever means are
provided in the design.
166
If you have gone to the trouble of dowelling the timing

cover in correct alignment, it can now be put in place with
the appropriate gasket or any good-quality jointing cement
thinly applied, and the screws or bolts finally tightened. If
not dowelled, the cover may have to be juggled about a
little to ensure that the cam spindles are not binding.
Incidentally, although brown paper is commonly used for
gaskets, a higher-grade material, such as "Oakenstrong,"
is to be preferred for high-class work, as it is less likely to
require renewal if the engine has to be worked upon when
far away from its rightful garage. If the joint faces are in
good condition, perfectly oil-tight joints can be made by
omitting the gasket and using only a thin application of
.. Gasket-goo."
.j
167
I
L
CHAPTER XIll
IMPROVED CYLINDER FILL1NG

reason for providing much longer periods of valve
opening on racing engines, as shown in the tables in
Chapter XII, is purely a matter of mechanics, for the loads
in the valve gear are diminished (or, alternatively, higher
speeds can be attained with the same stresses) if the angular
periods of opening and closing are increased. Provided that
the lift is kept the same, an increase in speed from 6,000
to 7,000 r.p.m. puts up the loads in the ratio of 36 to 49,
an increase of 36%, but if the angular periods are increased
in the same proportion-for example, by increasing the
duration of opening from 260 to 283-the loads in the
gear will be brought back to their original values and the
higher revs. can be attained with the same valve springs
with no potential loss of reliability, as might be incurred if
the spring strength were to- be increased from, say, 120 lb.
to 150 lb.
Lightening the valve gear, of course, reduces its inertia
and consequently the force required to operate it, but this
. can be done only to any marked extent in the design stage.
Apart from the minor savings which have been previously
described, a really desperate attempt to reduce the weight
of a well-designed mechanism by one-third would, in
addition to introducing a distinct risk of failure, almost
certainly defeat its own object by bringing in a much greater
amount of flexure under load.
Despite their seeming lightness, all gases possess quite
appreciable weight (as a matter of interest, the air in a room
10 ft. square and 10 ft. high weighs 80 lb.) and consequently
also possess inertia; that is to say, they resent being rapidly
accelerated, and when they are on the move are reluctant to
168
ONE
stop. The effects are similar to the inertia of the valve

gear, inasmuch as they become increasingly serious at high
speeds. Fortunately, the expedient of lengthening the valve~
opening period, used to reduce the valve gear stresses,
is also of the greatest value not only in overcoming the
adverse effects of gas inertia, but in turning it to good
account.
When the exhaust valve opens, the pressure in the cylinder
is quite high, somewhere in the region of80 lb. per sq. in. (the
exact figure depending, of course, upon the characteristics
of the engine), and the bulk of the exhaust products are
entering the pipe as a "slug" of gas at well over atmospheric
pressure. Towards the end of the exhaust stroke, i.e. as the
piston is slowing down, the column of exhaust gas tends,
under the influence of its own inertia, to continue travelling
along the pipe and to draw the gas still remaining inside
the cylinder out through the valve. Superimposed upon this
so-called extractor action is an even more valuable effect
. created initially by the "slug" of high-pressure exhaust
gas.
This "slug" forms the starting-point of a pressure-wave
which travels down the pipe at the speed of sound, until it
reaches the open end. Here it is reflected back up the pipe
but is reversed in the process, becoming a wave of low
pressure instead of high. This low-pressure wave travels
back to the valve and augments to a considerable extent
the extractor action already mentioned, provided that it
reaches the port at the right time, which is of course at or
near top dead centre. It is possible for the wave action to
create negative pressures of up to six pounds per square
inch in the port and almost all of the residual gas can be
evacuated from the cylinder head if the exhaust valve is still
open by a considerable amount at t.d.c. and does not close
until 45 or 50 after. Furthermore, the inlet valve can
commence to open long before t.d.c. without fear that the
spent gases will try to get out of the cylinder by flowing back
through the inlet tract: in fact if everything is correctly
proportioned the depression created in the exhaust port by
169
TUNING FOR SPEED
the combination of column inertia and wave action will even

cause a fresh charge to start moving into the cylinder before
the piston begins to move downwards on the induction
stroke. The inlet valve can be lifted t in., or more, off its
seat at t.d.c. and can attain fu1llift nearly as soon. as the
piston reaches its maximum speed, near mid-stroke, without
undue stress in the cam gear such as would be caused if the
valve had not commenced to lift so soon.
The speed of sound in exhaust gas-surroundings varies
with the temperature and pressure, but, for practical purposes, can be taken as 1,500 fto per sec., considerably faster
than its speed in air which is 1,100 ft. per sec. The time
taken for a wave to travel down any given pipe and
back again to the port is practically constant, so that
the effect of the wave action just described is most beneficial
'Only over a limited range of speeds. At other times the
effect will be absent or can even be adverse, when there is a
pulse of high pressure instead of the required low pressure
in the exhaust port at t.d.c. Such a pulse will cause a back-.
flow of gas out through the inlet valve and may upset
the carburation to such an extent that the engine cuts out
completely.
Altering:the natural frequ.ency of vibration of the exhaust
system by altering the pipe length, or diameter, or adding a
plain or lipped megaphone, will alter both the maximum
power range and the minimum usable speed. Shortening the
pipe and, to a lesser extent, reducing its diameter, raises
the frequency. Adding a megaphone accentuates the wave
effect considerably but tends to narrow down the usable
range of speed. High-revving, road-racing, or record-breaking engines, therefore, employ short pipes with megaphones,
but, where greater flexibility and a somewhat lower peak
power speed are required, a longer, straight pipe is better.
For extremely high r.p.m. the correct pipe length may be
too short to be practicable or it may not comply with regulations, and it is then necessary to use double the optimum
length.
The whole subject of wave formation in exhaust systems
170
is very complex and not a great deal has been writt~n about
it. For those who wish to delve into the matter further
Dr. ~chweitzer's book, The Porting of Two-stroke Engines:
prOVIdes some very useful information, and in Volume
XXXIV of the Proceedings of the Institution of Automobile
Engineers there is a very informative paper on exhaust
systems written by J. O. Morrison.
By mid-stroke the iqcoming charge is rushing through the
inlet pipe at several hundred feet per second, and it will
continue to do so until the piston has commenced to travel
upwards again on the compression stroke. If the valve and
inlet pipe areas are inadequate in size, or of poor aerodyna~c shape, t~e cylinder will not be completely filled by the
tIme b.d.c. IS reached, and holding the inlet valve open
for a further 60 gives a little more time for filling to
be completed. But, of course, as the piston has by then
risen some distance, the amount of new charge taken in
will be considerably less than the swept volume of the
cylinder-,-in other words, the volumetric efficiency will be
low.
On the other hand, if the valve area and pipe size are large
enough the cylinder should be very nearly filled by b.d.c.,
and after this the high-speed column of gas will continue
under the influence of its own inertia to travel into the
cylinder. It is possible by making full use of this "ramming"
effect to obtain a small, but very useful, amount of supercharge at certain speeds.
The distance from the valve seat to the outer end of the
air-intake of the carburetter has a large influence on this,
and on most racing machines is of the. order of 8 in. to
10 in. This is several inches longer than is usual on
or~inary o.h.v. ~nd s.v. engines-particularly the latter, in
which the whole Inlet tract is frequently very short-whereas
~eader~ with long memories will recollect the much greater
InductIOn length used on fast side-valvers such as the longstroke Sunbeam. The excellent results which used to be
obtained from fitting a Binks " mouse-trap" or " rat-trap ,.
carburetter were possibly attributable as much to their extra
171
TUNING FOR SPEED

0
a)
length as to the variable choke-tube which was the main

feature of these instruments.
This may seem somewhat of a digression from the subject
- of cylinder-filling, but it has to be mentioned because filling
and pipe design are to a large extent interdependent.
If all the factors involved are mutually in harmony a very
large amount of overlap can be employed, and the long
valve-opening periods so obtained will,. in turn, permit
higher lifts to be used without unduly stressing the valve
gear, thus leading the way to a further increase in volumetric
efficiency. The extent to which one may go is well illustrated by the 500 c.c. Norton timing diagram, originally
published in an I.A.E. paper read by Mr. J. Craig during the
war. The Norton diagram also shows very clearly the
higher lift, longer -duration of opening and larger port
diameter of the inlet valve in comparison with equivalent
figures for the exhaust. It is also noticeable that the exhaust
opening point is a little earlier and the inlet closing point
a few degrees later than usual, but the outstanding features
are the 115 overlap and the amount by which each valve is
away from its seat at t.d.c. Since that time, further research plus the ability to run at still higher r.p.m., gained
by using larger valves in "square" engines, have permitted
even longer timings to be employed. Reference to the valvetiming tabulation in the previous chapter shows that the
inlet valve on the 76 mm. by 76.7 mm. "Manx" Norton
model 40M, after allowing for clearance, opens at 75
before t.d.c., which is actually when the piston is just halfway up on the exhaust stroke, while the exhaust valve does
not close until the piston is nearly half-way down the inlet
.stroke. It would be impossible to get an engine to run at all
on such a timing unless full advantage were taken of
correctly proportioned inlet and exhaust systems.
Diagrams of this nature are most easily obtained without
loss of mechanical reliability on engines with valve gear of
the double overhead camshaft type-and at the other end
of the scale, in side-valve engines-where the best possible
combination of lightness and rigidity is attained by reducing
172
.L
,.---,
K'"
......1/
..... ~
....
~""
~,.
"
~~........
<~
tuo~
-' ~ .
z<x
-0<
1:
ci
\:l
1-1 I-
> r---I
v;
0
<
0
Z
..:
..:
~
VI
I--
<
\~
1:
.........
~
c-
o-
0
0
0
0
0
:il
-.:
<'j
~
":
__
<'j
....
...
j:::
0<
"-
.... VI
:r:
....
... <z""
0
;'
...
0
'!::. :
I-
VI
r--
:l ,< f- - I-.-<
- u.
:r: 0:::;
~ ~x
0<
~1: ~
.....
0..;~~~
(S3H:>NI 'Os) Y3't1Y 3AWA
173
U
0
a>
0
....
\.
.., ...
<'j
cu
I..;.
~
a)
.0
,00 ~ 1::1 00
i:i's:,~.5
6 ~3<<:.
:::'- s
~ cu~ "'"
.... :5:2J!
~~~"'"
",:!~'-
,.tJ}
~,
. . . 1'-.
~~~
9.'- 1:::
t:S .. ,~ q)
C>.~<::s9.
~<:.~
'-- ~
"",1;; <:>
a)
'~
8@
("oINM-
~~7
18
VI
...
:::
1::1,5
~.S~
TO
t::':. i"~
r......
'tl:::"a
-
....
t.....
0<
If-X
x
~~E'
U 0
0'"
"\
Vl
<.!!::
~o-'
~.o~
... ~'~
<:.
...
- .. 1::1
.... 1-
.-
1::1
~1::1 ..
oZ ...
3<
1'".
j:":~
>
:i
...-
. / V"
1:::
Q~~
./
..,
~:::~
~~ 3:
r...... ......
a)
5
~
00
........ ~
...
~-",
"
I/~;"
: _~ 's:,
'\~
...... 1:::
~
0
a>
~.,
o:J
.......
~Jl~
l;.f
" -........'"
~f-:t:
~
"...
V /
.5'!i ~
~~.
"'=:::9.
-.~~
~!1 ~
?\.~ C)
~~~
'- .. 1::1
.00
~~ :a
...:~~
-..,~
1...~-..
Oo.()
:?3 ~.s
"'E ..
.g~~
'~.g, ~
~~,~
,,
,
TUNING FOR. SPEED
the number of components between each valve and its cam

to a single tappet in direct compression. With such a layout
the actual motion of the valves follows very closely the motion
theoretically imparted to them by the cams, whereas in
designs where flexure exists, even to a limited extent, this is
not the case and the valve-motion departs more and more
from the theoretical as the speed rises.
Extremely good results can be obtained from push-rod
engines if this fact is recognized, and legislated for in the
cam design or timing, by making allowance for the loss in
motion which occurs most noticeably at the commencement
of opening of each valve. This can be done by (a) advancing
one or both cams by, say,S"; or (b) modifying the cam
contour to give a slow initial rate of lift commencing 10
or 15 before the actual point at which the valve is desired to
open. Scheme (a) has the merit of being simple to carry into
effect, and comparative tests with various cam settings are
easily conducted, but there are limits to what can be done
in this direction, since any alteration in the opening points
automatically changes the closing points by an equal
amount.
However, an engine with "30, 60; 60, 30" timing will
almost certainly perform better at high speed with an open
exhaust if the inlet can be advanced 5, making the timing .. 35, 55; 60, 30," particularly if the inlet porting has
been improved and the compression raised, for which
improved volumetric efficiency given by the two latter
factors the inlet valve' should not require to be held open
for so long as it was before such improvements were effected.
Scheme (b) is standard practice with some makers whose
~ngines are notable for their good all-round performance;
It has the effect of taking up all clearances, and part of the
unavoidable flexure, comparatively steadily and before the
valve has lifted off its seat by any appreciable amount. At
low speed the valve will leave its seat a little earlier than
it does at high speed, but the actual area of opening is so
small that there is scarcely any adverse effect on the lowspeed performance.
174
BLEND IN
BY OIL.STONING
NOTE THINNING OF HARD
CASE OVER HEAVILY.lOAOED
FLANKS
-INCREASE IN VALVE LIFT
(Above) A method of increasing simultaneously the valve lift

and period ofopening by grinding the cam base circle (Below)
To obtail' increase in lift whils! retaining the origindl opening
perIOd.
(; ....---r-- "6:s.
START OF tifT
..
~,..V"'.rlt
ORIGINAL CONTOUR/;::'\O
....y
'0.'
NEW
CONTOUR
100>""\t
END OF OROI'
ORIGINAL CONTOUR
NEW
10
CONTOUR
'"
f-=~+-'
AMOUNT GROUND OfF

SASE CIRCLE
-INCREASE IN VALVE LIFT
Cam design is, in fact, quite a cO.mplicated business.
A~art from determining the most suitable lifts and opening.
pomts. the contours of the cams and followers need to be

very carefully worked out in order to give the minimum
s~resses in th~ gear consistent with the type of valve-lift
diagram reqUll."ed; the most that the majority of private
owners can do is to experiment with different settings or to
obtain a set of racing cams.
. There.ar~, h0'Yever, on~ or tw~ simple methods of modifymg the tmung dIagram Wlth ordinary workshop equipment,.
175
TUNING FOR SPEED
although experiments of this nature must, of course, be

conducted solely at "owner's risk"!
One of these methods, which has as its objects increasing
both the lift and the angular period of the valve opening,
is to reduce the radius of the cam-base circle, a process which
needs to be done very accurately to achieve good results,
and requires the use either of a proper cylindrical grinder or
a lathe equipped with a tool-post grinding attachment. The
base circle is reduced a thou. or so at a time in successive
cuts, the cam being partially rotated by hand between the
centres of the grinder or lathe.
To avoid grinding too far in either direction, and thereby
ploughing a hollow into the flanks, it is advisable to determine beforehand-by rotating the cam in contact with a dialgauge-the points at which the cam-flanks join the base
circle, and to mark these iIi some manner which will not
subsequently become obliterated. Being dead-hard, it is
next to impossible to mark the surface with a scriber.
The best method is to make up a solution of copper sulphate (or " bluestone ") in water and apply it to the cam
surface. Instantly a very thin film of copper will make its
appearance, and scriber lines will show up on this surface
very clearly.
A carrier should be attached to the cam spindle and two
rigid stops rigged up so that the wheel cannot grind beyond
its allotted area-in fact, it is necessary to run out the cut
some distance before the' marks are reached, and subsequently to blend the new base circle into the existing flanks
by handwork with an oil-stone-an operation requiring no
little skill to carry out in such a manner that a gentle rise
is obtained without forming any suspicion of a flat or
hollow at the new junction of the flanks with the base
circle.
Many workshops are now equipped with tap-grinders
which form the lead-in portion of the tap with the correct
cutting clearance. These machines can be adapted without
much trouble to modify cams and can even be used to grind
on new cams completely from master-cams by fixing a disc,
which is the same size as the grinding-wheel, to the bed of

the machine and mounting the cam to be copied and the
rough cam-blank on a mandrel held in the swinging head
normally used to carry the tap when it is being sharpened.
The master-cam then rotates against the fixed disc, and
moves the head in and out as the spindle is rotated with the
blank in contact with the grinding wheel. This process
gradually forms the new cam, with a high degree of accuracy
if the work is not rushed.
A simple method of generating cams with longer timing is
to cut through an existing cam and rotate one half backwards
in relation to the other by the desired amount, perhaps 8,
then sweat or dowel the halves together with a thin piece
of steel between them. This filler piece is then filed to
make the contour and to provide a smooth radius at the nose
of the cam and the resulting article makes a very satisfactory
master for use in a conventional cam-grinder or in the converted tap-grinder just described. It is possible to make up a
whole family of cams in this way without much trouble, and
the scheme eliminates all calculation or guesswork in determining the correct lift curves.
The majority of cams are hardened to a d~pth exceeding
.040 in. As there is practically no load on the base circle
(unless the tappets are wrongly adjusted, or heavy push-rod
return springs are fitted) it is feasible to grind off as much
as .025 in. and still retain sufficient hardness to give satisfactory running, if not for an indefinite period at least for
long enough to prove the success, or otherwise, of the
experiment.
Unless the new contour blends correctly into the base
circle, valve clatter may be caused and, worse still, surprisingly heavy loads may be generated. The final shape should
be checked with a dial-gauge and a degree-plate attached to
the cam-spindle, and supposing .020 in. has been ground off
and it is desired to increase the total angle of the cam by
20 (10 on each flank), quiet operation will be obtained
if the .020-in. rise is split up in the manner suggested on the
following page.
176
177
TUNING FOR SPEED
Measuring the lift directly.off the cam with a dial-gauge

does not give necessarily the actual lift which will be imparted
to the valve, so this should finally be determined by assembling the valve gear and measuring the lift curve off the valve
itself. In passing, a good idea of how much flexure exists in
the system can be obtained by first plotting the lift curve with
a very light temporary spring, and then with the correct
springs; the result is sometimes rather surprising! When
running at high speed the difference will be greater still,
particularly at the start of lift, because of the valve inertia,
which is, roughly speaking, about equal at peak revs. to the
spring strength.
through destroying the accuracy of other dimensions,

although the latter might be restored by grinding after
being built up with hard chrome plating by the " Fescol .,
process or kindred methods.
It has already been mentioned that altering the camfollower radius affects the acceleration rate but not the
timing, and this offers a method of varying the opening
diagram without altering the timing or the lift. A curved
follower foot, even though it is non-rotating, is moved by the
cam exactly as if it were a portion of a roller which does
rotate. The lift imparted to the follower is determined by
plotting out tb,e path of the centre of curvature of the foot or
axis of the roller, as the case may be.
DIlGREES FROM
START OF LIFT
1
2
3
4
5
RISE OF CAM
IN INCHES
II
--II
"':-1---:-~-~-5
.005
.007
DEGREES FROM
START OF LIFT
RISE OF CAM
IN INCHES
1 I\--r-~-!--
9
10
.017
.020
For quiet' operation, the cam-lift modification

described above sho.uld be graduated.
If this modification is correctly performed, a greater lift

will have been achieved without the necessity for an increase
in spring-strength. Should the period already be of satisfactorily long duration but the lift somewhat inadequate,
the latter can be increased by itself if the base circle is ground
off right up to the marks denoting the original start and
finish of the flanks, after which a lot of stoning will be
necessary to avoid a sudden and undesirably rapid rate of
rise at the junction of flank and base circle. The final effect
will be to reduce the case-depth over the heavily loaded
flanks and as, in general, stronger springs will be required,
subsequently the running life will probably be short. In
some designs it is possible to overcome this defect by rehardening the cams, but this may bring about difficulties
178
(Right) Plotting the positions of an imaginary

roller moving around a
cam will determine the
path of the centre of curvature and, correspondingly, the lift imparted (see
sketch below) to the cam
follower.
\
\
;-_-I-~L--+\__j+-'-(~/-
'- -~ /
ROLLE'" OF EQUIVALENT
DIAMETE'" TO
fOLLOWER FOOT
I I
I:
if .:.~
'" 'j-.-/
''''-.
l ~
"
IMAGINA"'Y POSITIONS
OF ROLLER MOVING
LENGTH Of THESE
"'.
\. \./
- \ , /" .
~r
\.J
'-.1)
RADIAL LINES GIVE LIFT

AT INTERVALS OF S'
~PATH
OF CENTRE OF
CURVATURE OF fOOT
OR ROLLER
179
TUNING FOR SPEED
.Even if the follower is flat, the same statement applies,

as a flat surface can be considered as part of a cylinder with
an infinitely large radius. The lift curve can be depicted,
first by plotting out the rise of the follower-centre as it
traverses the cam, and then, in effect, straightening out the
curved line so obtained; this has been done in the accompanying graphs (see page 181), which show the effect of
different follower radii. While the rate of rise or fall is
greater with the larger follower at the start of lift and finish
of the drop (which increases the stresses in the gear at those
times), the accelerations and decelerations near the crest
of the cam are less and higher revs. can be attained with
the same strength of spring. The main gain, however, is to
be expected from the increased valve-lift at the vital period
extending for a few degrees before and after t.d.c.
Followers can be ground to shape by hand on an emerywheel, using a sheet-steel template previously made to the
desired curvature as a gauge, but care must be taken to see
that the surface is at all points parallel to the axis of the
follower-pivot, otherwise highly concentrated loads and
rapid wear will ensue. A good method is to make up a
simple swivelling jig, which can be clamped to the machinetable and fed in to the required depth. If more than .020
in. is taken off, the follower will definitely require to be recasehardened.
A factor which must be allowed for when lever followers
are used is the effect of their angular swing; if the cam is
symmetrical, the effect of the swing will be to make the liftcurve unsymmetrical and the flatter the follower the greater
lA~GE
\
"j
/ ' T-I'{ADIUS J
---r--~-tJ
1 /1
/
180
Varying the radius

of a cam-follower
base circle serves
to alter the acceleration rate but not
the overall valve
timing.
'-rT
'/16"
) ....
I.J
1\
'/u"
V
/
so'
I,;,~.
1
~o'
1\
1 .1
1"'""1/
65> 60""
r---r-,.
I~l
r-r-r-
30>
10'
10'
'\.
r"-r---"
I
10'
20"
30'
~'
50'
60' 65'
Above is plotted the lift curve of a cam operating the valve

via afollower with ai-ill. radiusfoot. Below is the lift curve
provided by the same cam operating via a follower with a
i-in. mdillS foot, the shaded area representing the increase
in valve opening.
I
t-H
~~.
I
I ~~~,.
65' 60
~~
l..r
Ii
I~~
~j..o'
IJ
SO
1 ', .. -
Ll1l~
I~~
j..I
I~~
"tnT
)0
10
'/,,"
bfilrV'
~o
10
ot
'0
10
OF CAM
r~~~
20
30
~o
SO
60 65'
this effect will be. The distortion of the lift curve varies
according to whether the follower is leading or trailing the
cam, and occasionally the effect is turned to good use to
obtain a rapid lift and slow drop (or vice-versa) with a
symmetrical cam contour.
Ir: this volume it is possible to touch on the fringe of the
subject only. It cannot be emphasized too strongly that
cam .m~dification is a thing not lightly to be undertaken,
and IS ltkely to do more harm than good unless carried out
with extreme care and with the aid of good equipment.
181
IGNITION TOPICS
CHAPTER XIV
IGNITION TOPICS
IF there is any doubt about its internal condition, the
best plan is to get the magneto overhauled by the makers
or a reputable specialist in the art. Symptoms such as heavy
sparking at the points, a noticeable grey deposit in the region
of the points which this malady causes, or slackness or tightness of the armature spindle would justify such a course of
action, since they are indicative of actual or impending
troubles which are beyond the ability of most amateurs to
rectify. When having a magneto tested, particularly if it
appears to be correct but the engine misfires or just refuses
to run up to speed after a period of running, get the test
conducted after the instrument has been heated up to
around 100C. in a stove. This sometimes shows up a
fault which was not obvious before.
Dead-certain starting, even with fully retarded ignition, is
absolutely essential. The magneto must, therefore, give a
good spark even when just flicked over with the fingers. If it
does not do so, although otherwise in good condition as
regards the points, brushes and slip ring, observe whether
there is an appreciable magnetic resistance to rotation just
prior to the points opening. If this " armature pull" is weak,
it indicates, in conjunction with the poor spark. that remagnetizing is necessary. This is not a usual malady with
modern instruments having cast-in magnets of the" Alnico"
type, but it does occur in older patterns and also in flywheel
magnetos.
'.
While not wishing in the slightest degree to cast any
aspersions on them, combined magneto-dynamo instruments
are intended primarily for touring purposes. For racing it is
best to discard these devices and obtain one of the several
types of magneto which have been built for the job. The ideal
182
is, of course, a modern B.T-H. or Lucas instrument, but

failing these one of the old M;L. models will give excellent
results if properly reconditioned.
Given everything in order in the sparks department, the
magneto can be attached to the motor and, if driven by
chain, adjusted so that the chain has! in. up-and-down play
in the middle of its run. Turn the engine over several times,
feeling the chain meanwhile to check if its tension varies.
This will be the case if the chain is worn unevenly, or the
sprockets are eccentric, or one or both shafts are not dead
true. Uneven tension in the magneto drive means that the
ignition timing will be perpetually varied over a small range;
thus the cause should be discovered and eradicated.
If the magneto holding-down bolts pass through slots,
adjustment is effected quite easily, but in other cases (mainly
on o.h.c. engines) where there is no such provision the base
will have to be packed up on steel or brass shims of the same
area as the base and suitably driUed. Owing to the high
expansion of aluminium, the chain will usually tighten a
little as the engine warms up, and some allowance must be
made for this; besides wasting power, an overtight chain will
ruin the armature bearing in a very short time.
Frequently there is little or no provision for preventing
grit entering the slot provided in the chain case to allow for
adjustment. This omission should be rectified by cutting out
a piece of felt large enough to cover the slot and a little
thicker than the gap between the case and magneto. The
felt is slipped over the shaft before assembling the magneto,
and is of particular value if racing on sand or dirt is contemplated.
Gear-drive magnetos invariably have some form of oilretaining device fitted behind the pinion. This should be
examined and, if necessary, components such as felt washers,
which have a limited life, should be renewed. There is
a number of 100% efficient proprietary oil-seals on the
market to-day.' They cause only a negligible amount of
friction, and in obstinate cases of oil leakage it may be worth
while looking into the matter of fitting one, although this
183
IGNITION TOPICS
TUNING FOR SPEED
will, in all probability, entail some additional machining

operations. If a gear-driven magneto is found to be full of
oil, probably the seal surrounding the mainshaft has become ineffective, so that cleaning out will be merely a
temporary measure. The seal should be renewed; this c.an
be done externally in B.T-H magnetos, but Lucas lQstruments must be first dismantled. In this event, make
sure that the new seal is of the type which has a garter
spring to contract it on the shaft, and not one of the early
pattern which does not possess this feature.
There is little else to check in a gear drive unless the wheel
is made of some non-metallic material such as " Tufnol " or
"Fabroil." In such a case, if any teeth appear to be damaged
it is advisable to obtain a new wheel, as if one tooth fails
completely the rest will follow suit very shortly. Owing to
the high expansion rate of these materials the teeth should
have a perceptible amount of backlash, even when new;
also the tips of the teeth must be well clear of the roots in
the other gear, otherwise an annoying whine, which .will not
cure itself as time goes on, will be set up when runnmg.
Racing magnetos are commonly equipped with manual
advance mechanism, but on many sports machines fixed
magnetos are used in conjunction with a centrifugal A.T.D.
These are quite satisfactory, but for racing the schem,e ~as
two disadvantages: one is that the accuracy of the tImmg
cannot be checked when running; the other is that it is not
possible to drive a rev-counter direct from the magneto
nut, which is one of the accepted methods, because the
power absorbed by the drive, though slight, might prev~nt
full advance being obtained. There are methods by WhiCh
the drive can be taken off the fixed stops on the pinion.
allowing the advance mechanism to work freely, but this
is rather a matter for personal ingenuity. Some riders
prefer to lock-up the device by welding together the fixed
and moving stops after taking out the bob-weights, and
either convert the magneto to hand advance, or use fixed
ignition.
Engines running on alcohol, and not requiring more than
184
35" advance, will usuaUy start quite well with no method of

retarding but, if petrol is used, or much more than 35 is
required, the engine may be prone to kick back, and sO
some form of retard control may be essential.
With the adjustment of the drive correct, and the contactbreaker points set to the correct gap (.012 in. in the case of
racing instruments), the ignition can be timed in the usual
manner, making sure that the cam-ring is in the fully
advanced position and that the points are just breaking, with
all ,the slack taken up in the drive. With automatic advance,
it IS also necessary to wedge the device into the fullyadvanced position, as this is more accurate than setting at
full retard and trusting that the extra advance will be as
stated by the maker!
The metlfod of using a fiashlamp battery and bulb to
determine the exact point of breaking is very good, but the
time-honoured scheme of inserting a cigarette paper between
the points and noting when this just comes free has the
advantage of being available in almost any circumstances.
Whichever method is used should be adhered to regularlymixing up two schemes is bound to bring in discrepancies.
When one finally tightens the magneto drive nut,the spindle
frequently creeps on a little way and thus retards the timing
a degree or so; it is therefore essential to check the timing,
and, if necessary, to reset it, if some particular" spot on"
advance is being aimed at. The vernier device used on A.J.S.
The flashlamp
method of
checking ignition-timing
referred to.
NOle that the
centre screw
must either be
removed or ;nsu[atedjromthe
block holding
the fixed point.
G
185
TUNING FOR SPEED
A great aid towards obtaining op~imum .timing of ignitio~1

and valves is provided by a vermer adjustment. Here IS
the simple but excellent system applied to the magneto
sprockets of certain early A.J.S. models, also to the camshaft drive.
machines does not suffer from this trouble and provides a

very simple method of setting or altering the timing.
On twin-cylinder engines the timing after bei~g set o~ one
cylinder must be checked on the other one. It WIll very hkely
be found to be a degree or two different owing to small
variations or errors in the contact breaker or cam. If the
error is only slight, the easiest way is to split the difference
between the two cylinders; if more serious, careful grinding
or oil-stoning of the cam rin.g will effect a cure. On singlecarburetter V-twins the rear cylinder filling is rarely so good
as that of the front one; consequently it will stand a little
more advance and thus a difference between the two timings,
provided it is not more than a degree or so, ~11 n.ot be a
disadvantage. This does not apply to V-twlllS wIth two
carburetters, which should take the same advance on both
"lungs." Vincents, prepared for fast road work, run
rather more sweetly low down if any difference is arranged,
give the front cylinder less advance than the rear, rather
than vice-versa.
In Chapter XII various makers'. timings are qu~te~.
These should provide a useful gUIde so long as It IS
appreciated that in many cases they are intended ror use ~ith
comparatively slow engines. Broadly speakin~, rac~ng
engines require less, not more, advance than theIr tounng
counterparts for the following reasons.
186
IGNITION TOPICS
Rapid combustion is aided by high compression-pressure,

a clean mixture (the result of good scavenging) and a high
rate of turbulence within the cylinder. All these factors are
present to a greater extent in racing power units than in
touring engines. Matters of detail design, such as the shape
of the combustion chamber, the location of the plug and the
angle of inclination of the inlet port, exert ap. effect and
account for the big differences in timing found in engines
which, to a casual glance, appear to be fairly similar.
As a general rule, if an engine has had its compression
raised from, say, 6 to 8 to 1 and has been otherwise attended
to as regards its ports and exhaust system, the ignition point
will need to be brought back by about 5, and possibly
more. The spark advance also varies according to the fuel or
more accurately according to the flame-rate or speed of combustion of the mixture. The rate of burning of methanol is
slow compared to petrol and for this reason requires nearly
as much advance at 14 to 1 compression as petrol-benzole
does at its maximum permissible compression ratio in the
same engine. But with blended fuels such as Shell 811 or
Shell M. which contain a proportion of petrol and benzole
yet can be run at over 12 to 1 C.R., the ignition point may
be perhaps 4 later. Some representative examples may be
seen in the table .overleaf. The only way to discover the
best setting-and even an error of lOin either direction
will make quite a big difference in a highly tuned engine-is
by test under power, preferably on a brake, a method of
which few amateurs can avail themselves.
Failing that, the best method is by road test using a rev.
. counter, the ignition being first set a little on the early side.
When the engine is at its maximum revs. in top gear, retard
the ignition very slightly. If the revs. increase, retard a little
more until the needle begins to fall back again. Then leave
the control in its last position, check the timing, and reset the
magneto to give the same timing, but with the lever at full
advance. For this test it is best to select a road which is
slightly uphill, so that the engine is definitely pulling hard.
Another method is to leave the ignition at full advance and
187
TUNING FOR SPEED
j
;
MODEL
COMPo
RATIO
FUEL
IGNmON
ADVANCE
Petrol-benzole
B.S.A., B32
Alcohol
Petrol-benzole
B.S.A., B 34
Alcohol
NORTON30M
Petrol
(79.6 x 100 mm.)
Petrol-benzole)
or Shell X
Shell A or 811
Petrol
NORTON40M
(71X88 mm.)
I petrol-benzole}
or Shell X
,
I Shell A or 811
NORTON30M
Petrol up to 80
(86 x 85.6 rom.)
Octane
NORTON40M
Petrol up to 80
(76 x 76.7 rom.)
Octane
VINCENT
Petrol up to 80
Octane
Shell 811
Shell A
9
13
9
34!O
7.5
4!W
38tO
37tO
7.5 to 10
36
12t
7.5
34
37-1
7.5 to 10
36
13t
9.5
34
34
9.7
39
38
12t
34
34
11
13t
CHAPTER XV
38t
make several runs, altering the contact-breaker gap in stages

by one-sixth of a turn on the points. A good magneto will
fire perfectly between the limits of .010 in. and .014 in. gap
measurement. If a notable increase is gained by, say, the
wider gap, check the actual timing, then reset the gap to the
correct figure of .012 in. and re-time the magneto to ~he
same figure. This is about the only satisfactory method with
a fixed magneto or when using an A.T.D .
.Unfortunately, though the correct 'advance can be determined very accurately by either of these methods, they
cannot be applied until the whole machine is ready for the
road. In the meantime a fairly accurate estimate can be
made from the maker's figures, taking into account the effects
of any modifications which have been made to the engine
internally.
188
So far as the cylinder 'head and valve gear are concerned there is no fundamental difference between the
work to be done on a single-, a twin- or even a four-cylinder
engine, provided that they all have individual inlet and
exhaust ports and pipes of equal length. Multi-cylinder
engines built in that way act simply as a collection of
singles and will give off power proportionate to the number
of cylinders. However, when a single carburetter feeds
more than one cy1inder, the inlet passages are necessarily
less direct and, therefore, offer more restriction to flow,
whilst slight differences in the manifold shape frequently
cause the mixture supplied to one cylinder to be richer than
the other.
If for reasons of cost, or because the race regulations
demand it, a single carburetter has to be retained, there is
the choice of running one cylinder at correct mixture
strength and the other too lean, or of increasing the jet size
to make the weak cylinder correct and possibly dropping
speed by reason of the other being over-rich. If the difference
is slight, the last-mentioned expedient is the best alternative
as there is less risk of damage to the engine. but it is
usually possible to eliminate the trouble by inclining the
carburetter towards the weak side. For initial experiments, a
thick gasket of medium-hardness rubber can be fitted and
the inclination can be adjusted by tightening the flange bolts
more on one side than the other. This enables rapid
comparisons to be made in a short period but, when the
angle has been established, a solid gasket should be made up
for permanent use.
Most parallel twins have a manifold with fairly sharp
b~nds. forcing the mixture to turn two right-angled corners,
this arrangement having been found to give better all-round
189
TUNING FOR SPEED
Special tools are

necessary to remove or
replace (see clutches
above and below respective(v) the cam
wheels of Triumph
twin-cylinder models.
Similar tools are available from other
manu!acture,:s.
results than one in which the two ports blend together at a

narrow angle close to the carburetter flange. Formaximum
power, it is of course best to fit two carburetters mounted
on spacers of the sam~ thickness as the manifold, or possibly
a little greater. As the head studs and the carburetter bolt
holes do not coincide, a little ingenuity is required in making
tip the spacers if it is not possible to buy them as spares.
Owners of late-model Triumphs are fortunate in that
they can purchase a racing kit containing all the necessary
bits and pieces. This is intended for engines supplied as
standard with aluminium heads and "Grand Prix" type con190
rods, but can be for earlier engines with iron heads which
are quite satisfactory on alcohol fuel. The rods in these
engines were of a lighter pattern than those fitted to the
genuine" Grand Prix" engine and it is best to obtain a pair
of the latter type if possible. The change of rods and
pistons will call for some re-balancing and, though the
factory balanCe factor is 62%, some riders find that 70%
gives better results.
Triumph camshafts are driven by wheels which are a tight
fit on the shafts and have three keyways for accurate adjustment of timing. When fitting racing camshafts, the wheels
are pulled off with a special remover (part number D 178)
and forced on again with a replacer (part number DI82),
and it is wise to obtain these tools before commencing the
change-over. TIre sequence of re-assembly is to fit the
crankshaft and camshafts into the drive side, not forgetting
to place the rotary breather-valve disc and spring in the
inlet cam-bush, then the timing-side case with the usual
precautions for making an oil-tight joint and again not
forgetting the two internal bolts just inside the crankcase
mouth. Assemble the pistons, cylinder block and head in the
normal manner, but fit only the exhaust rocker-box and
push rods with the clearance adjusted to zero so that the
rockers are just free enough to slide sideways.
Replace the key and half-time pinion, fit a timing disc
and pointer, accurately adjusted to t.d.c.; set the crankshaft
at 70 before b.d.c. and rotate the exhaust camshaft
forwards until one rocker is just tight enough to resist side
movement. Without allowing anything to move, offer up
the camshaft wheel so that one of the three keyways is in
line with the key at the same time that the teeth line up with
those on the intermediate wheel. Pull the wheel into place,
and rotate the crankshaft backwards until the rocker
again becomes tight. This is the closing point and should
be 52 before t.d.c.
Check the figures on the other cylinder and if the discrepancy between the two is greater than 4 it should be
equalized by re-positioning the wheel. It is unlikely that
191
TUNING FOR SPEED
the exact figures quoted will be realised at all four points

on both cylinders, but aim to obtain the nearest approach,
remembering that the exhaust opening point is the least
critical and a degree or two early will make little difference.
Before worrying too much about minor differences, check
again with the clearances set to, say, .005 in. The figures will,
of course, then be different, but may be more consistent
between the two cylinders, in which event the difference
between the figures at zero clearance is not unduly serious.
Procedure for timing the inlet valves is exactly the same
as for the exhaust, after which the rocker-boxes are fitted
permanently and the clearances set to .002 in. inlet and .004
in. exhaust. The ignition timing varies according to fuel
and is 42 for petrol or petrol-benzole, or 38 for methanol,
or alcohol-rich, fuels with the appropriate high-compression
pistons. If alcohol is used with low-compression pistons
just for the sake of cooler running then an advance of up to
45 may be necessary.
Triumph crankshaft assemblies are built up with the flywheel sandwiched between flanges. This is a very rigid
construction but, with the main bearings spaced very widely
apart, flexure is bound to occur to some extent and it is
prudent to avoid exceeding the maker's recommended
limit of7,500 r.p.m. When working on an engine of unknown
history, get the crankshaft tested for cracks in the region
of the junction of crankpin and web and, if any defect is
found, replace the component. The halves are accurately
spigotted and dowelled to the flywheel and alignment will be
accurate provided cleanliness is observed during the changeover. This precaution should be taken with any make of twobearing shaft after a considerable period of use.
The Norton "Dominator" shaft is of similar three-piece
construction and it is wise, when tuning an example which
has done much road work to dismantle the shaft and clean
out the large cavity which forms a sludge-trap in the centre
of the flywheel assembly. Mark all three components beforehand so that they can be refitted exactly as they were so as to
preserve original accuracy of alignment and balance.
192
As a rule parallel twins utilize forged aluminium rods,

and though those in up-to-1955 Triumphs run direct on
the shaft with white-metalled caps to provide a measure
of safety in the event of lubrication failure, it is usual to
employ precision-type thin-wall renewable steel big-end
shells, such as are commonly used on cars. These components must be handled with care and, in particular, must
. not be assembled unless the locating nibs are snug in their
recesses in rod and cap. If the clearance with the bearing
dry exceeds .003 in. the shells should be renewed; the temptation to file the faces of the rod and shells must be firmly
resisted, as the amount of "nip" or pinch on the shells has
been very carefully determined and it is extremely difficult to
maintain the same fit' exactly. In addition, once a rod or
cap has been filed, replacement shells cannot be fitted
correctly unless they also are filed, and, thus, the valuable
feature of interchangeability is lost.
Plain bearings must be given a greater quantity of oil at a
higher pressure than roller-bearings; they are more easily
damaged by fine abrasive which embeds itself in the white
metal and acts as a lapping compound. The greatest care,
therefore, must be taken to see that every vestige of dirt
is removed from oilways and filters; morever, the efficiency
of the oil-feed mechanism to the crankshaft must be checked.
The "Dominator," for instance, has a rubber oil-seal installed in the timing cover; it fits over an extension of the
crankshaft. This cannot be removed without it being
destroyed completely, but it is a simple matter to fit a new
one, which is pressed into place with the metal-covered face
visible.
Matchless and A.J.S. twins employ a three-bearing shaft,
which imparts extra rigidity and permits oil to be supplied
through the centre main bearing, and with little likelihood
of over-supply to one big-end at the expense of the other.
When working on an early A.M.C. model it is worth considering changing the original head with small fins to the
later large-finned pattern, which affords considerably
greater cooling:area. Similarly several pounds of weight
193
TUNING FOR SPEED
All extellded impressioll ofthe A.J.S.,Matchles~ crallkshaft

layout showillg the camshaft alld celltre bearl1lg support.
A.M.e. Ltd. are ullique ill providillg this desigllfeature.
can be saved by fitting the late-pattern aluminium head to

an early Norton "Dominator" with a cast-iron head.
There are three examples of V-twins in current use, namely
the 1,100 c.c. and 1,000 C.c. J.A.P; engines, built ex~ressly
for racing, mainly in sprint events, and the 1,000 c.c. Vmcent
in its various guises, "Rapide," "Black Shadow" and
"Black Lightning." Additionally, there must ~till be. in
existence a number of obsolete makes, of deslgn which
could be well worth working on-in the light of present-day
knowledge. There are of course also the American Harley194
Davidsons and Indians from which formidable power can

be wrung.
Dealing first with current J.A.P. engines, the 1,000 c.c.
model, known as the "8.80," is virtually two speedway
singles on a common crankcase, and has cast-iron heads and
barrels with shallow finning and therefore can only be run
on pure methanol, Shell" A" or J.A.P. fuel consisting of
97!% methanol,2% octane and !% Castrol "R." Lubrication is either on the total-loss system with sight-feed pump,
or on the dry-sump system, according to model; the former
is usually used for pure sprint work and sidecar racing on
speedways, the latter is superior for longer distances, as
some measure of internal cooling is obtained from the oil
circulation. Being designed expressly for speed, this engine
does not call for much work apart from the routine assembly
procedure described in previous chapters.
One major constructional feature is the con-rod arrangement, the rear being forked and the front one plain and
running on needle roller bearings on the outside of the sleeve
which forms the outer race of the main big-end. The two
rods cannot be separated unless the sleeve is pressed out,
and as it has three diameters, it can only be pressed out in
one direction. This scheme is adopted mainly to simplify
assembly as the sleeve can be pushed in by hand for twothirds of the full distance.
When reassembling, locate the needle rollers in the plain
rod with thick grease or with a dummy disc which can slide
out ahead of the sleeve as the latter is pressed home.
The 1,100 c.c. version was designed for use in cars and
has alUlninium heads with inserted seats, and aluminiumjacketed barrels; though usually used with alcohol fuel at
around 14 to 1 C.R. it can also be run on petrol merely by
changing to 7.5 to 1 pistons and tuning the carburetters
to suit. The crankpin has parallel ends and shoulders, instead of the usual J.A.P. tapers, and floating thrust washers,
which must not be omitted, are used to locate the rod
assembly. Similar washers are also used on each side of
the flywheel assembly and their thickness can be varied if
195
TUNING FOR SPEED
necessary to adjust the end-float to the design figure of

.0lD in.
The overhead rockers and also the valve guides are lubricated by a suction system common to o.h.v. engines of this
make; oil-mist, escaping past the rotary release valve which
ventilates the crankcase, condenses in a box below the
timing-case and is drawn thence up small-bore pipes to the
rocker-boxes by virtue of the depression which exists within
the engine. This system only functions if the rocker-covers
are in place and are reasonably air-tight, so that it is not
possible to check the flow of oil to the rockers by inspection
with the covers off.
The rotary release valve is driven by
the rear camshaft at engine speed and must be replaced with
the timing marks in line. In the unlikely event of having to
renew the bronze bush, this must be fitted with the slot
vertically downwards so that the valve actually closes when
the crankpin isa few degrees past its lowest position in the
crankcase, or about 300 after rear cylinder b.d.c. The timing can be verified by inserting a piece of wire into the
outlet pipe after removing the screwed plug in the bottom of
the oil-box. The two rocker-oil pipes projecting into this
compartment clear the flQor by s1z in.; this dimension is
important and must not be altered.
The duplex pump controls both the pressure and scavenge
oil supplies, the latter being drawn from the sump by an
external pipe. The slightest air-leak in this suction line will
impair the scavenging much more than one would think, and
it is vital to make sure thatthe unions are seating properly
and that the nuts are undamaged, otherwise the engine wHl
be sluggish and prone to oil the rear plug.
As two magnetos are fitted, each cylinder can be timed
individually. The recommended figure is 360 for the 1,100
c.c. engine and 340 for the smaller one both at a 14 to
"Black Shadows" have the same cams as " Rapides, " except
that they are selected to give "long" timings within the
limits of manufacturing tolerances, and con-rods are
selected for 65-tons minimum tensile strength and polished.
"Black Lightnings" have different cams, 85-ton "Vibrac"
con-rods, and a steel idler wheel is used in some early
examples; larger inlet ports, and of course racing carburetters and a special racing Lucas magneto are used,
with fixed timing pinion instead of the A.T.D. used on the
two touring engines. It will be seen therefore that either of
these can be "Lightningized" so far as performance goes,
without too much expenditure of time or money, and
without fear of overtaxing the "downstairs" section.
Standard heads have It in. inlet ports which can be
opened out to any size up to 1-& in. according to the size
of carburetters to be used-Iii- in. T.T.IO Amals are the
usual choice, for which a pair of flanged adaptors, part Nos.
ET32/6 and ET 32/7, are required. The ports can be enlarged
on the lines described previously, taking care not to go
through into the rocker box. As the guide is very short, it
is not advisable to cut it off, and a streamlined boss leading
to the guide must be left to avoid cutting through into the
recess in which the guide lock-ring is screwed. When
ultimate in power output is desired, obtain another front
head and a pair of
in. carburetters if you can' manage
to do so, though very good results can be achieved with
32 mm. carburetters. The reason for using a front head on
the rear cylinder is that as the port is not positioned on
the same side as the rocker box, it 610 be opened up to 1t
in. diameter.
These engines are very suitable, for short sprint work such
as speedway sidecar racing, and quite good results can be
obtained by retaining the standard port size and using
ordinary~ 1t type 29 carburetters as fitted to the "Black
Shadow," but suitably modified for alcohol fuel. "Black
Shadow" adaptors, which are bronze, are also required, as
the standard aluminium ones are rat1:ter too thin for safety
when opened up to suit the larger carburetters. Inlet valves
197
1 C.R.
The three versions of the Vincent unit are basically the
same; crankcase, flywheels, barrels and timing gear are
identical, except for the cams, and so are the heads and valve
gear, except for the port shapes and polishing of the rockers.
196
In-
TUNING FOR SPEED
with the heads 1 lin.-larger than the standard Ii in.-diameter can safely be installed and the seat-ring tapered to
-suit; it "is, in fact, preferable to do this rather than 10 fit new
seat-rings if the latter are badly worn.
The inner valve guides are normally of aluminium bronze
and held in place with lock-rings. When alcohol fuel and
castor-base oil are used, guide-wear may be excessive due to
lubricant being washed out by the fuel and cast-iron guides
may be a better proposition. Their reduced heat ~on
ductivity is not a detriment owing to the cool running nature
. of the fuel. A special tool is required to remove the lockrings, which only need firm pressure when being replaced,
but must be retained by punch-dots at two points other than
-opposite the slots. "Black Lightning" cams have about
.040 in. more lift than standard cams and, unless the seats
have :sunk considerably, the guides must be shortened by
050 in. to avoid any chance of the collar contacting the
guide. This job can be done in situ with a spot-facing tOQl
cutting the guides back until there is
in. space between
the guide and the step on the valve stem. Do not use an
ordinary drill for -the work, as the countersink which it will
form in the bore acts as .an oil~collector and plug-.oiling.may
result.
Six types of piston are available, ,giving ra.tios between 6.8
and Ui .to:1. The latter ratio can :be used with Shell "M,"
.but Shell 811 gives slightly more .power. Although the
difference in weight ,between the extremes in the range.is
,over 3 ounces thechange~over causes noperceptable roughness up to t6,800r.p.m., .and, therefore, the motor can be
.changed from petrol to alcohol without beingre-balanced.
Although the ring equipment is apparently :the same, the
.compression :rings in pistons E 7/9, 10 and 11, which are
the three highest, are of thicker radial depth than those in
E 7/6, 7 and 8,and, while iUs possible but not advisable
to use standard rings in the racing pistons, it is fatal to use
thick .rings in standard pistons astheyTproject above the
:ring lands. The oil ring, however, is jdentical.
For adjusting the compression ratio, base washers up to
198
"*
.062 in. thick may be used, or a small amount-not more

than .?30 in.:-may be turned off the jacket, though. generally
speaking, this should be avoided. There is not much free
length on the rocker adjusters, neither is there much room
between the head of the adjuster and the inside of the
rocker inspection cap at full lift, and damage to the valve
gear will result if the adjuster cannot lift freely which will
be the case if the barrels are shortened excessivel;. However,
the pushrods are easily shortened by grinding up to .030 in.
off either, or both, ends, and, if necessary, new ones can be
made from ~ in. silver steel, with the ball-ends hardened
simply by heating them to redness and plunging into water.
Apart from polishing, do not endeavour to lighten the rockers
more will be lost in flexure than will be gained by th~
reduction in weight.
The rocker spindles are .497 in. in diameter and, if one
loosens in the rocker, it is often possible to fit a .500 in.
gudgeon pin shortened to suit, or else use a piece of
hardened silver .steel which is simply driven into place,
though, naturally, the rocker-bearing bore must be reamed
---11
I---~--"'''/..' . '/"""-I
DEAD SMOOTH FINISH

_ _-r-_ON ROLLEjI. .RACKS
.765'
2.215"
.093" MAX.,
.
.750" ----1.218': i--:OBo- MIN., GROUND FILLET
./.... Ii 2 CROSS HOLE$" I .~.r 'of
T---+--r--.-~~_
~ 15.
SCREW
7/ ur--1-.
'040"
X 20 T.P.!.,
.....L--+.""'/~".I-"""
U:t:t1J
: ' SCREW. '/," X 20 T;";;.P:O.I
1'/."
.'.
020" R
,. ~
ROCKWELL HARDNESS;.
~
C60-G2 ON 1.500" 0 .. DRILL./>
HOLD THESE!.~ '/,'

WIDTHS CLOSELY'
MATERIAL 3% C.H. NICKEL STEEL
CASEI-iARDEN .OW DEPTH ON ,1:500' AND 1.57"'- DIAS.
THREADS TO BE SOFT
Details of a modified crallkpill suitable for use ill a 1,000 c.c.

.
Villcellt ellgille.
199
TUNING FOR SPEED
rr~6r
.952"
...'1-.(
ON ALL
M
;'/,,;'
O:;!:lAl1fE~'
1.575"
'f---!11'
"'1'
~::::J
}"-o. -x
8
I
=~-1J-8-E:-:Q~U~AL:-:-L::-Y'-:S::;PA:::CED
.S70"
SLOT LENGTH
'Ill"
L .' ..
J i::tJ,7
-1-+--,;--;- 1.810"
1.815"
.
003"
.570" CENTRAL. WITHIN. '-
CORN:E:R~S_:..~~~
L5 77"
SLOTS
EXACTLY PARALLEL TO <t

TO. FIT
x
.ROLLERS
'i,,~
'/;~"
WOODRUFF CUTTER
ALfERNA'TIVe METHOD OF FORMING SLOTS
BEFORE FINISHING TO SJ.ZE.
MATERIAL-WROUGHT ALUMINIUM H
.
.tw.0
Dimensions a/the MSS-type bearing cage

a/which are .used
itl conjunction with the modified crankplll Illustrated earlier.
out to suit beforehand. Looseness of the rocker-bearing in

the head is of no consequence so far as speed is concerned.
In the crankcase section, the standard big-end bearing with
three rows of 3 mm. by 5 mm. rollers works satisfactorily
for a long period if revs are .kept below 5,800, but prolonged
work above that speed shortens its life, so also does racing
the engine from cold before the oil has a chance to warm up
and circulate freely. If genuine spares cannot be obtained, a
very satisfactory big-end can be made by making a new ~in
as per the diagram, using two MSS Velocette cages. ~Ith
18 ftr in. by
in. rollers in each. Four of the ongmal
roller spacers must be ground down to .078 i~. thicknessa~d
placed one each side of the cages to centrahze them, whIle
the original sleeves which. are har,d all through can be. ground
out without moving them froIn the rods to 1.875m. The
MSS cages must be bored out to slip neatly over the shoulders
of the pins but, if these components are not available, cages
may be made up. Their width should be .950 in. so that the
steel spacers can be omitted altogether, although, of course,
the centrat'distance-piece must be retained.
200
As the big-end oil is fed into an annular space, the oil

supply cannot be cut off by incorrect assembly, but the two
oil holes leading to the rollers should be placed at a three
o'clock position looking towards the drive side. If placed
at 12 o'clock, the additional oil pressure created by
centrifugal force causes almost all the oil to pass out through
the first hole, thus starving the drive-side big-end .
In the timing-gear department, the front exhaust camspindle should be locked, after fully tightening, by angledrilling a small hole from the end of the slot in the head back
towards the camshaft and fitting therein a split-pin. Worn
cam followers can be trued by grinding or oil stoning, but
the face must be kept parallel with the hole. Care must be
taken to note the positions of all the thin hardened washers
behind the steady-plate and to replace them in exactly the
same order. The lowest hole in the plate is purposely made a
loose fit on the idler wheel spindle to allow for adjustment, but
it is best to cut a strip of soft aluminium as wide as the plate
thickness and ~ in. thick, wrap it into a circle round a 1- in.
bar and pull it into position in the steady plate with the
idler-spindle nut. This holds the outer end of the spindle
positively in position, the soft bush being able to flow somewhat to accommodate itself to any slight eccentricity of the
hole and spindle.
Better crankcase ventilation is obtained by filing straight
across the breather-valve sleeve until the slot is fully i in.
wide, then rounding the outer edges of the flat so that the
outgoing air has a clearer passage; the edges of the slot
should be left sharp to act as oil scrapers. Replacing the
external banjo and pipe with a straight union and a large-bore
pipe running upwards and rearwards also helps ventilation.
Cams and camwheels are press-fitted together and are
regarded by the factory as inseparable units. Consequently,
"Black Lightning" cams are only supplied with the wheels
attached, but it is possible to change cams and wheels
provided that the interference fit is between .001 in. and
.002 in., otherwise either the wheel may slip or itmaycrack
through overtightness. All cams are ground in relation to
201
TUNING FOR SPEED
the slot in the end-face, which must be placed in correct

relationship to the marked teeth. This can be accomplished
by scribing lines on both the faces of each gear, and
accurately lining up the new cams with the lines before
pressing them home level with the face of the gear. There
are timing dots on both gear-wheel faces, so be careful to
place each gear on the same way as it was originally; t~e
timing should then be correct when all the gears are agam
assembled to the marks.
A rev. counter cannot be driven from the magneto-pinion
nut satisfactorily, partly because it is a clockwise magneto
and the nut would tend to unscrew, and partly because the
automatic advance would not operate correctly. It is
possible to devise a component which fits loosely over the
A.T.n. and is driven from the fixed stops on the fibre
pinion. but the better scheme is to obtain a fixed pini~n.,as
used on the "Black Lightning." together with the dnvmg
dog and special magneto pinion cover.
.
Strictly speaking, the magneto should be conve~ted ~o
manual advance, but with high-alcohol fuel the engme Wlll
start at the full advance of 34 at a 121 to 1 C.R.
. As there is no provision for measuring tappet clearance
the method adopted for checking valve timing is to attach
a dial indicator to one head stud and screw the tappet
adjusters down until the valve is just lifted from its seat by
.002 in. to .003 in., the exact amount being immaterial as it
merely serves to take up all backlash. Re-set the dial gauge
to zero and turn the engine until the valve is lifted a further
..005 in.-this is important. The point at which this lift
occurs is taken as the opening or closing point of the valve
according to whether the valve is lifting or dropping, and
all the points so obtained should agree with those quot~d
in the Table in Chapter XII. If there is an average dIScrepancy of more than 4 between the two cylin~ers. an~
there is still a discrepancy when re-checked at .010 m. lift, It
may be necessary to re-fit one of the two camwheels; the
timing as a whole can of course be set by re-positioning the
half-time pinion on anyone of its five keyways.
202
Before replacing the timing cover. see that all four of the
synthetic rubber seals are in place in the recesses provided,
and in no circumstances omit to fill the filter chamber with
the correct oil before fitting the banjo plug in the cover.
Failure to observe this precaution may lead to damage
occurring in the time lag before the pump fills the-chamber.
In final assembly, verify that all shock-absorber springs
are nestled in the recesses in the spring plate and then get
the mainshaft nut really tight. The hexagon size is t-in.
Whitworth (the same as the crankpin nuts), and an S.A.E.
spanner 1-.& in. across fiats also fits. The tube spanner in the
kit is inadequate for the job, which requires a solid steel
box or ring spanner.
The best scheme is to hold the crank assembly with a
i-in. bar through one small-end eye; in any case it is always
advisable to tighten up the shock-absorber assembly before
reassembling the cylinders, so being able to verify that all
is correct down below, and to measure the centrality of the
small-ends in the crankcase bores. The foregoing notes
are in the nature of underlining some features of this power
unit which are a little unusual to those not familiar with its
construction, and must be read in the light of what has been
said elsewhere in this volume.
c
Twin-cylinder Carburation
Mention has been made already of the advisability of

using separate carburetters on a vertical twin, and the need
is even greater on a V-twin since It is virtually impossible
to obtain equality of mixture with a single carburetter
owing to the unequal periods between induction strokes. A
duplex instrument with a separate choke for each cylinder,
or preferably two individual carburetters, should be fitted.
When experimenting with two carburetters on a V-twin
it may be necessary to use different forms of instrument on
the front and rear cylinders, in which case the main jets,
throttle slides and needle positions are almost certain to be
different. As an example a horizontal 1Ttr in. carburetter was
203
TUNING FOR SPEED
used on the front cylinder of the pre-war Vincent-H.R.D.

"Rapide," with a 6/3 slide and 180 jet, whereas the rear
carburetter was of the same size but had a vertical mixing
chamber, a 6/4 slide and 170 jet. For racing 11 in. instruments could be fitted with slides and jets to suit.
Generally speaking, on special layouts it is best to avoid
Conversion kit necessary to adapt a standard Arnal twist-grip

assembly for dual carburel1er control.
using horizontal mixing chambers because the main jet is

not fully submerged and the mixture strength, therefore, is
very affected by fuel-level. For that reason, post-war
Vincents were designed to use carburetters with vertical
mixing chambers, although the example which first raised
the American record to over 150 m.p.h. had both its carburetters bolted straight on to the heads with the mixing
chambers only 10 from horizontal, which was just sufficient
to bring the main jets below the fuel-level and ensure a
stable mixture.
The usual method of operating two carburetters is by
means of a single wire from the twist-grip to a junction
box, from which separate wires run to each throttle. This
system sometimes causes the grip operation to be rather
heavy and some riders are persuaded to shorten the throttle
springs as the easiest way out. This is a bad idea because, if
the slides do not cJose properly, erratic tick-over results.
Pay attention to the run of the wires, leaving off all clips
so that the wires can adopt their natural position. This usually effects a cure, but the best method is to obtain a dualcable twist-grip or modify the existing one to take two
wires.
It is essential to adjust each control so that both cylinders
accelerate absolutely in unison. The best method with
standard Amals is first to slacken both cables and set the
tick-over on each cylinder by means of the throttle stops
and pilot screws until even running is obtained. Next set
the cable adjusters so that the slightest grip movement
causes a rise in engine speed with. each cylinder firing
equally.
If with both slides apparently moving simultaneously,
firing is still uneven, try the effect of closing each air lever
in turn. If, say, one cylinder is missing or does not fire at
all yet cuts in when its air control is closed, less cut-away is
required on the throttle slide. In border-line cases, an
enriching of the pilot mixture may help the offending
cylinder to come off the pilot at the expense of slightly
erratic slow-running. It is, of course, essential to have an
204
205
'I'"~
DRILL 'h" DIA.

FOR SPECIAL NIPPLE
GROOVE TO TAKE
1 CABLES SIDE BY SIDE
-j ~/t
1,.
tJ
11ALMOST
1 CABL; H. OLEoS ~
TOUCHING
,
EACH OTHER
I
SPECIAL :1 CABLE NIPPLE
BORE .1:0 FIT GRIP SLEEVe- -.OUTSIDE DIAMETER AS ORIGINAL GRIP

FLANGES TO BE CUT. AWAY ALSO A$
ORIGINAL GRIP
i"-1
HOLE~
EXISTING"~-
AS
CABLE ST01
-l
i"
i"
t-
L.:EC. /
I
FOR
CAB~ES
-tm---,
~
2 SLOTS CUT
ON
ANGLE
'HOWN '0' CAm,
cr::=J
TUNING FOR SPEBD
individual air control to each cylinder; by intelligent use of

that control equal carburation can be obtained throughout
the entire range.
As there are no throttle-stops on racing-type carburetters, this system cannot be used; the idling speed can only
be adjusted by varying the cable lengths but, if cut-aways
on both slides are equal it is essential to verify that they both
disappear simultaneously at the tops of the choke as the
grip is opened, otherwise acceleration will not be good. If
there is a big difference it will be necessary to check for air
leaks or fuel blockage which may have affected the idling
speed. Incidentally, it is quite useless to attempt to adjust
the idling setting on one cylinder only with the other cut
out by shorting the plug. The amount of throttle required
to pull the engine round with only one lung operating is so
much greater than with both in action that the scheme gets
you nowhere.
. Accuracy of the main jet settings must be verified finally
by the appearance of the plugs, in exactly the same way as
for a single-cylinder machine.
CHAPTER XVI
TWO-STROKE ENGINES
TWO-STROKE engines differ fundamentally from four-strokes
insofar as the charge is not induced directly into the
cylinder, but is transferred to it from the erankcase, or
from a separate pump which may be either a rotary
blower or a cylinder-and-piston mechanism. Either of the
two last-mentioned devices are permissible for recordbreaking but, under the rule that prohibits the use of superchargers in road-racing, an auxiliary pump which has a
greater swept volume than that of the cylinder is definitely
"out" and so are additional pumping pistons which act to
increase the volume drawn into the crankcase.
Of recent years, therefore, the tendency has been to discard supercharged, or augmented-induction, designs which
were in vogue before 1939 and to concentrate on the much
simpler forms, relying solely upon straightforward crankcase compression such as are commonly used for touring
work. Some really amazing results have been achieved by
working upon these engines, speeds of 85 m.p.h. and over
being obtained by private owners from 125 C.c. B.S.A.
Bantams" without recourse to fuels such as nitrobetlzine.
and good results can also be obtained from Villiers engines,
which are used in the Dutch Eysink machines, and the more
complex but still relatively simple split-single E.M.C. and
Puch designs.
The basic line of development remains the same as for the
four-stroke, namely to get the maximum quantity of fresh
charge into the cylinder at the highest possible r.p.m., but
it is more difficult to carry out because there are two sources
of loss in volumetric efficiency-one in getting the fresh
mixture into the crankcase and the other in transferring it
to the cylinder. Consequently every endeavour must be
made to reduce any losses in breathing ability to a minimum.
207
U
206
TUNING FOR SPEED
Dealing with the crankcase first, the initial step is to

reduce the clearance space within it as much as possible by
using disc flywheels or by building up existing flywheels or
crankwe bs until they almost fill the surrounding space and
only just clear the connecting rod. This procedure may
introduce some problems of balancing which may have to
be solved later, but, at the outset, filling must be done in the
lightest fashion either by using magnesium blocks or by
sheet metal work, though in the latter case all joints communicating with cavities must be air-tight, otherwise the value
of the scheme is largely lost. There must however be some
clearance between moving and stationary surfaces. A
certain amount of gas movement takes place therein,' so it
is advisable to' polish both the flywheel assembly and the
whole interior of the crankcase to assist this movement.
When, as is usual, the inlet port is controlled by the
piston skirt, the time available for induction is extremely
short, less than 1/300 sec. at 6,000 r.p.m., which can be
considered a comparatively low speed for this type of
engine in highly-tuned form. The ability of the port to pass
gas depends upon its "time-area integral"; that is the
combination of the area open at any given instant and the
total amount of time elapsing between opening and closing
points. Widening the port around the circumference gives
a greater area, without altering the timing; deepening the
port increases both the area and the time of opening, but
must be done with great care, because the effective compression stroke of the piston in relation to the crankcase
does not commence until the inlet port closes, unless the
induction tract is of such a length that an appreciable
ramming effect is generated by the fast-moving column of
gas. To some extent, this ramming action is always present,
but unless it comes at the right time, which is just towards
the latter end of the closing period, it will be of little assistance. Actually, the combination of the crankcase and
induction pipe (including the carburetter) constitutes a
resonant system equivalent to a closed vessel with extension
pipe and has a natural vibration frequency which can be
208
TWO-STROKE ENGINES
used over a limited speed range to augment the ramming

effect very considerably. Whether the problem is amenable
to calculation with any accuracy is rather doubtful because
of the number of variables involved, so the solution is best
arrived at by trial and error.
Widening the port around the circumference is quite
permissible and has little adverse effect because the pistonrings do not usually traverse this port, and naturally all the
interior surfaces must blend well into each other and be
polished. Provided that one is prepared to sacrifice lowspeed torque, relatively enormous carburetters can be used;
1i in. or I,\- in. type 29 Amals give very good results, and
are less expensive than the R.N. or T.T. patterns.
From the crankcase, the gas, after compression, enters the
cylinder through the transfer ports, and the angle of entry of
these into the cylinder is of great importance in all flat-top
piston designs. The early versions used two opposed
exhaust ports and four transfer ports, the latter being so
arranged that the streams of gas from each impinged at the
centre-line and deflected each other up towards the top of
the cylinder whence they curled downwards towards the
exhaust ports. In later versions with only a single exhaust
port there are only two transfer ports, which direct the fresh
gas across the piston crown almost in the opposite direction
to the outgoing exhaust gas. The streams then combine,
travel upwards towards the head and then downwards
towards the exhaust ports, the principle being referred to
as loop-scavenging.
The angles of the entering gas streams are most important
to obtain best scavenging with least loss of fresh charge out
through the exhaust port and, though the ports can with
advantage be enlarged a little, this process is best done in
stages. In some designs, small cover-plates are provided
through which access to the ports can be obtained but in
others the casting is solid. In the case of the latter it is
possible to build up around the elbow of the transfer passage
with bronze, then machine the surface off level and break
into the passage. The holes can finally be closed by fitting
209
TUNING FOR SPEED
screwed-on cover-plates with internal extensions shaped to

conform exactly to the interior contour. Any obstruction
to gas-flow in this region is extremely detrimental, so that
the whole of the transfer passages must be smoothed off and
particular care taken to improve the entry into them from
the crankcase by rounding off all sharp comers.
Exhaust ports in touring engines are usually made with
the top edge inclined or angular; this gives a gradual
opening and, by taking some of the "crack" out of the
exhaust noise, simplifies the silencing problem to some
extent. The inclined edge gives the rings an easier passage
as they pass over it on the upward stroke. For racing, it is
desirable to obtain the quickest possible rate of opening
in order to make use of what is termed the" Kadenacy
effect." Kadenacy discovered that, if the ports are large
enough and are opened with sufficient rapidity, escaping gas
rushes out with such vim that the cylinder pressure drops
several pounds below that of atmosphere, and, by adding an
exhaust pipe of the appropriate length, fresh gas is drawn
into the cylinder through inlet ports even without the aid of
crankca~e compression or an external blower. In fact, one
industrial engine, the Petter "Harmonic," after being
started on crankcase compression, is run thereafter purely
on air induced through the agency of tuned exhaust and
intake pipes. This engine, being a Diesel, operates all the
.time with a full air supply to the cylinder and is thus not
directly comparable with the carburetter-type engines under
discussion, but it is mentioned to indicate the great influence
and importance of the exhaust system.
During the period of port-opening we have a cylinder,
with its capacity continually varying as the piston moves
towards b.d.c., connected by an orifice (the exhaust port),
also of varying area, to an exhaust pipe, which is the only
non-variable item in the system. As a further complication,
the natural frequency of any system depends upon the speed
of sound in the gas contained in it, and though it is usual
to take 1,500 ft. per sec. as being a reasonably close figure
for a normal four-stroke engine, sound-wave velocities of up
210
TWO-STROKE ENGINES
to 4,000 ft. per sec. may be attained in a short exhaust

system with a rapid discharge of high-pressure gas through
piston-controlled ports. This factor alone would make
calculation of suitable exhaust-pipe length difficult, to say
the least, and that factor, in conjunction with the other
variables, accounts for the wide diversity of systems in use.
The Lambretta, for instance, has, in effect, a curved megaphone which starts right at the port and widens to about
4 in. diameter in 16 in. length, whereas the Eysink Villiers
has two pipes, of more normal length, exhausting into fishtailed expansion boxes. B.S.A." Bantams" operate well
with a It in. pipe about 20 in. long, but for r.p.m. in excess
of 8,000 the pipe can be shortened to about 10 in. if the
regulations will permit a system of that length.
The functioning of a two-stroke exhaust is more complex
than that of a four-stroke; it must first assist in evacuating
the cylinder rapidly and in so doing is almost bound to draw
a certain amount of freSh charge out through the port. By
correctly proportioning the exhaust system, this charge can
not only be retained in the pipe but can be rammed back
into the cylinder by a wave of positive pressure arriving at
the port towards the end of its closing period. Conversely,
if the system is incorrect or the relationship of the transfer
and exhaust closing points is wrong, a considerable portion
of the charge may be lost irretrievably. Poor power development with a high fuel consumption and the strong smell
of unburnt fuel in the exhaust are all signs that serious
charge-loss is taking place. If the work is being conducted
under laboratory conditions, an Orsatt gas analyser is an
invaluable instrument for determining charge-loss as, by
its use, an accurate measurement of the amount of free
oxygen in the exhaust is obtained. There is bound to be
some free oxygen, up to perhaps 2%, due to combustion
not being absolutely complete, but anything above this
figure (provided the mixture is correct) represents an actual
charge-loss of five times the amount of excess, because, of
course, only! of the atmosphere is oxygen. With the aid of
the Orsatt device, therefore, the effect of port- or exhaust211
TUNING FOR SPEED
TWO-STROKE ENGINES
pipe modifications can be determined quickly. But without

it, recourse must be had to the time-honoured "cut-andtry" method which, fortunately, is not too difficult or costly
with a small single-cylinder engine.
A general scheme which has possibilities is to raise the
upper edge of the inlet port until it is t in. or so higher than
the edge of the piston crown at b.d.c.,. so placing the
carburetter in direct communication with the cylinder for a
few degrees of crank-travel when the exhaust port is almost
fully open. The reduction of cylinder pressure which comes
about by the Kadenacy effect then causes fresh mixture to be
induced through the inlet port at the same time that the
normal supply is being fed into the cylinder via the transfer
ports. There will be no supercharging created thereby, but a
gain in power will result because of improved scavenging
of the residual exhaust products. This device is essentially
one which will work well only over a limited speed range
and may result in very poor power output at other speeds.
When modifying port heights, the simplest method for
trial purposes is to alter the cut-off edges of the piston instead
of altering the barrel. The effect is not exactly similar because whilst the port timing is altered the actual area is not,
but it will at least give an -indication, and if a mistake is
made, it is cheaper to replace a piston than to start all over
again with another barrel. One of the most exasperating
things about two-strokes is the difficulty of duplicating
results; two barrels with ports which appear to be identical
will rarely give equal power, possibly because small differences in port contour, which are difficult to measure, create
differences in the direction of gas-flow and modify the
degree of scavenging to a much greater extent than might be
expected. The moral of this is that if you do get a good
barrel, look after. it; do not think that all you have to
do when the experimental one is worn out is make another
which looks the same.
One of the difficulties with a two-stroke is cooling; the
cylinder does not have the benefit of the idle inlet stroke, nor
is the fuel vaporized actually in the cylinder as in a four212
stroke. Consequently the heat-loss to the walls is high, yet a

large part of their area is occupied with ports which cause
unequal temperatures and local distortion. Some riders go
to the trouble of turning off the fins above the ports and
shrinking on a finned aluminium jacket, but this does not
really get down to the root of the matter and it is doubtful
if it is worth doing on an engine used on road circuits where
the greater cooling required by the increase in power is
automatically supplied by the increase in air-speed.
Cylinder-heads with larger fins and a smaller combustion
space giving compression ratios of about 12 to 1 are becoming commercially available for B.S.A. "Bantams";
as the pattern-work is relatively easy these are not difficult
to make, though trouble with cracking round the plug-boss
may be experienced unless the material is good. Heat-treated
Y -alloy or R.R.53 will be perfectly satisfactory. In designing
such a head it is easy to obtain an impressive appearance
without much gain in cooling; to be of real value, fins must
spring directly from the actual metal around the combustion
space, there being little or no virtue in putting a fin on a fin.
The old Villiers system of radially-disposed fins is excellent
as the positioning of the fins is good and the air-stream has a
clear path past each one, which is not always the case with
deep parallel fins.
Flywheel magnetos appear to be able to run satisfactorily
at any speed which can be got out of the engine but some
experimenters prefer to utilize a 180 twin-cylinder magneto,
running at half-engine speed with both leads going to one
plug, which is standard practice on the E.M.C.-Puch, or
with each lead going to its own plug in a special two-plug
head, a scheme which gives the plugs an easier life. The
reason for using a half-speed magneto is simply to reduce
stresses in the instrument, as a normal magneto is not
ordinarily called on to run at more than 4,000 r.p.m. It is
essential, however, to have the magneto checked electrically
to verify that the sparks do occur precisely at 180, otherwise
correct ignition cannot be obtained as a one-degree variation on the magneto means 2 variation on the engine.
213
TUNING FOR .SPEED
TWO-STROKE ENGINES
It is general practice to use uncaged roller big-ends, there

'being insufficient lubrication to permit the use of cages. In
:some Villiers engines, the steel rollers are alternated with
bronze rollers of slightly smaller diameter, the idea heing that
-the bronze rollers do not carry any load but merely separate
'the steel rollers and rotate backwards; other models use the
.conventional crowded-row system. The big-end assembly of
tbe B.S.A. was enlarged in 1954 and the later pattern should
;be used as a replacement on early models.
Interference-fit crank-pins without retaining nuts are
-commonly employed and can be assembled either in a press
or a stout vice with parallel jaws. As there is some loss of
interference-fit caused by removal and refitting, the replacement pins should be .001 in. larger than .the originals.
Villiers spares are supplied as .standard with this amount
Jof oversize. Greatcaremust betaken to start the pin squarely
in the hole to avoid any chance of damaging 'the surface,
,.and a piece of.-shim steel.OlO in. thick should be interposed
between the rod and one crank-web to limit the final position
-of the crank-web ;and provide the necessary amount of side
.clearance. For finally truing-up the shafts, the easiest way
us to holdtheassembly in one hand and knock one crank-web
'in the required 'direction with a copper hammer; this is more
.effective with small assemblies than the "bumping" method
,employed with much heavier four-stroke flywheels.
Although seemingly simple, the design of a twin two-stroke
..crankshaft is .not easy, because of the necessity to provide a
~sealed bearing between the crank chambers. On the
Exce1sior"Talisman" the right.handassembly is built up
, and fitted to the case; one-half of the left-band assembly
'follows and the central retaining nut fully tightened. Then
. the con-rod is fitted to this assembly and the last crank-web
is placed in position. To facilitate this part of the work,
"the crank-pin is not a very tight fit in the web, the necessary
,rigidity being provided by supporting the main-shaft in two
:rrelatively widely-spaced bearings.
Crankcase ;pumping efficiency depends largely upon the
~seals provided at the main bearings, and which vary from
214
make to make. The system common to all modem Villiers

engines consists of flanged bronze bushes which are a sliding
fit in the crankcase and have a slight clearance on the shafts.
The flanges are pressed lightly against the crankcase by
star-washers; none of these parts can be examined unless
the case is heated and the bearing tapped out but they give
little trouble unless the engine has been run with loose main
bearings.
.
On B.S.A. "Bantam" engines there is a number of shims
arid one collar on the mainshafts; these must be replaced
in exactly the same order; so their initial position should be
remembered.
Mention was made in an earlie~' chapter of the necessity
for accuracy of fit of two-stroke pistons, which have to be
reasonably gas-tight on the skirt and must not run with
excessive clearance. One line of attack is to fit the piston
deliberately a little on the tight side, and give the motor a
short burst on full throttle after sufficient running to warm
everything up thoroughly. Then strip down and very carefully e'ase off piston areas on which excessive pressure is
evident; a fairly rough file is better than a smooth one for
this work, as the file marks act as oil reservoirs and may save
a bad seizure. This process should be repeated several times
until the motor will hold full throttle for a
couple of miles with no sign of tightening.
It is useless attempting to run-in a twostroke piston by covering a long distance at
part-throttle openings; it is almost bound to
seize the first time it is given a prolonged
burst at full bore, because the temperature
A step in the crankcase-transfer

port joint hinders gas-ftow
and, therefore, impairs performance.
215
TUNING FOR SPEED
distribution and, consequently, the shapes and dimensions

of piston and barrel are then quite different from what
they were during the light-load running period.
When using alcohol fuel, Castrol "R" or a similar type of
lubricant must be used because it is soluble in the fuel,
whereas mineral oil is not. The proportion of fuel to oil
should be raised from the usual 16 or 20 to 1, to about 10 to 1,
at which strength a continuous faint blue haze will be visible
from the exhaust. For distance racing, where full throttle is
held for long stretches, an auxiliary drip-feed from a gravity
tank could be arranged to supply oil into the carburetter
intake, the feed being put into action either by a separate
control or by utilizing the last part of the throttle cable travel
in some manner which the reader can puzzle out for himself!
The foregoing is essentially of a general nature. Specific
data regarding the B.S.A. "Bantam" was the subject of
"Speed from a Two-Stroke," a two-part article featured in
Motor Cycling in 1954, at a time when the revision of this
book was being prepared, and I noted then that the writer,
A. E. Rose, imparted considerable wisdom. Some of his
text overlaps, or enlarges upon, my own remarks earlier
in this chapter but, for the -benefit of" Bantam" owners, it
will stand repetition. Mr. Rose advised noting carefully the
crankshaft angles at which the ports open and close on
standard "Bantams." They are as follows:
Exhaust opens 70 b.b.d.c.
Transfer ports open 60 b.b.d.c.
Exhaust fully open 10 b.b.d.c.
Transfer ports fully open 10 b.b.d.c.
Transfer ports close 60 a.h.d.c.
Exhaust closes
70 a.b.d.c.
,
Opening period slightly later on recent engines.
The carburetter inlet aperture has a V-notch at the bottom.
and for this port the timings are:
216
217
TUNING FOR SPEED
Closes t6 bottom of rectangle 50 a.t.d.c. (130 0

b.b.d.c.)
Closes to bottom of V-notch 60 a.t.d.c. (120 b.b.d.c.)
Opens to bottom of V-notch 60 0 b.t.d.c. (120 a.b.d.c.)
Opens to bottom of rectangle 50 b.t.d.c. (130 a.b.d.c.)
Fully open but shrouded by:fz in. of piston skirt at t.d.c.
For these checks an ordinary timing disc is necessary;
it is attached to the mains haft and, in conjunction with a
fixed pointer, indicates the crankshaft angle at which the
piston reaches the upper or lower port limits. Having
recorded the standard port timing, start dismantling the
head, barrel and primary drive in the usual way.
Further extract from" Speed from a Two-stroke" covers
practical work. In planning port modifications it is suggested that it is best to bolt the cylinder to the offside crankcase half, using the head, or two distance sleeves, to pull it
down into the normal position. Inspect the junction of the
transfer port in the crankcase with that of the cylinder at
the joint faces. Probably it will be found that a step exists
between the two and that the corners do not blend. Carefully
note these conditions and log the measurements involved.
Repeat by bolting cylinder to the nearside crankcase half.
Repeat the whole procedur~ with the crankshaft assembly
and piston in position and observe how the cut-away of the
piston and cylinder mate at b.d.c. Probably they will not
line up with each other and the condition must be noted.
It may be of interest at this stage to note standard "Bantam"
ratios and the "close" alternatives. These are shown in
the Table opposite.
The parts required for a conversion to close-ratio gears
in the "Bantam" are:
(1) Gearbox main shaft and fixed pinion with 19-t.
B.S.A. Part No. 90-445.
(2) Layshaft pinion 23-t.meshes with (3). Part No. 90-447.
(3) Mainshaft primary gear (sleeve gear) with 24-t.
Part No. 90-474.
218
TWO-STROKE ENGINES
(4) Layshaft sliding pinion (C.R.) 25-t. Part No. 90-449.

(5) !:thaft gear-meshes with (1)-28-t. Part No. 90Use. the standard layshaft but remove the 19-t. gear, observIng correct way round, and press on the 23-t. gear pinion.
The standard "Bantam': engine has a quoted b.h.p. of
4.5. at 5,000 r.p.m. and this corresponds with a maximum
speed of almost 50 m.p.h. in good conditions. Without
going too deeply into technicalities, it can be said that an
~ppreciabl~ increase will not be obtained unless the engine
IS able to Induce a high charge of mixture which with a
~tanda.rd i in. choke carburetter, is not pos~ible. I~creased
InductIOn volume automatically demands an enlarged
c?oke. Comparison of the values here can be assessed by a
SImple squarIng of choke diameters.
For example, a choke of i in. diameter compares with the
standard size as follows:
ixi=gQ
iXi=~=25: 49
Therefor~ a choke diameter only slightly larger than
standard ~ves nearly double the choke area. It might well
be that a slightly smaller choke than i in., or a slightly larger
STANDARD "BANTAM" RATIOS
SLEEVE
SilDING
INTEGRAL
Mainshaft
Layshaft
28-t
19-t
22-t
2S-t
lS-t
32-t
Mainshaft
Layshaft
24-t
RATIOS
Top 7 :1
2nd 11.7 :1
1st 22 :1
As above with 15-t gearbox sprocket and standard
47-t rear wheel sprocket.
CLOSE RATIO
23-t
22-t
2S-t
219
19-t
28-t
Top 7
:1
2nd 8.4:1
1st 10.78 :1
TUNING FOR SPBBD
one, would give equally good results; the fonner would, no

doubt, be better for grass-track or similar purposes where
better acceleration from comparatively low engine revolutions is required. A choke of about 11 in. is recommended for
this purpose, while a choke of up to 1 in. could be used for
fast circuits but acceleration might be affected until 4,000
.r.p.m. or so was reached.
The Amal T.TAype carburetter is not available in choke
sizes below t in. There are certain advantages in using this
type of instrument; it is easier to determine and maintain
the correct mixture with the T.T. pattern and the writer's
experience has shwn that it provides very easy starting, no
flat spots. good acceleration and a satisfactory maximum
speed when fitted to "Bantam" engines with a suitably
modified inlet port. Settings for this instrument are: pilot
jet, 1t to 2 turns open; slide, No.4; needle position, 2nd
groove from top.
Main jet recommendations vary considerably according
to fuel and exhaust-pipe lengths. Jets larger than 300 can
be fitted to highly-tuned engines, intended for fuel of around
80 octane value, with a greater increase still for alcohol fuels.
If the older type brass T.T. instrument is used, the main jet
sizes required may be 25% smaller than with the modern
alloy T.T. Amal component.
Carburetter Tuning
In order to keep all information about carburation
. compact and to avoid splitting these notes into disjointed
paragraphs later on, the carburetter-tuning technique to be
employed when the engine is ready to run is given now.
The method recommended by A. E. Rose is an adaptation
of that used by the well-,known exponent of two-strokes,
John Hogan, and a great deal of time is often saved by adopting it. It is assumed that a t in. alloy T.T. Amal is used and
that the pilot jet, slide number and needle settings, as given
above, are adhered to. A No. 260 main jet is suggested to
commence with, but the engine may accept a larger one.
220
TWO-STROKE ENGINES
An exaggerated example of irregularity

in the mating of the
piston cutaway and
cylinder port areas.
. The object is to get the engine running at maximum r.p.m.

will cause it
Just to begin to four-stroke when the air lever is moved
to the i-open, from the fully-open position. It may be
that the jet initially chosen will cause four-stroking right
from the start, even with the air lever fully open, or prevent
a smooth take-over from the pilot jet. The cure is obvious
and smaller jets should be fitted progressively until reasonably clean running is obtained with the air lever fully
open.
When this is achieved, give the machine a few runs and
try ~losing the air lever down to the i-open position. If the
e!lgme four-s!rokes badly ~ reduction of 10 or 20 c.c. in jet
SIZe should give a clean nuxture at the fully-open position.
Be careful not to go too low in jet sizes. If one of 220 is
found to be too rich, try setting the needle in the top groove
before going down farther in jet size. Spitting-back indicates
a very weak mixture. New experiments should also be
carried out whenever inlet or exhaust-pipe lengths arevaried
however slight these variations may be. If possible, a tuned
two-stroke should be run on a slightly rich mixture rather
than a slightly weak one, as the former keeps the engine
cooler and the occasional stutter through richness will not
reduce speed appreciably.
The air lever is best fitted on the left handlebar as it is
much easier to manipulate there; this leaves the right hand
free to hold open the throttle, which race regulations require
to roll back on release.
221
~n top gear and to choose a main jet which
TWO-STROKE ENGINES
TUNING FOR SPEED
In order to minimize the amount of grinding work and

filing, and to improve access to the port, the inlet stub
should be sawn off to within .fo in. of the rear fin tips. Three
first-class riffler files, double-ended and of assorted shapes,
should be obtained by people not in possession of a grinding
shaft. The standard port outline, where it enters the
cylinder,is shown sketched with the new shape superimposed.
Gradually open out the port and blend it into a 1 in. diameter
orifice at the cut-off stub.
No more than ~ in. should be removed from the top of
the inlet port. The sides may be swept out up to i in. and
the base curved around to about ~ in. below the original
"V." Do not exceed a total depth of ! in. or the width
shown, for the piston-ring ends lie comparatively near to
this port at the bottom of the stroke. Owing to the size of
the "Bantam" cylinder casting at the cut-off stub, it may
not be possible to attain a truly circular port of i in. diameter but this may be overcome by introducing a slight
ovality, blending into the increasing width of the port.
A piece of Ii in. O.D., ~ in. LD. tube should be counterbored at one end to accept the iii in. stub (here only about
--
'\.. PISTON SKIRT
The original inlet-port

profile and (shaded) the
area of metal to be removed to procure a
modified outline suggested to improve
" Bantam" induction.
.'
:~
'..
222
OPTIONAL
\STIFFENING
PLATE
Details of modifications to the inletport stub.
~ in.~thick)
and bronze-welded (not brazed) to the cylinder

barrel. The bore of the tube may then be blended into the
port so as to leave no step. A further sketch illustrates this
job. Alternatively a flange may be welded on.
The shape and position of their upper orifices at the
cylinder bore determine the "opening" and .. closing" of
the transfer ports, and the behaviour of the jets of mixture
issuing from them. In standard form they are not capable of
passing the greatly-increased quantities of mixture now
induced into the crankcase and it will be necessary to modify
them. This can be affected in two ways, (1) by greatly
increasing the width circumferentially, or (2) by moderate
increase of the whole cross-section 'of the ports, very
slightly altering the height of the cylinder orifices and
reshaping the width to a streamlined form. The latter
course is preferable.
Be careful not to leave a lip at the rear of the transfer port
exit; this will cause the incoming column of gas to swing to
the front of the cylinder, which is undesirable. It should
be noted that the orifices are offset to the rear of the bore
centre line for a distance of i\ in. This assists the new gas
223
TUNING FOR SPEED
to move toward the rear of the cylinder to displace the

residual rgases. One further point remains. The cylinder
cutaway at the spigot should be streamlined where it is
bevelled at the lower entrance to the port. Later "large
fin" barrels have transfer ports adequate in size.
Modified port dimensions. The reference

.. A" denotes the point of commencement of
swell.
iHAUiJ
'---;
.1
1", INLET
l, . . ___ ,
The timing of the exhaust port, as determined by its upper

boundary, presents a problem. If the port is opened considerably earlier, its closing is similarly delayed, with the
result that, when using a highly extractive exhaust system,
a proportion of the hard-won fresh mixture may follow the
exhaust gases. No doubt this does occur, anyway, at
224
TWO-STROKE ENGINES
certain engine revolutions, but it is not a characteristic

which should be encouraged. At the same time it is important that pressure, caused by the burning charge should
drop quickly, and the exhaust be expelled at th~ earliest
possible moment without the serious loss of fresh mixture
or thrust on the piston.
Therefore, it is suggested that the top edge of the port be
modified only by straightening out its curved outline and the
area of the port increased by radiusing the sides. The shape
to be attained and the dimensions concerned are as shown
in one of the sketches. The cleaning of the top edge will
alter the timing slightly. It is most important that the widths
shown are not exceeded; failure to follow this advice may
lead to broken piston rings.
It is now feasible to consider the crankcase and the modifications to be carried out in order to increase volumetric
efficiency. First apply a smear of engineer's blue around
the bottom of the transfer ports at the cylinder base and
lightly bolting the crankcase halves together, pull the cylinde;
down on the bolts so that a blue impression of the ports is
left on the crankcase.
Dismantle again and, tackling each crankcase half
sepa,ratc:ly, file back the aluminium until the cylinder port
outline IS reached. Do not overdo it. At the same time, the
_comers of the crankcase ports should be nicely radiused to
blend. Do not ma~e too large a radius as the cylinder spigot
has to be blended mto the crankcase radius later on. Now
bolt the cylinder on to each crankcase half in tum and finally
?lend the, whole por~ length, leaVing no steps or rapid
mcreases In cross-section. Use a small electric torch and a
Removal of the lip

of metal facilitates
gas-flow from the
tra1lsfer port.
225
TUNING FOR SPEED
dentist's mirror will assist vision. File back the cylinder cutaways and radius the edges to blend into the crankcase port
radii. Polish off finally with emery paper. A small additional
improvement may be made at this stage by opening out the
entrance to the main bearing oil-feed holes a trifle.
Where the last ounce of power is desired it is possible to
pad out the crankcase a little. Two aluminium discs should
be made to fit inside each crankcase half to reduce the
volume. These should be a little less than iF, in. in thickness
in order to clear the flywheels and they should be cut away
in the centres to permit main bearing replacement and also
at the top so as to coincide with the" step" in the crankcase.
Fix the plates to the crankcase walls by drilling in suitable
positions, tapping and using fs B.S.F. set-screws. The discs
should be drilled and countersunk; screws with countersunk
heads should be used. The heads should be filed flat if they
protrude in order to clear the flywheels. Coating the threads
with jointing compound will eliminate the chances of leakage
should the walls be penetrated when drilling, and also help
to secure the screws.
It may appear that these discs have little practical value.
Calculation of their volume will show that they account
for several cubic centimetres and an appreciable percentage
of the crankcase and under-piston volume. The general
effect will be to maintain higher crankcase pressures, possibly
increasing carburetter blowback at low engine speeds, but
assisting proper transference of the charge at useful racing
speeds.
Piston shape determines, with the port positions, the
timing of the engine, and perhaps the most important
dimensions are those which operate over the inlet tract. It
will be seen that the longer the rear skirt is, the later it will
uncover the inlet port when ascending and the sooner it will
close it when descending. Provision has already been made
for "longer" timings by bringing down the bottom of the
inlet port and a further improvement may be made by filing
away the bottom of the piston skirt at the rear. It can be
seen that, in a standard engine, the skirt does not clear the
226
TWO-STROKE ENGINES
Driving out the crankpin and parting the flywheels in order to get
access to the big-end
bearing is simplified by
Service Tool 61-3206.
port when the piston is at t.d.c. This factor in addition to

reducing the effective area of the port, also ca~ses turbulence
around the skirt, reducing the port efficiency.
If necessary, t~erefore, the skirt should be cut away i in.
at ~he rear ?r fo m. ~up to. the base ring of aluminium) for
racmg. ThIS alteratIOn WIll have the following effects: (1)
to open the P?rt earlier and close it later. (2) to eliminate the
port obstructIOn. (3) to allow a "dwell" period offull openmg of the port as the crankpin swings across its top arc.
NEW
Exhaust opens, b.b.d.c.

Exhaust closes, a.b.d.c.
Transfer ports open, b.b.d.c.
Transfer ports close, a.b.d.c.
Inlet port opens, a.b.d.c.
Inlet port closes, a.t.d.c.
75
75
65
65
115
65
227
OLD
70
70
60
60
120
60
INCREASED
PERIOD
10
10
10
TUNING FOR SPEED
The next operation is to build up the flywheel assembly,

piston and cylinder on one crankcase half and mark the
piston so that it may be radiused at each cut-away to blend
with the cylinder and crankcase ports which have already
received attention. Repeat on the other side and then bevel
the front of the piston skirt slightly and put a i in. radius on
the base of each cut-away corner. It is the writer's practice
to use second-hand pistons, suitably polished; they have
settled down and. therefore, have little tendency to seize.
Having described various methods of facilitating the
introduction of extra mixture into the combustion space, the
next aim must be to extract as much power from the charge
as possible, and here the question of compression ratios
arises. Generally speaking, the compression ratios may be
kept quite high without incurring risk of detonation or hot
running; standard ratios vary from 6.5 to 1 to 8.25 to 1.
For all purposes except racing it is suggested that the C.R.
be increased to about 9 to 1; this should produce speeds up
to 65 m.p.h. from a 125 c.c engine. For more serious purposes, or when petrol-benzole mixtures or 80-0ctane fuel
are used, the ratio may be stepped up to 11 to 1 to 13 to 1
with confidence. In modifying a "Bantam" to a compression
ratio of 11.25 to 1, the capacity of the head should be 15 c.c.
and this takes into account a raised piston-<:rown volume
of 3 c.c. i.e.:
_ 123
12 _ 135 _ 11 25
1
C.R. 12
- 12 . to
The head should then be drilled through the centre to

accommodate the plug and turned out to the correct size for
a 14 mm. tapping later on. The head should then be reversed
and the fins carefully turned away so that a 14 mm. plug and
spanner can be inserted. This turning should be deep enough
to provide a tapped hole just sufficiently long to allow the
sparking plug (with washer) to reach the inside of the
combustion chamber. A centrally-disposed plug increases
maximum speed slightly and does not oil. With a C.R. of .
6.5 to 9 to 1, a Lodge HNP plug gives a good result; from
228
TWO-STROKE ENGINES
A sectional impression of a modified

high-compression
cylinder-head
showing filling
material machined
to the required
hemisphere.
11 to 13 to 1 a Lodge R47, or equivalents, in my experience,

are satisfactory.
A further factor related to the combustion phase is the
length of time the charge is burning prior to release through
the exhaust port. This time can be varied only by selecting
the moment of ignition. If the standard C.R. is retained,
there is no need to alter ignition timing; experience has
shown that standard ignition setting is satisfactory with a
C.R. of 8 to 1, especially if fuel of high octane value is
employed.
Any further C.R. increase may call for slight retarding of
the ignition because the higher ratio and gas temperature
at the normal ignition point, cause the flame to spread at a
quicker rate. The resulting knocking, or pinking, is not
always readily apparent in a two-stroke engine and it may
occur "silently" and unnoticed.
One cannot lay down hard and fast rules about a suitable
degree of spark retardation, for a great deal depends on the
volumetric efficiency of the engine and the fuel. For racing,
with C.R.s of 11 to 13 to 1 an ignition timing of 21 to 19
b.t.d.c. is worth trying. This may be achieved by cutting
away the contact-breaker housing in the region of the three
slots and clamping down by means of short distance pieces
and washers and the original screws. The breaker points
should separate with the piston at .110 to .096 in. b.t.d.c.
according to the 21 or 19 advance setting required. It is
vital to check the contact-breaker gap first.
Earlier it was arranged that the exhaust port and the
229
TUNING FOR SPEED
TWO-STROKE ENGINES
exhaust stub should have increased cross-sectional area in

order to pass the exhaust gases but it is necessary, also, to
arrange the exhaust system so that gas-flow is unobstructed,
and so that the lowest possible pressure-even a negativemay prevail in the pipe at the moment when the port is uncovered by the piston. This can be achieved by experiments
with varying lengths of exhaust pipe and by the use of the
well-known megaphone component.
Experiments have been carried out with a 12-in. long
megaphone and a 12 pipe divergence, the exhaust pipe
being shortened gradually. Sometimes a length which gave
good power at high revolutions introduced a "miss", or
flat spot, in the carburation at other speeds. A fairly sound
scheme is to have an exhaust pipe which terminates at the
"Bantam" front chain-cover studs, or up to 3 in. shorter
if a megaphone is attached.
If the big-end is in good order, there is not much more to
do. An improvement to the lubrication of the "Bantam"
big-end is possible, however, by belling out the slots of the
connecting-rod eye on the outside. This can be done by
running the lip of a grinding wheel around the edges. The
1953-type rod is cut away at the big-end faces to achieve the
same result. This, with the more liberal oil supply coming
from the new, larger jets, results in bearing reliability.
Attention is drawn to the fact that the" Bantam" 1953 bigend bearing is !- in. wider than its predecessors and, therefore, has a better load capacity.
An improvement to magneto-side main-bearing lubrication can be carried out by drilling the duralumin distance
coilar, which fits against the bearing, so that the holes line
up with the balls in the bearing. Ten i in. holes will do. If,
when reassembling the crankshaft and case, there is a sense
of stiffness, or a tight spot, suspect your initial inspection of
the bearings. New bearings will be slightly stiffer in rotational
movement. Or possibly the shafts are not in line; they should
be tested in the usual way between centres. True them by
tapping-before the covers are put in position-and re-test.
Obviously, gearbox bearing and bushes should be
checked; the splines and dogs should be inspected because
Belling out the oil slot

in the connecting-rod
big-end eye of pre1953 model "Bantams", tends to improve lubrication at
this highly stressed
point.
230
Special tool No. 613191, in conjunction

with a screwdril'er
facilitates the dismantling of the "Bantam" clutch-retaining
cir-clip and main
clutch assembly.
the increased power it is hoped to attain may reveal weakness

in the transmission component. Irrespective of condition
the fol!owing improvement is suggested. The second-gea;
notch III the selector should be deepened to a i\ in. "V"
and ~he selector spring shimmed to a depth of i in. in its
hOUSIng. The gear selector-arm retaining spring, located
by a cup and secured by a split-pin, should be reinforced
b~ adding a short spring inside. This spring should be about
i m. O.D. and care should be taken that the coils do not bind
at normal movement of the selector mechanism.
231
TUNING FOR SPEED
TWO-STROKE ENGINES
certain port-timing or C.R. data, hased on the T.T. "works"

~ntries of such-and-such a factory upon which to draw. It
lS all a matter of compromise founded on commonsenseand therein lies all the fascination of two-stroke tuning.
For removing and rep/acing the B.S.A.
"Bantam" flywheel
mogneto components
the spec/alfactory tool
No. 61-3188 is useful.
Alternatively. a heavier main spring may be fitted but this

"mod" must not prevent the change mechanism from
functioning properly. These improvements to a .. Bantam"
box are necessary to avoid the tendency for second gear to
be forced out of engagement due to the selector mechanism
side-thrust under racing loads.
The clutch on the average two-stroke unit is quite up
to its job and will perform well under racing conditions
without any increase being made in spring pressure.
Inspection should be made, of course, to ensure that the
friction inserts are not worn. Running under oil-bath con
ditions, these clutches are most reliable and heat-resistant.
The magneto flywheel is fitted with the engine installed;
care should be taken to see that it is well home on its taper
and tightly locked up. The Wico-Pacy instrument will give
very good service and the engine can be revved to well over
7,000 r.p.m. without misfiring.
Now it remains only to test your handiwork on the roadassuming you are not one of the select few with access to the
services of a dynamometer and skilled operating staff.
The result of your tests may inspire further modification to
any, or, indeed, all, of the components previously dealt with,
for the art of extracting speed from a two-stroke is one of
continual trial and error. There are no recorded and
232
233
CHAPTER XVII

that the frame of the machine is not suspected
of having been damaged or is not out of line in any way,
the next step is to assemble the engine into it; this is
usually a fairly straightforward matter calling for little
comment in itself.
If the design is such that the unit can be built up as a whole
on the bench and then dropped into the frame, that is the
best sequence to employ, because it pe~its t~e rear eng~ne
plates and bolts to be fitted without any Jug~l~g ?r forcmg
into line. Whatever the method, however, It IS Important
that those bolts and others which connect the plates to the
frame lugs really are a good fit in their respective holes, and
that all mating holes are accurately in line. Slack-fitting
bolts are very prone to work loose, however tight the nuts
may be, and if the holes are out of line to an extent, calling
for a considerable amount of levering (and perhaps a few
hammer blows) to work the bolts into position, peculiar
internal stresses are created, which accentuate any tendency
there may be for vibration to be set up at certain speeds.
Commercial engine bolts are generally made from mild
steel, but for serious racing something better is .worth while,
particularly in view of the fact that the bolts wIll have to be
ASSUMING
'11th
234
Waisting bolts down to a

diameter equal to that of
the bottom of the threads,
leaving a central paraUel
portion for registering purposes, or, if a bolt has a
solid head, drilling down the
centre makes for lightness
without sacrificing strength.
taken out and replaced several times in a season. For such

bolts, heat-treated K.E. 805 steel cannot be bettered; this
material is very strong and tough, and is not liable to strip
the threads if tightened by an over-zealous hand. A less
expensive grade, such as 3% nickel steel, is a good substitute.
A not inconsiderable saving in weight can be effected by
waisting the bolts down to a diameter equal to the bottom
diameter of the threads, leaving a short parallel portion at
each end to centralize the bolt or act as a register for the
engine plates if required; alternatively, if the bolts have
TABLE OF DRILL SIZES
BOLT SIZE
CORRECf I
HOLBl
DIAMETER
1/4 in. B.S.F. (26 T.P.I.)

5116 in. B.S.F. (22 T.P.I.)
5/16 in.x26 T.P.I.
3/8 in. B.S.F. (20 T.P.I.)
3/8 in. x26 T.P.I.
7/16 in. B.S.F. (18 T.P.I.)
7/16 in.x20 T.P.I.
7/16 in. x26 T.P.I.
1/2 in. B~S.F. (16 T.P.I.)
1/2 in. X 20 T.P.I.
1/2 in. x26 T.P.I.
0.151
0.185
0.174
0,208
0.185
0.239
0.228
0.202
0.271
0.244
0.217
i NEAREST
DRILL
SIZE
I!FRACfrONAL
i
SIZE
No.
No.
No.
No.
No.
25 I 9{64 in.
13
3/16 in.
17
11/64 in.
4
13/64 in.
13
3/16 in.
SizeB
15/64 in.
No. 1
7/32 in.
No. 7
13/64 in.
Size 1
17/64 in.
SizeC
1/4 in.
I
7/32 in.
7/32
in.
I
I
solidiheads they can be drilled down the centre, as shown in

the diagram, the hole (which must be accurately concentric
with the outside) extending very nearly to the beginning of
the thread. Far from weakening the bolts, this treatment
actually makes them less prone to fatigue-failure and, if
carried out consistently throughout the whole machine,
a perceptible mass of excess metal will be eliminated.
The Table above gives the right drill sizes to employ for
the varying diameters and thread pitches which are
commonly used for engine and frame bolts.
When making up those engine-securing bolts which also
235
TUNING FOR SPEED
Spigot nuts made to fit

enlarged engine-pJate
holes cwe wear and
obviate the risk of
vibration.
pass through the engine plates, the length of the thread

should be such that the platesbave at least half their thickness bearing on the full bolt diameter and not on the crests of
the thread. otherwise the latter will rapidly hammer down
and looseness will develop. Occasionally in old machines
some of the plate holes will be found to be oval or enlarged,
in which event they can either be reamed out to tin. oversize, and bushed to bring them back to original size, or else
special spigot nuts can be made up to fit the enlarged holes.
The latter is a good idea, particularly if the plates are thin,
because the bolt threads are then protected from further
damage from the action of the plates.
CRANK CAse
Whether or not a bolt, or stud, is waisted, be

care/ullo see that at least half the width of the
engine plates bear on the /ull boit diameter-not
wholly 'on the thread crests which will rapidly
wear.
236
With all engine and gearbox bolts tight, the chain alignment should be carefully checked. Two straight edges
should be used in preference to one, since this method shows
up any errors more clearly. The most common error is for
the ~prockets, though paraliel, to be displaced sideways
relative to each other; in some designs this may be caused
by clutch inserts of incorrect thickness, for which the remedy
is obvious, but in other patterns (such as the Burman in
which the insert thickness cannot influence the chain-~e)
the spacer-washers between the clutch body and the shoulde;
on the gearbox mainshaft may be of incorrect thickness.
Another cause met with when an engine-shaft shockabsorber is fitted, is wear on the inner face of the engines-,?rocket, but. in. any event this form of error is usually not
dIfficult to elimmate. Lack of parallelism is rare which is
fortunate, as it is more difficult both to trace the s~urce and
effect a cure.
If the engine plates are interchangeable a quick check
can be made by swopping them over, side for side, and if
~e er;or has yanished, or is still there but in the opposite
dlTectlOn. obViously the plates are at fault. If this course
cannot be a~opted, or yields negative results, the faces of the
box lugs may not be square to the mainshafts which can
be verified by bolting the box down on to a 'block on a
surface plate and checking the parallelism of the sprocket
with the plate.. Should an error thereby be brought to light,
the lug faces WIll have to be carefully filed up to eliminate it.
;\\'hen under load, the pull of the chain has a tendency to
den~ct the shafts towards each other and, consequently,
a sli.ght outward divergence of the shafts is not necessarily
detnmental. But if the shafts are inclined towards each other
when at rest, both sprockets will be badly out of line when
under load. It is not sufficient merely to check the alignment along the top run of the chain; the bottom run should
also be checked, and any noticeable difference between the
tw~ is evidence that the box is twisted relative to the engine.
!hlS c~n be caused by lack of flatness of the engine plates, or,
m deSIgns where the box is mounted on a bracket separated
237

CLUTCH
SPROCKET
HARDENED
m>."""r--- WASHERS
h~~
ALTERING THICKNESS
AT LEAST fr~ OF SPLINE

MUST BE IN ENGAGEMENT
Self-explanatory methods of (above) adjusting

chain line from the clutch end, alld (below)
making similar alteratioll to the ellgille sprocket
assembly.
CLEARANCE TO ENABLE CAM

TO MOVE OUTWARDS
TOWARDS NUT
WASHER ADDEO TO
MOVE SPROCKET OUTWARD
SHOCK ABSORBER 1X'~~~,.m.;
CAM SLIDING
ON SPLINES
from the engine, there is a possibility that the frame itself

is twisted.
t,~
Accuracy of alignment is particularly important at speeds
of 6,000 r.p.m. and over; for instance, on A.J.S. racers of
recent manufacture, provision is made to obtain almost
dead accuracy by the addition of shims behind the sprocket.
It is somewhat confusing, therefore, that speedway J.A.P.
engines are fitted with floating sprockets with about i-in.
laterial movement, but this is to allow for the flexure which
is inherent in the very light frames employed for this work.
So long as the chain does not come off the sprockets, it does
not matter much if its life-span is short.
Chain lubrication is vital for high speed; the best scheme
with an open chain is to adopt the Norton practice of feeding
mineral oil through a suitable jet to a divided pipe from which
the oil drips on to the side-plates just before the chain starts
to run on to the clutch sprocket. The aim is to get the
lubricant directed into the actual joints, as the rollers themselves need little lubrication externally.
When adjusting a primary chain in a layout employing
aluminium engine plates, or in a unit-construction plot,
allowance must be made for thermal expansion, by permitting about 1 in. up-and-down play; a little too much slack
does not harm a chain nearly so much as too little. The best
plan is to adjust a little on the tight side then to operate the
adjuster to bring the box slightly forward: the adjuster then
holds the box forward against the pull of the rear chain, and
there is no possibility of the box being shifted rearwardly.
SPROCKET INWARDS
OBSERVE POSITION OF THIS
SHOULDER AFTER BORING OUT
238
239
MAl NTENANCE OF RACING CARBURETTERS
CHAPTER
xvm
MAINTENANCE OF RACING CARBURETIERS

THE standard needle-jet Amal carburetter is a very
simple instrument which, owing to its "straight through"
construction, gives a higher power output, size for size,
than other types which make use of butterfly throttles.
From the standard pattern there have been developed a
number of special racing editions with the object of obtaining maximum power and acceleration throughout the whole
of the speed range. Of these the best known are the T.T.
type and the remote-needle or R.N. type. In addition,
non-needle models have been produced for track and dirt
racing, where engines are operating at full bore most of
the time.
In the T.T. pattern the needle-which meters the fuel at
certain throttle openings-is suspended from the top of the
throttle slide, and thus hangs ~cross the centre of the choke.
In the R.N. pattern-as its name implies-the needle is
suspended from an extension on the throttle slide and is
housed within a chamber cast on the side of the carburetter
body; thus the needle offers no obstruction to air-flow, and as a result the R.N.
type is slightly faster than the T.T.
pattern for any given choke size.
The more devious fuel passages in the R.N., however,

render it rather more sensitive at low engine revs., which
means that the T.T. pattern offers more foolproof carburation and is a little easier to tune than the R.N. The latter also
occupies more space (due to the presence of the remoteneedle chamber), which sometimes makes it difficult to
house.
The position ofthe taper

needle, which is "remote" from the main
intake tract, gives the
initials R.N. to this
type of racing carburetter. Being carried
in a separate chamber
off-set from the main
barrel ofthe instrumelll,
the needle offers no obstruction to the air-f/ow.
PRIMARY
CHOKE
The main choke, showing the small

primary choke protruding into tIre airstream. The block is of the standard
type but T.T. and R.N. types are
similar in principle.
240
In both types the diameter of the main choke, at the point

where the small primary choke protrudes upwards into the
airstream, is usually bigger than the bore of the instrument
immediately on the engine side of the throttle. This latter
and smaller diameter is actually the choke size and corres241
TUNING FOR SPEED
ponds to the figure stamped on the outside of the mixing.

chamber for identification.
The increase in diameter referred to is provided (a) to
minimize the 0 bstruction to air-flow caused by the uprush of
liquid fuel from the primary choke, and (b), in the case of
the T.T. type, to help overcome the obstruction caused
by the needle. Therefore, when discussing choke sizes,
it is important to remember that the choke size refers
to the diameter of the bore just past the throttle valve, and
not to the larger diameter at the point where the throttlevalve works. In 1 in. and 32 mm. carburetters, however,
this reduction in diameter is not present.
The normal R.N. type is essentially a petrol-benzole
instrument, and cannot be used with a main jet size larger
than 900, whereas simply by fitting a larger size of needle
jet, the T.T. type can be used with straight alcohol up to a
jet size of around 1500. The needle jet is formed by the
hole in the upper end of the brass component into the lower
end of which the main jet is screwed, and is varied for
different sizes of carburetters.
Up to I-in. choke diameter, the standard size is .107 in.,
and over I-in. diameter it is '.109 in. For alcohol fuel the
.107-in. jet must be replaced by one of .113-in. diameter,
while the .109-in. jet should be replaced by one of either
.1l3-in. or .120-in., according to the size of main jet which
the fuel necessitates. Incidentally, alcohol fuels vary considerably in their chemical make-up and some require a
much larger jet than others; one characteristic of these
fuels is their relative insensitivity to jet size as compared
with petrol, and excessively rich mixtures can be handled
without the loss of power which would be associated with
petrol.
A special type of R.N. carburetter has been evolved for
use with alcohol, primarily for use on the dirt; this has a
.160-in. needle jet, as opposed to the .109-in. size fitted in
the petrol-benzole edition. But other considerations are
involved, and if the two sorts of fuel are to be used, then
two R.N. carburetters will be required, whereas a single
242
REPRESENTATIVE CARBURETrER TyPES AND SElTINGS
CARBlJREI'nR
MODEL
----,------
_ _ _ _ _ _ _ _ _ i_~T~Y_PE_ _
A.J.S. 498 e.e.

A.J.S.7R
Ariel 350 c.c. "Red Hunter"
Ariel SOO c.c. "Red Hunter"
BoS.A. 350 c.e. "Gold Star"
B.S.A. 3S0 c.e. "Gold Star"
BoS.A. SOO c.e. "Gold Star"
B.S.A. 500 c.c. "Gold Star"
B.S.A. 500 e.c. "Gold Star"
B.S.A. 500 c.c. "Gold Star"
J.A.P. 500 c.c. Speedway
and 880 VTwin
Matchless G45
Norton 350 c.c. "Manx"
Norton 3S0 e.c. "Manx" 1946
Norton 350 c.c. "Manx" 1946
Norton 350 c.c. "Manx u
Norton 3S0 c.c. Short stroke
Norton 490 c.c. "Manx!)
Norton 499 c.c. Short stroke
Triumph Tloo
Vincent "Bluck Lightning"
Vincent "Black Lightning"
Vincent "Black Lightning"
Velocetto KIT
TIOTr9
lORN
ISIT38
IOIT38
JETS
---~--
CHOKE_~I
Ii. in. '12S0
1'>". in.
320
260
320
j"m.
It in.
~\~~
.1,,1, in.
I~gg
IOIT38
RN
TIOTr9
TIOIT9
27/013
1,. in.
113 in.
Ii. in.
It. in.
I in.
520
; 360
I 1700
i 960
10GP
IOIT38
IOITRNI
TlOIT
TlORN
TIOGP
TIOGP
10IT38
10ITRNI
TlORN
tORN9
TlOGP
SGP
76
TIOTr9
T10Tf9
TIOIT
TIOTT
t!i\ in.
l!i\ in.
ts; in.
l~j !n.
I:I.z m.
Ii in.
Ii. in.
t!'i'IJ in.
1:1\ in.
It\ in.
It'h in.
trill in.
1~~ in.
tin.
lih in.
t1'1, in.
32mm.
HI, in.
3S0
i
!
240
350
500
350
500
230
350
460
600
560
S60
260
310
190
360
I {m
FUEL
NEEDLE _ _
.120
.109
.107
.109
Ale.
Petrol
Petrol
Petro!
::ro
~~~ol
.109
.109
.109
.120
Petrol
Petrol
Petrol
Ale.
NIL
.109
.109
.109
.109
.109
.109
.109
.109
.109
.109
.109
.109
.109
.109
.109
.120
.120
.109
Ale.
Petrol
Petrol
Pelrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Petrol
Ale.
Ale.
Petrol
For various alcohol-blend fuels, the following percentage increase in

jet sizes over those fitted for petrol-benzole may be used as starting
points for tuning purposes.
FUEL
JET INCREASE
Discol P.M.S.2
Discol R.D.l
J.A.P. Racing Fuel
Straight Methanol
60%
80%
100%
150%
For an increase in altitude, the percentage decreases in jet sizes

over those used at sea level are 5 %at 3,000 ft., 9 % at 6,000 ft., 13 %at
9,000 ft., 17 % at 12,000 ft. It will rarely be necessary to alter the jet
more than one size for conditions as widely different as those found in
Holland and Switzerland.
243
TUNING FOR SPEED
On the T.T. pattern

carburetter it is the
air slide only which
operates in a separate
chamber.. the taper
needle is centrally disposed as in the standard illstrument.
T.T. type will suffice for both fuels with the requisite change
in needle and main jets.
To obviate this situation, and also to incorporate some
design features which render the carburetter less sensitive
to "megaphoning" and yet retain an unobstructed choke,
the G.P. type has been evolved. The salient feature of this
type is that the needle is suspended from the throttle-valve
so that it is just to one side of the choke, and the fuel enters
the choke through a spray-tube inclined at an angle instead
of directly underneath'as in the R.N. and T.T. types. The
passage from the needle jet to the end of the spray tube is,
therefore, quite short and direct, a most desirable ~eat~re.
To furnish a more precise tuning control, the metenng Jets
are supplied in two tapers but there are only five adjusting
grooves in each, instead of seven, and there is an additional
244
mixture control provided by an air jet, situated between

the air control slide and the needle-jet passage. This jet,
which can be unscrewed after removing the hexagon plug at
the base of the air slide, is made in two sizes, .100 in. and
.125 in bore and the appropriate size is fitted to each
carburetter by the manufacturers. Normally no alteration
is necessary but, if difficulty is found in obtaining absolutely
precise mixture control, it may be advantageous to alter
the size from that already fitted. Increasing the size weakens,
and decreasing it richens the mixture.
The main jets are interchangeable with those used in
R.N. and T.T. carburetters, but the needle jets, throttle
valves and needles are different. There are three basic
models in the G.P; series, the 15 G.P. in choke sizes from
i-in. to It\in, 10 G.P. from 1,.\ in. to 1~ in. and 5 G.P. from
liz in. to Ii in. which is considerably larger than that of any
R.N. or T.T. previously obtainable commercially. The two
smaller groups utilize similar needles, designated and marked
G.P. (standard) and G.P. 6 (weak) and are flange-mounted
with bolts at the regular 2 in. centre distance, but the 5 G.P.
range uses its own needles, marked 5 G.P. (standard) and
5 G.P. 6 (weak) and has flange bolts spaced at 65 mm. or
2.56 in. Hence, while all the other types are interchangeable,
sO"far as the flanges are concerned, the 5 G.P. cannot be
fitted directly unless the head has been drilled especially for
. it, as in the case of the short-stroke "Manx" Nortons.
G.P. instruments are fitted with long air-intakes and although
less bulky laterally than the R.N. types, may be difficult to
instal on account of extra length. In this connection it is
essential to place the mouth of the intake in a position where
carburation cannot be upset by stray air currents blowing
across or into it. Some odd effects can be created by, for
instance, the flat surface of an oil tank within an inch or so
of the intake; carburation trouble was prevalent on the
early A.J.S. "Porcupine" owing to the air whistling at high
velocity through the tunnel in the saddle tank in which the
carburetters were placed.
245
TUNING FOR SPEED
Float Chambers and Location
P.V.C. plastic tubing, making sure that it is not bent so

sharply that it may kink and shut off the flow. A bronze
spri~g pushed inside the tube will obviate kinking.
It IS preferable to place the float chamber fairly close to the
carburetter, but the hose must be long enough to absorb
engine vibration without undue flexing and possible damage
Both T.T. and R.N. carburetters use orthodox single

float chambers, with enlarged passages capable of passing
as great a flow of fuel as the double float chambers which
were popular some years ago, and which are still used at
times with alcohol fuel.
With twin floats, the fuel lines also can be duplicated and
there is less risk of complete fuel supply failure. But it
must be remembered that, if one side blocks, the other
cannot pass enough methanol to supply the engine and a
hole may be burnt in the piston crown very rapidly, this
being the inevitable outcome of running with a weak
mixture at ratios of more than 12 to 1.
Twin-jloat carburetters offer advantages where alcohol

fuel is used.
. ..-.,
Variation in fuel level,

caused, for example,
by cornering with a
sidecar, gives this
effect, and causes
"cutting-out. "
The latest Amal single-float chamber, type 302, as fitted

to.all G.P. carburetters can handle all the fuel required for
the largest jets ever used, and are also equipped with an
anti-frothing baffle. Frothing can be a puzzling reason for
loss of revs., owing to the rich mixture which it causes, and
to minimize this possibility, it is advisable to mount the
float chamber on adjacent part of the frame connecting it to
the jet block by a flexible petrol-proof hose, or a piece of
to the lining; the relative movement which occurs between

the cylinder head and frame on some machines is much
larger than might be imagined. .
One method of remote mounting is to clamp the bowl
bodily t.o an object such as the tank; another, and very neat
system IS to screw a rod into the chamber top or union nut
and hang the chamber from a flexible rubber mount bolted
to some convenient spot such as a tank bracket.
246
247
TUNING FOR SPEED
The float should not be placed appreciably to the rear of

the jet, as the fuel will tend to lag behind. during vi~Ient
acceleration, and this may cause the engme to heSItate
when opened up on the road. Somewhat the same effect
occurs when cornering sharply on a sidecar outfit, and can be
entirely eliminated only by using a double float chamber,
or by the addition of a" swill pot" (which resembles a small
float chamber minus the float, needle valve and petrol
union) placed' at the side of the mixing cha~ber farthest
from the main float chamber, the latter bemg remotely
mounted in the approved fashion. Under the action of
centrifugal force the fuel level drops in one chamber and
rises in the other; thus the level of fuel at the jet remains
approximately constant.
Carburetter Tuning
On page'243 is given a range of carburetters an~ approximate main-jet sizes which have been found sUItable on
various machines but it must be understood that any such
list can form oniy a very rough guide. So many varying
characteristics are involved that hard-and-fast rules cannot
be laid down and every racing engine must be treated as
an individual' unit when fin,al tuning is being performed.
Note that the main jet sizes for G.P. carburetters are very
.
. much smaller than for any of the other type?
The fuel used, the nature of the course and Its geographIcal
position all have their effects, which can be allowed for only
by ex~ence coupled with trial-anderror metho~s of .
obtaining the optimum settings. The exbaus7system 1:' use
also exercises a profound effect on carburatlon; as Will be
seen later, a larger jet is requir,:d with a me~ap?one ~an
with a straight-through pipe, whilst a smaller Jet 1S reqUIred
if any form of silencer is fitted.
.
The best method of determining the correct main jet IS
to drive the engine (after it is fully w~rmed up) "flat ~ut"
over a distance of at least a quarter-mde, and at the finIsh to
slam the throttle shut, cut the engine dead and immedi!t;tely
lift the clutch, thus avoiding any alteration to the sparking-
248
MAINTENANCB OF RACING CARIWRBTTERS
The tendet/cy mentioned above may

be corrected by
Jilting what is
knowil as a "swill
pot."
plug condition caused by the slow-running mixture or excess

oil. The sparking plug is then removed and examined.
If the end of the body is of a grey or lightish colour then
it is
the ~ixture is weak and a larger jet is required.
heavIly coated with soot, the jet is too big and a smaller one
is required. With a correct mixture, the end of the plug
should have the appearance of polished ebony; but, of
course. this condition will be achieved only if the plug was
clean (and preferably lightly polished) on the face before
fitting. "Reading" a plug is quite an art which takes a
little ~xperience to acquire. Also. modern plugs with
ceramIC Insulators are less easy to read than the now almost
obsolete mica patterns. The former exhibit a rather harsherlooking brown colour with a correct mixture instead of
the ebony of the mica type, but, in either case, any suspicion
o~ greyness on the electrodes or body is a sign of weak
mIxture and the jet ~ize must be increased. Some people
prefer to go up one SIZe further as a safety precaution if the
. race is on a course with long sections calling for fullthrottle work unless they are able to check the plug reading
on the actual circuit during practice.
On T.T., R.N. and G.P. types the air-lever-or more
accurately, the mixture control-operates an air-v;Zve on
the side of the carburetter, and so controls the depression
249
If
' .~ Ii
<1
.'.
:~
I
TUNING FOR SPE'ED
on the main jet via the primary choke tube; consequently,

this control can be used as a mixture-correcter without
reducing the effective bore of the carburetter. The effect on
mixture-strength between the fully open and fully shut
positions is approximately equal to three or four jet sizes. It
can therefore be used as a rough indication of correctness of
jet size; if the engine runs happily with the control shut, the
jet is too small, whereas if it sounds woolly or misfires
with the control fully open, or nearly so, the jet is too large.
This, however, is not in any sense a true test of mixture
correctness, for which the sparking-plug test is- the only
safe method. The real value of the air control is in its use
as a correcter for temporary conditions. In the Isle of Man,
for instance, it can be used for the first mile or so on the first
lap, when both the motor and air temperature are definitely
cold. To permit of such use the main jet should be of a
size which provides a correct mixture with the air control
fully open when the engine is hot.
Having 'satisfied oneself that the main jet. is correct,
attention must next be given to obtaining clean running and
acceleration throughout the entire throttle range. The three
factors which affect this are: (1) the pilot adjusting screw,
situated on the side ofthe mixing chamber; (2) the cut-away
of the throttle valve; (3) the adjustable taper needle.
The pilot adjuster screw on these racing instruments
controls the amount of fuel supplied to the engine for slow
funning; its operation must not be confused with the
standard-type pilot air-screw, which controls the amount of
air to the pilot jet. On the racing types, screwing the pilot
adjuster clockwise weakens the mixture, and, conversely,
screwing it out enriches it. The adjuster should be set as
rich as possible consistent with a good tick-over; the rate
at which the engine turns is controlled by adjusting the
throttle-cable stop to give the necessary small opening of
the throttle valve when the twist-grip is shut, but as the
position of the valve also affects the slow-running mixture
strength, the cable adjustment and the pilot adjuster
position are to some extent inter-related
250
61S
612
A diagrammatic explanation of the tuning

rule referred to in the text, With (lelt) a
straight:through exhaust pipe, the jet
needle IS fitted higher and the throttleslide cut-away is greater thall when (right)
a megaphone exhaust is employed,
Eventually, a setting for both can be arrived at which

gives a reasonably good tick-over, erring, if anything, on the
fast side, and with the mixture strength slightly rich. If
pains are taken to obtain a really super tick-over it is sometimes .found that snap acceleration suffers; ciearly, it is
essentIal that the engine should never fail to respond
instantly if the throttle is slammed open.
The throttle valve cut-away (i.e., the bevel milled on the
sli~e on the side farthest from the engine) controls the
mIxture strength from low openings up to about halfthrottle. The cut-aways vary in steps of
in. in height;
the No.5, for instance, has fir-in. cutaway, and consequently allows more air to enter than a No.3, which, as its
number indicates, has a *-in. cut-away. Generally speaking,
the lower the cut-away can be kept, consistent with clean
running, the better will be the snap acceleration, and this
should be borne in mind when tuning. It is advisable,
therefore, to have on hand a range of slides varying from,
say, No.2 to No.6, and to select from them the one which
is the most suitable.
These racing slides are very delicate, and a close fit in the
body; if they are handled roughly or assembled in a diTty
251
nr
TUNING FOR SPEED
condition, binding may occur, and it may be impossible to

get even running at small throttle openings.
Position of the Jet Needle
The taper needle is adjustable to five or seven positions in
the valve by a spring clip; the lower the needle the weaker the
mixture, and the higher the richer, for any particular throttle
position. Any flatness or tendency for the engine to eightstroke at over half-throttle can always be eliminated by
needle adjustment. A very common position on both T.T.
and R.N. types is 4 : that is, the fourth groove down from the
top; the final setting should, if anything, be on the rich side.
By systematic attention to the adjustments described, a
setting can be obtained which furnishes clean and lively
running from tick-over right up to full-throttle, particularly
when the engine is being used with a straight-through
exhaust system.
As is well known, the fitting of a megaphone exhaust
affects the running of the engine adversely at low and
medium speeds, a disinclination to run at anything below
3,000 r.p.m. and a tendency to cut right out when the throttle
is opened rapidly from low speed being typical symptoms.
Almost invariably the carburetter is blamed unfairly for these
troubles-a change over to a straight-through pipe will
prove otherwise--but it is necessary to alter the carburetter
tuning in order to minimize the ill-effects as much as possible.
A megaphone exhaust system usually requires a main jet
about 40 c.c. larger than that used with a straight pipe-in,
cidentally, Amal jets are calibrated in c.c.'s., the numbers
stamped on them indicating how much fuel the jets will
pass under a given standard set of conditions, although not
necessarily in the carburetter. To counteract the tendency
to cut out, it is necessary to fit a throttle-valve with greatly
reduced cut-away. Very often a No.2 slide has to be used.
but whilst this helps appreciably in getting the.engine to
open up from low r.p.m., it may introduce a rich spot at
small throttle openings. To attempt to balance out this
252
excessively rich setting it is usually necessary to employ a

very weak needle setting; for instance, in conjunction with
a No.2 slide one may have to lower the needle to position 1,
i.e. with the clip in the topmost groove.
Unless the increase in lap speed gained by fitting a megaphone really warrants its use, the carburetter people advocate straight-through pipes every time; but if a megaphone
has to be used the foregoing instructions should be followed,
otherwise a chaotic setting will result.
Racing carburetters are made in brass or light-metal, the
latter being most usual. When used with alcohol, a precipitate in the form of white powder gradually accumulates; thus it is very necessary frequently to dismantle and
clean the whole instrument, paying particular attention to
the small fuel passages and preferably blowing them clear
with compressed air.
Alcohol fuel is prone to leave a deposit of jelly which
can block the fuel flow completely, and it should be a firm
rule to drain the carburetter by loosening the jet-plug after
every race meeting.
Throttle slides in light-alloy carburetters are very sensitive
to grit and before a slide is pulled right out, the interior
of the mixing chamber must be cleaned, for if a grain of
sand does get in and jam the slide, the only safe course will
be to dismantle the whole thing, including removing the
jet-block from the body, in order to remove the offending
particle from the narrow annular gap in which the throttle
valve works. For very dusty conditions brass bodies are
superior to those made of light alloy, but, in either case
efficient air filters should be made up, not only to prevent
premature wear, but as a safety precaution to obviate a
jammed throttle. To be fully effective, this filter or a small
auxiliary one, must also cover the slot in the side air-control.
On light metal carburetters the threads of screwed parts
tend to seize rather easily, and any brute force exerted will
result in a scrapped component. The correct treatment,
when partial seizure is detected, is to dose the thread with
penetrating oil, leave it for a few minutes and then carefully
253
TUNING FOR SPEED
ease the component out in a series of quarter-turns. Before

replacement, the threads should be smeared lightly with
graphite grease. To conclude these remarks on racing
carburetters, it cannot be emphasized too strongly that they
are precision instruments which will not perform properly
unless well cleaned and nicely adjusted; it is, unfortunately,
a common sight at race meetings to see riders putting jets,
throttle slides, etc., down on a dusty road, forgetful of the
fact that any dirt or grit picked up may, a few minutes later,
be the cause of trouble in the race.
the standard models is screwed in, not out, to enrich the

mixture; as a general rule a setting of one to one and a half
turns "out" is about right. The tick-over speed is first
set by means of th.e throttle-stop on the side of the mixing
chamber, after whICh the throttle cable is adjusted so that
the engine responds immediately to grip-movement. It is
surprising how much a well-adjusted and smooth-acting
throttle control helps in improving the general" feel" of
the carburation.,
The standard air control differs from the racing type in
that, except when fully open, the air slide obstructs the airflow, and thus simultaneously enriches the mixture and
throttles the engine. It is, therefore, not of the same value
a,s a mixturc-c,orrector, but it can be used as a guide to jetsIze when tUlllng. If the speed increases when the lever is
closed slightly, the jet is too small; if the speed falls off. the
'
jet is either correct or too large.
Here" again, the appearance of the sparking plug is the
only rehable guide to correct jet size, but when conducting
plug tests it is essential (unless a magneto cut-out is fitted)
to have the carburettcr so adjusted that the engine stops
i'lstant.'y whe,n the throttle is snapped shut; even a few
explOSIOns WIth the engine idling are sufficient to mask
the true appearance of the plug.
Plug tests must be conducted with whatever exhaust
system is to be used when racing; the' removal of even an
effi,cient s,ilen.cer usua.uy necessitates an increase in jet size, a
POInt whIch IS sometimes overlooked to the great detriment
of the engine.
.
Tuning a Standard Carburctter

In the absence of one of the racing carburetters described,
a standard model can be made to perform quite well if it
is sufficiently large in the bore. Most sports engines will
accept a choke size slightly larger than that normally fitted
for touring use, particularly if a lot of work has been done to
improve the breathing ability of the engine. (An indication
of the sizes which can be employed was given in Chapter II
of this book.) On the other hand, it is easy to make the
mistake of fitting too large a carburetter; this will result in
an engine which can only scream around at high revs.,
lacking bottom-end power and acceleration.
A small amount of hand-work in smoothing and polishing
the bore and carefully blending-in the junction of the airintake with the carburetter body will improve the air-flow,
but care must be taken to see that the tiny slow-running
mixture hole (drilled at an angle into the bore immediately
on the engi~ side of the throttle) is not blocked in the
process. As with the racing instruments, it also. pays to
dismantle the mixing and float chambers completely to
make sure that the small fuel and air passages which control
the idling mixture are fnie from dirt or sediment.
Basically, the principles of operation and tuning are
similar for racing and touring models, thus the tuning sequence already described holds good for the latter. It is
important to remember, however, that the pilot adjuster on
Standa~d touring Amals are not suitable as they stand

for use WIth alcoh~I blend,S, for none of the passages is large
e?ough to cop: WIth the Increased fuel flow, but it is quite
SImple to modIfy them to use any blend containing up to
60% alcohol such as Shell "X", "Y", or "M". Fuels such
as 8 I 1 or '.' A ", containing much methanol, are rather beyond
the capacIty of these instruments.
254
255
Tuning for Alcohol Blends
TUNING FOR SPEED
The first step in conversion is to enlarge the area of every

fuel passage from the connection to the fioat-bowl right up
to the jet. Double the number of drilled holes in the banjo
connection and remove all sharp corners of the holes.
Bottom-feed, fioat chambers need the fioat-needle hole
enlarged to a size only .030 in. smaller than the head of the
fioat needle; drill out the hole in which the head is housed
about .030 in. larger. Careful lapping-in of the head and
seat may be necessary after this operation.
.
Enlarge the holes leading from the needle valve seatmg
into the fioat chamber, being very careful not to damage the
seat, as there is not much room in this region. Remove the
small hexagon screw in the banjo end of the fioat bowl and
open out the hole right through about .005 in. larger, which
will still leave enough thread for the screw, and double-up
the number of holes in the jet-plug. A good check on the
efficacy of the work can be made by rigging up a pint tin
with a piece of hose to couple up to the fioat bowl, and
timing the interval required to pass a pint of fuel: it should
eventually be reduced to less than 40 seconds.
With top-feed bowls, the rate of fiow is limited by the
feed hole which obviously must be smaller than the needle.
It can be enlarged a little, butthe best scheme is to make up a
large head for the needle and drill the needle seating to suit.
The next step is to remove the needle jet and jet block
from the mixing chamber. Enlarge the idling jet, which is
the tiny hole at the bottom of a larger hole in the engine side
of the jet block, to about twice its present diameter and drill
out the needle-jet to .113 in. for" X" or "Y" fuels and .118
in. or .120 in. for" M"; it may be necessary to experiment
later on with these sizes if tuning proves difficult.
Main jets of the size required are not very readily obtainable and, although the makers hate the idea, the only
scheme is to drill out some standard jets, the size for" M"
being in the region of .068 in. (No. 51). For this work, a set
of number size drills (ranging from No. 48 to No. 55) is
necessary, and a set of jet reamers (whicb can be obtained
from suppliers of tools to the watchmaking trade) is also
256
advisable. The trouble with drilling is that jets opened out

with the same drill do not necessarily pass the same quantity
of fuel owing to variation in the diameter, length, and
roughness of the hole and the only course open is to make
up a number of jets and go through the tuning sequence
previously described, starting first with the largest sizes to
avoid premature damage to the engine.
Other factors, however, may be limiting the flow of fuel,
and the precautions to be observed if the engine does not
respond to variation in jet size are dealt with in the chapter
on Testing the Finished Work.
Multi -cylinder Carburation
Although the basic tuning sequence is the same as for a

single, carburation can be somewhat more difficult on
multi-cylinder engines, and their special problems are dealt
with in Chapter XV.
257
THE correct selection of sparking plugs is of the utmost

importance in high-speed work; the time has long
passed when one could stop for a plug change and still have
sporting chance of winning. Prior to 1939, special racing
plugs were produced which were able to stand up to continuous full throttle with unfailing regularity, and in most
cases they were of the non-detachable type, with mica
insulation. Then" hardness," or ability to withstand heat,
was conferred by the use of heavy, high-conductivity, central
electrodes, by reducing the internal area of the plug body
and by minimizing the amount of insulation exposed to
flame round the central electrode. The insulation was also
protected to some extent by almost closing the end of the
plug body, and the path by which heat could escape from the
central electrode to the plug boss was kept as small as
possible.
Unfortunately, all such measures designed to keep the
plug cool also have an adverse effect in that they make the
plug much more sensitive to oil, which may either lodge
between the points, and so prevent the passage of a spark,
or may give rise to a conducting coating on the insulation,
through which the current may leak in preference to jumping
the gap.
Speaking very broadly, a hot-running engine is not usually
oily, but there are always occasions, such as when starting
or at the end of a long downhill stretch, when the engine is
not really hot, and it is at these times that the susceptibility
of the plug to oil is an important factor.
Each plug maker produces a range of plugs which can
cover almost all known engine types and, in addition, supplies
what are known as warming-up plugs, for lISe when riding
to and from the start or when not actually racing. These

types are usually detachable, and so can be taken apart for
cleaning, whilst the non-detachable racing versions can only
be properly cleaned by the makers. Conversely, long-reach
plugs should never be used in short reach heads; even the
expedient of using several plug-washers to shorten. the
effective length is bad, because they are seldom gas-11ght
and, besides causing loss of power, gas-leakage heats the
plug excessively.
Just prior to 1939 plugs with sintered aluminium-oxide
insulators, such as Sintox, Corundite and Sillimanite, made
their appearance and are now here to stay. This is not the
place to enter into a discussion as to the merits of mica or
the newer type of insulation, except to say that the former
cannot be used for any length of time with fuels containing
much tetra-ethyl lead; the lead salts formed during combustion react with the mica to form a conducting coating
which it is almost impossible to remove. Apart from this,
it may be taken that, provided the selection of type is correct,
any of the known makes of plug will function satisfactorily.
However, although various grades are supplied in each make
none are directly comparable with each other in their characteristics, and it may be found that a change of make is beneficial (or otherwise) merely because the heat value of one
type is intermediate between two types of another make.
There may be a few engines still using 18 mm. plugs, but
the 14 mm. size is at the moment almost universal; a description of the method of conversion from 18 to 14 mm.
was given in Chapter T. In addition, 14 mm. plugs are
available with two lengths of thread, known as " shortreach" and "long-reach," the latter being generally
employed in engines having aluminium heads. Short-reach
plugs should not on any account be used in long-reach heads,
except as a purely temporary measure, otherwise the exposed
threads in the hole become filled with carbon. When next the
plug is changed there is every chance, then, of the threads
becoming seized, and it may be very difficult to extract the
plug without ruining the thread in the hole.
258
259
CHAPTER X1X
" HARD" AND "SOFT" TYPES OF

SPARKING PLUG
TUNING FOR SPEED
The accompanying tables give 14 mm. types made by

Lodge, Champion, and K.L.G., arranged in order of
hardness-i.e., those at the lower ends of the tables will
stand the most heat and the least oil; the grades of different
makes are, however, not necessarily equivalent. and the
pre-1939 grades are quoted only as a guide to assist those

who were familiar with those grades but not with the later
types.
The symbol denoting the plug-type indicates whether it is
of short or long reach, but when ordering it is essential to
quote the full symbol or to mention which length is required.
For warming-up. any of the softe<;t plugs in the ranges
quoted, i.e., those at the upper end of each group, are very
suitable, but one of the "hard" varieties of sports plugs, such
as the Lodge HN or HHN, can be used quite satisfactorily.
For actual racing, the correct plug is that which will stand
up to the engine under the conditions prevailing, which in
this sense embraces the length of race and severity of the
course, and the fuel in use. There is nothing to be gained by
using a harder plug than is necessary, and by so doing one
runs the risk of the plug failing through oiling-up halfway
through a race, when, perhaps, the piston rings are not as
good as they were when they started.
Courses where full throttle is held for long stretches will
call for harder plugs than short courses, such as that at
Cadwell; but the most careful choice has to be made for
long-distance races, such as the T.T., where it is not unusual
to hear engines missing or cutting-out entirely after shutting
off for Creg-ny-Baa corner. This happens less often to topflight riders who come down the Mountain really hard than
to those who start to ease back a trifle up near Kate's
Cottage. As an initial choice, a plug near the middle of the
range should be selected, such as K.L.G. 646 or 689, or
Lodge R49. It has been found, however, that, owing to their
construction, the latest Lodge plugs with Sintox insulators
have a very wide heat range, and even the hardest, R51 or
RL5I, will deal with a range of conditions which once had to
be covered by four types, BR48 to 54, in the old mica series.
On the other hand, when conducting mixture tests, as
described in the part dealing with carburation, these plugs
will appear, even with a correct mixture, to be running much
hotter than would be considered advisable with the mica
variety. This point must be borne in mind when tuning,
261
LODGE
REMARKS
SHORT REACHI":ONG REACH
R47
R49
R50
R51
R52
R53
RL47
RL49
RL50
RL51
RL52
~ Current production with Sintox

insulation.
I
! I
I'
jJ
CHAMPION
i
SHORT REACH, LONG REACH
REMARKS
-----1----1--------------i
LAlO
LAII
LAI4
LAI5
NA8
NAlO
NAI2
NA14
NA18
NAI9
cu.rrent production with Sillimanite

insulation.
K.L.G.
REMARKS
SHORT REACHjLONG REACH
690
646
689
731
875
690LR
646LR
'689LR
731LR
875LR
All mica insulation and exactly

similar to ore-I 939 grades. Available
also in "New Corundite" material in
the 250, 280, 290, 303, 310, 320, 340
FE series.
260
~I
TUNING FOR SPEED
otherwise too large a jet may be fitted in an attempt to attain

the correct appearance of the plug.
open ends protected, to prevent damage to the threads or

the ingress of dirt or fluff, which may get up inside the plug
and cause an internal short. Spare plugs should always have
fresh washers in place, the best form of washer being the
solid copper type, as this provides the best heat conductivity,
and does not pack down and allow the plug to loosen, as
either the copper-and-asbestos or rolled-copper varieties can
do; a loose plug will overheat very rapidly.
Racing plugs are invariably supplied with the correct gap,
and in many instances the gap cannot be altered; in others
this is not the case, but if the gap has to be set, it must be
done by bending only the earth electrode, for attempting to
bend the central electrode is almost certain to damage the
insulation. Non-detachable racing plugs are not easy to
clean: the best method, if convenient, is to return them to
the factory for reconditioning, which is done for a moderate
fee. They can also be cleaned by sand-blasting in an ordinary garage plug cleaner, using a low air pressure to avoid
damage to the surface of the insulator and subsequently
taking particular care to see that every vestige of sand is
removed from the thread and from the small annulus between
insulator and body.
For most work, there is nothing to beat the K.L.G. springwire clip for retaining the high-tension lead, but when
excessively wet conditions are the rule it is better to use one
of the proprietary types of umbrella terminal which protect
the plug from water or mud.
Another point is the plug spanner; it pays to obtain a
really good example, which fits the hexagons properly and
with the shank bent to the minimum amount necessary to
clear the adjacent head fins. A badly fitting spanner can
easily cant over in use, until it bears on the upper end of the
plu.g, and may then either bend or crack the insulated core.
K.L.G.
(Ceramic S.749)
K.L.G.
(Mica.)
LODGE.
CHAMPIOt'.
----1-----1----
F.320
F.E.320
"HARD"
F.340
FE. 340
875
875LR
I
I
I
.
R.53
RL53
R52
RL52
LA15
NA19
731
731LR
R51
RL51
LA!4
NAI4
689
689LR
R50
RL50
F.280
FE.280
646
646LR
R49
RL49
LA1!
NAI2
F.250
FE.250
690
690LR
R47
RL47
LAlO
F.310
F:E.310
F.300
FE. 300
.MEDIUM
F.290
FE.290
'''SOFT''
F.220 (detachable)
FE.220 (detachable)
'''HARD''
T.280
'''MEDIUM''
T.260
"'SOFT"
T.240
Having settled by experiment on the most suitable plug, it

is not a bad scheme to carry, as the spare, another which is
one grade softer, for obviously if the hard one oils up, another
of similar grade is likely to do the same thing. Spare plugs
should be looked after very carefully both when in store and
'when being carried, and always kept with their threads and
262
263
CHAPTER XX

THE ideal method of testing an engine is, of course, on
a brake. With just a small Heenan and Froude brake,
type DPXl for preference, an accurate rev.-counter and an
Amal flowmeter, much information can be collected, and the
effects of varying induction and exhaust pipe lengths,
ignition and valve timing and compression ratios can be
studied and tabulated. Naturally, to do this sort of job
thoroughly takes a fair amount oftime, and is rather beyond
the scope of this book, but apart from such development
work, a brake is invaluable simply for determining whether
an engine is, or is not, developing its correct horse-power,
and for finding out its peak r.p.m.-that is to say, the
speed at which it gives off its maximum power, not necessarily maximum speed at which it wiII turn over.
Very briefly, the method is to run the engine against a
light brake load until it is thoroughly warmed up, and during
this period the lubrication system can be checked to ensure
that oil is reaching everywhere it should. Then the throttle
is opened up and the load on the brake increased until the
engine is pulling the maximum load it can at a medium
speed of, say, 3,000 r.p.m. (most racing engines do not like
. full-throttle at anything below this speed).
Readings are then taken of the r.p.m., brake load, and
fuel consumption, and by means of the air control an indication can be obtained as to the correctness of the jet size.
Obviously, if closing the air-control causes the speed to rise,
the jet is too small and should be increased before proceeding
further, as much damage can be done by running for long
on a lean mixture.
Likewise, the ignition setting can be checked by operating
the ignition control: if the speed rises as the spark is retarded
264
the advance is too great, but if it falls there is insufficient

advance. The ideal setting for bench-testing is one which
gives maximum power when the lever is pulled back just
a trifle, as then you can always check the effect of giving a
Uttle extra advance at any time. It is, however, not a good
plan to run continuously with the magneto considerably
retarded, as the spark then being delivered is not of maximum intensity; it is wiser to stop the engine and re-set the
timing to the correct figure. Incidentally, when observing
the effects of adjustments to ignition or mixture, it is better
to watch the pointer on the load dial of the brake rather than
the tachometer, because the former is far more sensitive to
a slight variation in speed than the latter.
The brake does not give a direct reading of the horsepower; this has to be calculated, but the arithmetic involved
is very simple. All one has to do is multiply the load in
pounds shown on the dial by the r.p.m., and divide by a
fixed figure, called the "constant," which depends upon the
type of brake in use; on modern DPX1 brakes the constant
i.s 4,500. Put into the shape of a formula:
R.P.M. x load
P
BH .....
4,500
As an example, if the engine is pulling 28 lb. at 5,000
. IS
. then d eve1opmg
. 5,000
x 28 311 b h.
r.p.m., It
4,500
... p.
It will be seen that the accuracy of the b.h. p. figure depends

upon the accuracy of the rev. counter; should this be at
fault, the whole business is worse than useless, it is most
misleading, and it is easy enough to. delude oneself into
believing an optimistic result without the added complication of inaccurate instruments. Consequently it is best
always to have at least two separately driven rev-counters,
one which, preferably, w~ll.be used on the motorcycle itself.
This serves as a check on the proceedings and avoids subsequent discrepancies in readings when the model is tested
on the road. A simple method of checking the speed indication stroboscopically consists of painting a white mark on
K
265
TUNfNG FOR SPEED
the drive coupling, and observing this solely by the light

from a neon lamp. In England and all countries using
50-cycle mains frequency, the mark will appear stationary
at 3,000 and 6,000 r.p.m. while in the U.S.A., Canada and
all countries with 60-cycle current, the stationary speeds
will be 3,600 and 7,200 r.p.m. There are, of course, more
accurate methods of checking such as by Strobotac or the
Ashdown Rotoscope, and these devices can also be used for
observing the behaviour of valve springs etc.
By manipulation of the brake, the horse-power developed
at various speeds at intervals of, say, 500 r.p.m. can be
determined and from the figures so obtained the power
curve can be plotted on graph paper. Readings should
always be taken after the engine has settled down to a steady
speed; "snap" readings can be deceptive and are usually
optimistic. It may be found that the ignition timing and jet
size which are correct at low speeds are not necessarily
correct at high r.p.m., and if so, it is best to re-set them to
suit the latter condition. The effect of varying the induction
and exhaust pipe lengths can be determined very easily, and
sometimes quite surprising results can be achieved by
obtaining just the right combination. Sometimes, of course,
such a combination cannot be used on the eventual installation, either for structural reasons or because of regulatio~s
which restrict or define the layout of the exhaust system, In
which case the next best combination which does comply
with the rules must be sought. In this connection, builders
of 500 ,c.c. racing cars are somewhat better off than motorcyclists, because they have room to use almost any conceivable arrangements, and are not restricted by regulations
regarding pipe lengths.
'
Usually a gain in power at the upper end of the speed
range, where it is mostly required, is accompanied by a drop
in power somewhere lower dowI!, although this is not of
much consequence on a racing motorcycle, where the revs.
are always kept up by use of a close-ratio gearbox. There
are, however, conditions where it may be better to sacrifice a
tUttle at the top end in order to get more power low down, or
266
to eliminate a flat spot in the power curve; for instance, on

dirt-track machines, which are single-geared, it is essential
to have plenty of power available to give traction and keep
the rear wheel spinning when coming out of the corners, and,
therefore, good pulling power at around 4,000 r.p.m. is a
vital necessity. Then, again, the comparatively great weight
of a 500 c.c. motorcar demands wider gear ratios, particularly for standing-start hill climbs, and if the engine
torque is inclined to fall off rather badly with a diminution
in speed, the acceleration following a change to a higher ratio
may suffer.
Mention was made earlier of taking fuel consumption
readings on a flowmeter, of which the Amal pattern is the
one most frequently used. This instrument indicates the
rate, in pints per hour, at which the engine is consuming
fuel and from this the " specific consumption" can be found
by dividing the consumption by the power being developed.
For short-distance racing, a lavish consumption does not
matter greatly, but if the specific consumption, when using
petrol, works out at much more than .8 pint per b.h.p. per
hour, there is something wrong somewhere. Probably the jet
is too large; but if reducing the jet also reduces the power or
causes overheating of the plug, then either the valve timing
is not correct, the exhaust system is of the wrong length. or
the compression ratio is too high necessitating a large flow
of fuel to keep the internal temperature down.
The flowmeter may show also that an increase in power
gained by some alteration has been accompanied by a disproportionately great increase in consumption, this being
particularly liable to happen with two strokes. That may be
a serious matter for long-distance work, as the increase in
performance may, or may not, be nullified by the necessity
for carrying more fuel, or perhaps making an extra pit
stop. In such cases the alternative settings should be carefully noted, so that subsequent experiments can be conducted in practice on the course, to determine which is the
better of the two over the distance of the race.
Arguments frequently arise as to the power which various
267
TUNING FOR SPBED
engines can or have developed. Few motorcycle engines

are subjected to officially observed tests as aero engines are,
and' such arguments are likely to continue. It may be said,
however, that-anything over 45 b.h.p. for a 500 C.c. engine
and proportionately less for other capacities, is very good
indeed. Atmospheric conditions exert quite a large effect
on power, which increases when the barometer reading is
high, or when the air temperature is low, and vice versa.
For that reason, it is always advisable, particularly in a
changeable climate, to correct all power readings to N.T.P.
-i.e. normal temperature and pressure, which are respectively internationally fixed at 150 C. and 29.92 in.
mercury. Correction is made by the following formula:
Corrected B.H.P.=
Observed B.H.P. x 29.92 x 400 ... air temp. (degreefl C.)
barometer reading
415
of speed provided the factors stated do not vary. Air resistance depends upon the frontal area and the aerodynamic
shape, and increases as the square of the speed, and, therefore, the power absorbed in overcoming it increases as the
cube of the speed, whereas that required to overcome rolling
resistance increases only directly as the speed. In consequence, on a good surface, the power absorbed by wind
pressure at speeds in the region of 100 m.p.h. is nearly 10
times that absorbed by rolling resistance. hence the necessity
for adopting the most compact riding position possible.
The air resistance can be found from the formula:
From this it will be seen that a drop of 1 in. in the barometer reading lowers the horsepower by approximately
3%, and although the effect of temperature change is
less important Over the ranges normally met with in an
enclosed room with the engine running, it is certainly
advisable to correct the readings fully to avoid subsequent
confusion.
Other figures, such as the torque of the engine and its
B.M.E.P. can also be obtained from the brake test, and for
those who are interested the relevant formulae are included
in an appendix.l
A, the frontal area, is difficult to determine accurately for

such an irregular shape as a motorcycle and rider, but it can
be taken as being about 5 sq. ft. for'a rider of the average
physique, lying down on a Senior T.T. machine; it would
obviously be less, perhaps
sq. ft. for a small man on a
125 C.c. model. For the factor C, a figure of .0008 has been
found by experiment to give results which tally reasonably
well with known performance figures, though it may be
slightly less for a clean design of machine with the rider
exceptionally well tucked in.
Using these figuJ;'es. the air resistance to a Senior T.T.
model at 124 m.p.h. (equal to 182 ft. per sec.) is
Power Requirements
The power required to propel a vehicle depends upon two
main factors-rolling resistance and air resistance. The
former depends upon the gross vehicle weight, the state of
the road surface and the size and inflation pressure of the
tyres, but it remains substantially constant over a wide range
1 For those who are really interested in this work, A. W. Judge's
book The Testing ofHigh-speed Ihternal-combustion Engilles (Chapman
and Hall Ltd.} provides a wealth of information.
268
Resistance (pounds)=A.V2.C.
where A = the frontal area in square feet.
V = velocity in feet per second.
C = a constant called the" drag coefficient."
4,
5 x 1822 X .0008=1321b.
Taking the total all-up weight as 520 lb. and rolling resistance at 2 per cent, which is a good average figure to
employ. then the rolling resistance is approximately 10 lb..
a very small figure compared to the air resistance. The total
resistance is therefore 132 plus 10=142 lb.
The horsepower required to overcome a known resistance
at a known speed is found from the expression
269
TUNING FOR SPEED
H .P .
= resistance x feet per second

70
&"
.
H .P. 142550
x 182 = 47
t heretore,
m thi S Instance,
This is the power required at the rear tyre; assuming 96

per cent transmission efficiency; 49 B.H.P. would have to
be developed by the engine to achieve the quoted speed of
124 m.p.h. This result agrees very closely with the power
development and maximum speed of known racing
machines, particularly the 1938' Senior Norton, but there is
always bound to be a certain amount of guesswork in
estimating the factors involved. The drag coefficient" C "
for instance can be brought down to .0003 by full stream~
lining and though this usually results in a simultaneous
increase in frontal area, the net gain is very considerable.
, To avoid much laborious calculation, the accompanying
graph has been compiled from a number of known performance figures and gives a fairly accurate guide to power
requirements; the lower curve is for light, small-capacity
machines in track condition, the middle curve being for
heavier machines in road-racing trim, and the upper curve is
for large, heavy mounts in the big-twin class. These curves
do not appear to rise so rapidly as might be expected with
increase of speed, but this is because in the examples chosen
greater efforts had been put into reducing wind resistance by
streamlining, and rolling resistance, partly by reducing
weight, and partly by using special tyres. With the aid of
this graph and the performance figures of the engine an
estimate of the speed attainable and of the gear-ratio to
employ can be obtained which will be sufficiently close to
serve as a " jumping-off place" pending final settlement on
the course.
Most amateurs will, of course, have to do what they can
without the advantage of knowing the actual developed
horsepower, unless they can get their engines tested for
them. In any case, even if the figures. are available, final
tuning has eventually to be done under race conditions, as
270
A. -VERY LIGHT, STRIPPED MACHINES
550
B - ROADRACING MACHINES WEIGHING ABOUT 300 lb.
60 f - C -HEAVYWEIGHT MACHINES
50
/'
30
--
10
a.:
17
E:7V
,... ~ V
20
8/
/~
V
~
./
...---;:
:i
ai
60
70
MP~
eo
90
100
110
120
130
140
In graph form this approximate guide to the tieup between

b.h.p. and m:p.h. gives the tuner an idea of the results his
efforts should produce-always bearing in mind wind and
other resistance factors.
it is almost impossible to duplicate them on the bench. For

this work a correctly gearedfront-wheel-driven speedometeris a
great help, although so far as the indirect gears are concerned
it is not so useful as a rev. counter, because it is difficult at
racing speeds to convert m.p.h. into r.p.m.The rev. counter is
also considerably more sensitive to small variations in speed.
Runs can be made over a stretch of the course, or a similar
piece of road, using the correct exhaust system and the
" official" fuel, if any is stipulated. First of all, the correct
jet size should be determined by the method described in
Chapter XVIII and experiments made, when the jet is approximately correct, to check the accuracy, or otherwise,
of the ignition timing.
This is done when running on full throttle by moving the
271
TUNING FOR SPEED
lever back a very small amount in progressive stages, and

noting the rise or fall of the r.p.m. Alternatively, the timing
can be altered very easily over a range of about 4 degrees
by opening or closing the contact-breaker points, remembering that one-sixth of a tum on the adjustment is roughly
equivalent to two degrees. Care must be taken to see that
the effect of one adjustment does not mask the effect of the
other; for instance, if the advance is too great, the plug tends
to run hot and an over-large jet will have to be fitted in an
attempt to obtain the right appearance of the plug-face;
speed will suffer, both on account of the excessive advance
and the over-rich mixture.
When using alcohol fuels, it is not uncommon to find that
an increase in jet size has no effect on the mixture strength.
This is a sure sign that the fuel flow is being restricted elsewhere; possibly the needle-jet is not large enough, or else
the float chamber or fuel pipes cannot deliver a sufficient
quantity to keep the jet supplied. If this symptom is noted
it is useless to continue going up in jet sizes. The real cause
of the trouble must be found and cured, because, although
an engine runs very cool in a correct alcohol mixture, piston
failure will take place with great rapidity if the mixture
becomes weak through any cause, and there may be no
premonitory symptoms of engine distress to warn the rider
of the impending trouble.
Modem testing plant and the wealth of technical data
now available do much to obviate trial-and-error methods
of eliminating the minor snags enumerated in this concluding chapter. Such facilities, however, must be regarded
as going hand in hand with the skill of the man doing the job,
and in putting finishing touches to work assembled carefully
down to the finest detail, it is patience and perseverance
which ultimately produce the desired result.
272
APPENDIX
USEFUL FORMULAE
FRACTIONS OF AN INCH INTO DECIMALS OF AN INCH
1/1000 ............... =0.001

1/100 ............... =0.010
1/64 .................. =0.015625
1/32 .................. =0.03125
3/64 .................. =0.046875
1/16 .................. =0.0625
5/64 .................. =0.078125
3/32 .................. =0.09375
7/64 .................. =0.109375
1/8 .................. =0.125
9/64 .................. =0.140625
5/32 .................. =0.15625
11/64 .................. =0.171875
3/16 .................. =0.1875
13/64 .................. =0.203125
7/32 .................. =0.21875
15/64 .................. =0.234375
1/4 .................. =0.25
17/64 .................. = 0.26562~
9/32 .................. =0.28125
19/64 .................. =0.296875
5/16 .................. =0.3125
21/64 .................. =0.328125
11/32 .................. =0.34375
23/64 .................. =0.359375
3/8 .................. =0.375
25/64 .................. =0.390625
13/32 .................. =0.40625
27/64 .................. = 0.421875
7/16 .................. =0.4375
29/64 .................. =0.453125
15/32 .................. =0.46875
31/64 .................. =0.484375
1/2 .................. =0.5

33/64 .................. =0.515625
17/32 .................. =0.53125
35/64 .................. =0.546875
9/16 .................. =0.5625
37/64 .................. =0.578125
19/32 .................. =0.59375
39/64 .................. =0.609375
5/8 .................. =0.625
41/64 .................. =0.640625
21/32 .................. :=0.65625
43/64 .................. =0.671875
11/16 .................. =0.6875
45/64 .................. =0.703125
23/32 .................. =0.71875
47/64 .................. =0.734375
3/4 .................. =0.75
49/64 .................. =0.765625
25/32 .................. =0.78125
51/64 .................. =0.796875
13/16 .................. =0.8125
53/64 .................. =0.828125
27/32 .................. =0.84375
55/64 .................. =0.859375
7/8 .................. =0.875
57/64 .................. =0.890625
29/32 .................. =0.90625
59/64 .................. =0.921875
15/16 .................. =0.9375
61/64 .................. =0.953125
31/32 .................. =0.96875
63/64 .................. =0.984375
273
TUNING FOR SPEED
Power CalCulations
B.M .B.
P
~
B.H.P. xC
R.P.M.
C depends upon size and type of engine, though not

upon the number of cylinders, and i~ as follows:
For 250 c.c. 4-strokes
350 c.c. 4-strokes ..
500 c.c. 4-strokes}
250 c.c. 2-strokes
350 c.c. 2-strokes . .
1,000 c.c. 4-strokes}
500 c.c. 2-strokes .
52460
37480
26230
18740
13125
For other capacities the value of C can be obtained by

direct proportion.
.
.
PLAN
B.H.P. = 33,000
where P ==Brake Mean Effective Pressure (B.M.E.P.).
L - Stroke in feet.
A = Area of one piston in square inches.
N Number of explosions per minute.
:2
T orque
B.H.P. x 5,250
. R.P.M.
Therefore at 5,250 r.p.m. the torque in lb. per ft. is numerically equal to the b.h.p.
Fronde Brake Calculations
== P x R.P .M.
BH P
C
where P -Pull in pounds shown on dial.
C = A constant, usually either 4,500 or 5,500, depending on type of brake.
To find Torque or B.M.B.P. directly from the" poundspull" reading, multiply by the following factors, irrespective
of the number of cylinders in the engine:
274
THE BESTM()TORCYCLE IN THE "UJ'ffl'"

TRIUMPH ENGINEERING Co. Lid., Meriden WorkJ, Allule,. Covenlt)'.
275
APPENDIX
AR
HUR
HEELE
THE RIDERAGENT
The successful privateer, winner of many races

Including 1954 ITALIAN GRAND PRIX 250 c.c. and
1955 Hutchinson "100" 250 c.c., offers you personal
attention on all your problems and requirements.
A&ent for I
A.J.S., ARIEL, D.M.W EXCELSIOR, JAMES, NORMAN,
PANTHER, TRIUMPH, VELOCETTE. CANTERBURY,
. WATSONIAN, SURREY, BUSMAR, SWALLOW, SPARES,
ACCESSORIES and RePAIRS
WHEELER
MOTORS
45 & 51.AWATERLOO ROAD, EPSOM

Phone: 4505/6
REBDRES,
SlEEVING
Specialists In the manufa(:ture of all known
roller bearing replacement assemblies for
British and most
Foreign engines.
WORLD'S LARGEST
BIG END STOCKISTS
Valves, guides,
springs, main bearings,
pistons, etc.
C F SMITH & CO .,
83/85 ST. JOHN'S HILL

OLAPi-lAM JUNO., S.W.11
T.I.phone, Battetseo 0871
276
BRAKll CONSTANT
ENGINE SIZE
TORQUE
B.M.E.P.
ALL CAPACITIES
250 c.c. 4-str.

350 c.c. 4-str.
500 c.c. 4-str. }
250 C.Co 2-atr.
350 c.c. 2-str.
1,000 c.c. 4-str. }
500 c.c. 2-str.
4,500
5,500
1.167
.965
11.56
8.26
5.78
4.13
2.89
9.46
6.76
4.73
3.38
2.36
Example: If a 500 c.c. four-stroke engine is pulling 30 lb.

at 5,000 r.p.m. on a 4,500 constant brake, it is developing
30 x 5,000
4500
==33.3 b.h.p.
,
Its torque is 30 x 1.167 .... 35 pounds-feet.
Its B.M.E.Pis 30 x 5.78 -174poundspersq.in.
Engine Calculations
Mean Piston Speed
(ft. per min.) ...
R.P.M. x stroke in inches

R.P .M. x stroke in mm.
or ....----~~~------...
Mean Gas Velocity Through Port
Piston speed Dl1
60
x
Mean Gas Velocity Through Valve Seat
Piston speed DII x 22
(ft. per sec.) ""
(ft. per sec.) ...
60
xd v xLx7
where D = Diameter of piston.
d ... Diameter of port.
d v == Diameter at throat of valve.
L =Lift of valve.
277
TUNING FOR SPEED
(Either in. or mm. can be used on the right-hand side of

these equations, provided the same scale is used for each of
the dimensions used.)
C:R. COLLIER
Inertia Load of Reciprocating Parts

Load at T.D.C. (in pounds)=.OOOO142WN2S (1 + 2i)
Load at B.D.C."".OOOO142 WN2S
(1- 2i)
where W=Weight of components in pounds.
N =R.P.M.
S -Stroke in inches.
L = Length of connecting-rod in inches.
Centrifugal Load on Crankpin
Load == .OOOO142WN2S
where W = Weight of big-end and rollers.
N =R.P.M.
S = Stroke in inches.
To calculate speed in m.p.h. when time and distance
factors are known :
Distance (in miles) 3600
Time (in seconds) ,x ,
For speed over quarter miJe, divide 900 by the time in
seconds.
Conversion Factors
To convertMiles to kilometres
.
M.p.h. to k.p.h.
Kilometres to miles ..
K.p.h. to m.p.h.
Gallons to Htres
Litres to gallons
Pounds to kilograms
.
.
Kilograms to pounds
.
Cubic centimetres to cubic inches

Cubic inches to cubic centimetres .
M.p.h. to feet per second
278
Multiply by1.609 (or 815 approx.)

1.609 (or 8/5 approx.)
.621 (or 5/8 approx.)
.621 (or 518 approx.)
4.536 (or 4t approx.)
.2205 (or 2/9 approx.)
.4536 (or 9/20 approx.)
2.205 (or 11/5 approx.)
.061 (or 1/16 approx.)
16.39 (or 33/2 approx.)
88/60 (or It approx.)
Manufacturers 01 Highly Successful

Motor Cycles for over Half a Century
~I1I.1
MATCHLESS MOTOR cYCLES' PLUMSTBAO" ROAO .! LONOON.-S.E;llS
279
APPENDIX
CoBFFlcmNTS OF TtmRMAL EXPANSION
(per
<>
Centigrade for temperatures between 0 and 200 C.)

COBFFlCIl!NT
MATI!lUAL
.. Iovar"
Cast Iron
Mild Steel
Hardened Steel
Phosphor Bronze
Aluminium Bronze
Austenitic Valve Steel
(D.T.D. 49b, K.E. 965. Jessops G2)
Austenitic Cast Iron
Austenitic Stainless Steel (18/8 grade)
BURMAN &: SONS LTD
WYCHALL LANE
BIRMINGHAM 3()
White Ad 3124
VALVE
SPECIALISTS
Contractors to the
leading Manufactu rers. Also
valvesprings, collets.
big ends, piston
rings and liners.
Cylinder boring and
sle.ving
G. &. S. Valves and Guides

are stocked by all leading
dealen and beton.
.. LoEx " Aluminium Alloy

L 33 High-silicon Alloy
.000011
.000012
.000018
.000018
..000018
.000019
.000020
.000019
.000019
R.R.S3 B
.000022
.000022
.000022
Y-alloy
D.T.D. 424 Aluminium Alloy
Most Wrought High.strength Aluminium Alloys
8 %Copper Aluminium Alloy, L8
.000024
.000026
Heat-treated Cast Magnesium Alloys

Magnesium Alloys as Cast
Wrought Magnesium Alloys
.000024/26
.000029
.000029
Multiply coefficient by 5/9 to obtain coefficient per

degree F.
Example: To find expansion at 200 C. of a 3 in. bore in a
jacket case in L33 material.
Assuming the bore to be measured at 15 C. (normal
atmospheric temperature) the temperature rise is 185 0
C. Expansion therefore is 3 in. X 185x .000019=.0105 in.
HI LFORD, nr.Godalming,
Surrey. Phone: lOIfalmlne 1844

(2I1net>
280
.0000009
.000011
281
SPIlEO AND
R.P.M.
FOR 3.25-lN. AND 3.50-lN. X 19-IN. RACING TyRES

Difference Table
R.P.M.
m.p.h.
Gear
ratio
3.5
3.6
3.7
3.8
3.9
4.0
4.1
4.2
4.3
4.4
4.5
4,6
4.7
4.8
4.9
5.0
5.1
S.2
5.3'
S.4
40
30
1,382
1,422
1,462
1,500
1,540
1,580
1,620
1,660
1,698
1,738
1,777
1,817
1,856
1,896
1,935
1,975
2,015
2,054
2,093
2,133
2,172
2,212
2,251
2,291
2,330
2,310
S.S
5.6
5.7
5.8
5.9
6.0
I -- -60- -90 100

50
70 ! 80
110
130
120
1,844 2,304 2,765 3,226 3,688 4,146 4,608 5,068 5,530 5,990 6,452
1,895
1,948
2,()()()
2,054
2,106
2,160
2,212
2,264
2,318
2,370
2,422
2,476
2,528
2,580
2,634
2,686
2,738
2,792
2,844
2,896
2,950
3,002
3,054
3,108
3,160
2,370
2,436
2,500
2,566
2,634
2,700
2,764
2,830
2,896
2,964
3,030
3,094
3,160
3,226
3,292
3,360
3,424
3,490
3,556
3,620
3,686
3,752
3,818
3,884
3,950
3,316
3,410
3,502
3,593
3,687
3,180
3,870
3,963
4,055
4,148
4,240
4,332
4,424
4,515
4,608
4,702
4,793
4,885
4,977
5,070
5,160
5,253
5,345
5,438
5,530
2,843
2,930
3,002
3,080
3,160
3,240
3,318
3,391
3,476
3,SSS
3,634
3,713
3,792
3,871
3,950
4,030
4,108
4,187
4,266
4,345
4,424
4,503
4,582
4,661
4,740
3790
3:896
4.002
4,108
4,212
4,320
4,424
4,528
4,636
4,740
4,846
4,952
5,056
5,161
5,267
5,372
5,476
5,584
5,688
5,792
5,898
6,004
6,109
6,216
6,320
4,264
4,386
4,503
4,620
4,740
4,860
4,978
5,096
5,214
5,332
5,450
5,570
5,688
5,805
S,925
6,045
6,162
6,280
6,400
6,517
6,636
6,754
6,873
6,990
7,110
4,740
4,872
5,003
5,134
5,266
5,400
5,S30
5,660
5,792
$,925
6,057
6,188
6,320
6,452
6,584
6,716
6,847
6,979
7,110
7,241
7,372
7,504
7,636
7,768
7,900
5,210
5,360
5,503
S,646
5,794
5,940
6,082
6,227
6,372
6,520
6,664
6,807
6,952
7,096
7,242
7,388
7,532
7,677
7,822
-
5,686
5.846
6,004
6,160
6,320
6,480
6,638
6,794
6,952
7,110
7,268
7,426
7,584
7,742
6,160
6,334
6,504
6,673
6,847
7,020
7,188
7,360
7,530
6,632
6,820
7,004
7,186
7,374
7,560
----5
3
4
-----
46
47
49
SO
51
53
54
92
9S
97
100
103
138
142
146
150
154
108
111
113
U6
118
121
124
126
129
132
134
137
140
142
145
148
162
166
170
174
178
182
186
190
193
197
201
205
209
213
217
221
225
229
233
237
140
5S
57
58
59
61
62
63
65
66
67
68
70
71
12
74
75
76
18
79
lOS
ISO
153
ISS
158
IS8
184
190
195
200
205
211
216
221
226
232
237
242
248
253
258
263
269
274
279
284
290
295
300
305
311
316
230
237
244
250
257
263
270
276
283
290
296
303
309
316
323
329
336
342
349
356
362
369
375
382
388
395
277
284
293
300
308
316
324
332
340
348
355
363
371
379
387
395
403
411
419
427
434
442
450
458
466
414
1-7-[ 8
367
379
390
400
411
421
432
442
453
464
474
485
495
506
516
527
537
548
323
332
341
350
359
367
378
387
396
405
415
424
433
442
452
461
470
479
488
498
507
516
525
534
544
55l
SS8
567
579
590
600
611
622
632
I 9
41S
426
439
450
463
474
486
498
SIO
521
533
54S
557
569
580
592
604
616
628
640
652
664
675
687
699
711
---
Note.-For 3.25-;0. and 3.50-io. x 20-io. tYl"'$ subtract 4% from T.p.m. figures, for speeds intermediate between even u.n.,
add r.p.rn. from" Difference Table.' For ratios lower than si~ select any sub-multIple and multiply r.p.m. accordingly.
i.e. For 12 to I, select speed for 6 to 1 and multiply by 2.
-------_.
~.'-.'-
~.
--
i,l[~
E:s~
....
0
;J!
A
III
'1.
M
t;~2
!2!
~S"" ~
N~~
ft.
*[~ ~
()
ii:
S-...
.
~~.
ii'"
~~
!l
~
!:!.
!:i'
....
~
M
SO
fl,)1;:
'6'-
ll:S'~~
~2~
~.
!:.Il, ..-
.ii
....
f<
(tl
~1;'g t;5
";.,,
'"
:...~~ m
-~"
",,-
~:;~
"g:f[;
'9."'-
til
ll3<f
"::
;<I
;,~~~
~ ~ ;o~Q"
"~.
~:; ~~
:!
00
"'~~~
~~~e. trS
;:i"'~~ ~
.l!~
~
N
Q~~'"
"'''''
~~Q;
l?
'"
~~
~p
('Jf>
<;:>
~~ ~
5::1 .
'aR. t:1
2:
('.>1> ..
~~
':15:
;.
~S' ~
-'"
~~
-"tl.
.'"
CIl
'$"
GEOFF MONTY & DUDLEY WARD

The Speed Specialists
OFFER
Complete Tuning Facilities

Machining and Welding
The Finest Selection of Genuine Racing Machines
and Spares
HIRE PURCHASE
INSURANCE & EXCHANGES
Showroom
. 47 HAMPTON ROAD, TWICKENHAM

i Phone:
MIDDX.
Popesgrove 5040
Supertllne
. Motors
226 CLAPHAM
ROAD,
LONDON S.W.9
FOR. HIGH SPEED
TUNING
ROAD AND RACING
MOTOR CYCLES
Ride away H.P. Terms
TELEPHONE
RELlance 6420
284
INDEX
Air resistance, formula and
calculations, 269, 270
A.).S., 18, 96, 117, 161; threebearing crankshaft, 193, 194;
Vernier device for ignition
timing, 186
.. Allin", 63
Allen screws, 34
"Alpax",59
"Alnico ", cast-in magnets, 182
Amal, types of carburetter, 240,
250
Ariel, 161
Austenitic valve inserts, 12, 63
Big-end bearings, various modifications, 85, 86
Binks "mouse-trap" type of
carburetter, 171
"Birmabright," 59
Black and Decker tools, 12
Brake Horse Power, cordination
with m.p.h., 271; formula and
calculations, 265, 269, 274
Brake Mean Effective Pressure,274
"Brico", 10
. "Brightray", 30
B.S.A., 161
B.S.A. B32/B34, 188
B.S.A. "Bantam", 207, 215; a
modified cylinder head, 229;
conversion to close-ratio gears,
218; exhaust-pipe length, 211;
modified choke sizes, 219; port
modifications, 221,226; standard port timing, 216; various
compression ratios, 228-229
B.S.A. "Bantam", modified exhaust arrangement, 230; work
on the transmission, 231, 232
B.T.H. type of magneto, 183, 184
Cadwell Park, 261

Cams, basic design and modifications, 175, 181
Cam-followers, offsetting to mini
mise wear, 151
Cams, Vincent" Black Lightning"
type, 152
Camshaft drives, correct meshing,
146, 147, 148
Carburation, Amal G.P. instru
ments, 245; modifications for
"Twins"; 203,206; R.N. and
T.T. Amal instruments, 240,
249; representative settings,
243; various modifications, 248,
257; "Swill pot", 249
Champion plugs, 260, 262
"Cobalite",30
Co-efficients, comparison in varying materials, 281
Compression ratio, adjustment,
48, 49, 55; measuring procedure, 2, 5; ready reference
table, 53
Conversion factors, 280
Craig, )., LA.E. paper, 172
Crankcase details, the use of test
mandrels, 11, 112
Crankpin, centrifugal loading,
280
Crankpin modifications, 96
Crankshaft, checking end-play.
125, 126, 127
Creg-ny-baa, 261
"Cromard ", 58
Cylinder filling, improved induction,168,181
Cylinder-head joints, 136; copper
gasket annealing, 138; elongation of bolts, 138; the "doubleground" principle, 138
285
INDEX
Detonation, a limiting factor, 54,

55
"Dope" fuels, 2, 44
D.T.D. steel specifications, 138,
148
Duralumin, light-alloy, 41, 92,
145, 148
"Durex", abrasive tape, 23
Dykes ,type piston ring, 74
n.
"Elektron 59
E.M.C., 207
EN, steel specifications, 81, 92
Engine assembly, 234, 239
Engine ,balayce, calculating balance factors, 105, 108
"Enots" adjustable drip-feed,
118
Excelsior "Talisman", 214
Exhaust port, modifying work,
26,28
Eysink,207
"Fabroil", 184
Firth-Brown N.M.C. material,
139
Flywheel aligmnent, methods of
checking, 100
"Gasket-goo", 128, 167
Giiera,18
Guzzi, 18, 1~2
Heenan and Froude, 264
Hepworth:and GrandageLimited,
74
"Honeychrome", 58
Ignition advance, co-ordination
with compression ratio, 188;
"hard" and "soft" plugs, 258,
263;Itiming, an electrical check,
185
Induction, improved o.h. v., 21,22
Inlet port" modifying work, 18,
]9
J.A.P., 16,35,62,94,161,162
J.A.P., model "8.80",118: splittype collet, 40; type of fuel, 44,
195
Jessop G2 type of steel, 8, 9, 138
Judge,A. W.,The Testing ofHigh.
speed
Internal-combustion
Engines, 268
"K-Monel",]l
Kadenacy, 210
Kate's Cottage, 216
KE 9651 type of steel, 8, 9, 29,138
KoL.G., 260, 262, 263
READ WHEREVeR
MOroRCYCL.ES ARE RIPf)EN
Lambretta, 211
Laystall Engineering Co. Ltd., 58
Levis, 162
"Limalloy" piston rings, 74
"Listard' chrome7process, 58
Lister, R. A. and Co. Ltd., 58
Lodge, 260, 262
"Lo-Ex ", 59, 62
Lubrication, gear-type oil pumps,
118, plunger-type oil pumps,
120; reciprocating plunger-type
oil pumps, 120; total-loss type,
118: Velocette double-gear oil
pump, ]19
Lucas types of magneto, 183
Main bearings, removal and
refitting, 97, 98
Mainshaft aligmnent, method of
checking, 101, 102,103
"Manx" Norton, 18, 89; Vernier
settings, 158; valve timing, 173
Matchless, 117, 162; G3L main
bearings,12]: G3L main bearings, method of assembly, 123:
_ three-bearing crankshaft, 193,
194
Mean gas velocity, formula 277

Mechanical condition, an initial
check,6, 7,8
Monochrome Limited, 58
286
Every week, Moto Cycling is packed with ~he

latest news of every aspect of the motorcyclmg
sport and pastime. Results of races, trials and
scrambles; descriptions of new machines,
accessories and clothing; touring news, club
news and general interest items-all these
aspects of motorcycling find their place in the
journal. Motor Cye/ing's team of expert
writers and on-the-spot cameram~n ens!lre
that you get something new and mterestmg
every week.
EVERY THURSDAY EIGHTPENC
TEMPlE PRESS LIMITED .' BOWLING GREEN LANE' LONDON EC1
287
L.
E.
RTLE
The well-known engine development consultant

is able, as the result of over 35 years' ungrooved
experience with many engine types, to offer
unbiased advice on preparation and modification,
for all racing, trials, or touring purposes.
HARTLEY "GROOMED" ENGINES

have taken over 4,500 awards. Varying from fastest time
Brooklands Hutchinson Hundred and over 100 other
Brooklands awards, to records and awards at Cadwell,
Alton Towers, Oulton Park, Brands Hatch, Osmaston
Manor, Warminster, Silverstone, Abridge Aerodrome,
Brighton Speed Trials, Shelsley Walsh, and S.E. Centre
Grass Track Championships (solo and sidecar).
HARTLEY
FUEL USERS
have taken over 5,000 awards.
WHY NOT GET IN TOUCH NOW?

289, PI umstead High Street, London, S.E.18
Telephone: Woolwich 1895
The rider agent enthusiasts will be pleased

to supply your new, competition, racing or
standard machine.
Let our long experience be your gUide

Our well equipped and expertly manned
workshops at yo.ur disposal for major overhauls, repairs or tuning.
Personal service and attention given at all times

Advice
and
estimates
given
freely
264, MILLBROOK ROAD

so UTHAMPTO N
2RR
27744/5
INDEX
R.R.56 type of alloy, 41, 92,

148
R.R.77 type of alloy, 41
Rudge, 163
Morrison, J. G., 171

M.T. 80 type of fuel, 50
New Imperial, 162
"Ni-Resist" nickel-iron alIoy, 10
Nitro-benzine type of fuel, 207
Norton, 162;
"Dominator"
crankshaft details, 192; 30j40M
models, 188; main-bearing
details, 124
Orsatt type of gas analyser, 211
Panther, 162, 163
Petrol-benzoIe, 7, 50
Petter "Harmonic" type of
engine, 210
"Phormicast" piston rings, 74
"Pickavant" hydraulic extractor,
79
Piston crown, detail modifications, 133, 135
Piston, essential clearances, 68,
69; speed formula, 277
P.M.S. type of fuel, 44
Porting of Two-stroke Engines,
The, 171
Proceedings of the I.A.E., 171
Progress in Motorcycle Engines,
1.A.E. paper by J. Craig, 172,
173
Puch,207
Python engines, 29
S.A.E., American steel specifications,82

Seat grinders, Black and Decker,
12
"Seeger" type of piston circlip,
76,77
Side-valve engines, 16, 19
Sifbronze welding 13
SmalI-end bearings, modification
to con-rod eye, 89
Speed from a Two-stroke, article
in Motor Cycling, 216
Steels, expansion co-efficients,
139
"Stellite". 30, 149
Sunbeam, 163
Tappet adjustments, 41, 42

Tappets, direct-impact type, 151
Tetra-ethyl lead, fuel additive, 50
"Thermochrom " , piston rings,
74
"Top-half" modifications, 132,
135
Transmission, lining up chain
drive, 238
Triumph,
163;
crankshaft
assemblies, 192; special extractor tools, 190
"Tufnol", 184
Rocker mechanism, Rudge type, Twin-cylinder engine tuning
methods, 189, 206
142
Rocker modifications, lightening Two-stroke engine tuning modifications, 207, 233
o.h.v. mechanism, 143, 144
Rockwell
diamond hardness
"Unbrako",34
numbers, 81
Useful formulae, 273, 282
Rolling resistance, 269
Rose, A. E., Speed from a TwoValve gear adjustments during
stroke, 216
reassembly, 154-159
R.R. light alloy specifications, 11,
Valve-seat angles, 31
12
289
INDEX
Valve settings, representative data

161,163
Valve springs, an auxiliary return
device, 38
Valve springs, hairpin type, 36, 37
Valve timing, degree/linear conversion, 159
Van der Horst chrome process,
58
Velocette, 163; KTT head studs,
139; types of rocker box, 37;
MSS, crankpin details, 200
Villiers radially disposed finning,
213
Vincent, 117, 163, 188; adjusting

the compression ratio, 198."Black Shadow" and "Black
Lightning",
197;
big-end
details, 85; details of a modified crankpin, 199; main
bearing details, 124; rocker
lubrication, 118
TWO OUTSTANDING BOOKS

FOR PRACTICAL MOTORCYCLISTS
MODERN
MOTORCYCLE
MAINTENANCE
2nd Edition, 2nd Impression
By Bernal Osborne of "Motor Cycling"
Wellworthy Limited, 63
Wellworthy pistons, 10, 74
Y-alloy, 11, 12
An ideal book for those who take pride in the reliability

and efficiency of every component of their machine, outlining the efficient use of tools, the overhaul of working
parts, carburetter adjustment, lubrication, and the repair
and maintenance of electrical equipment.
Of equal interest to novice and expert alike, the practical
value of this up-to-date handbook is much enhanced by the
numerous explanatory illustrations and cutaway drawings.
Crown 8vo (paper boards) Illustrated 68. net, by post 6.>. 3d.
"MOTOR CYCLING"
ELECTRICAL
MANUAL
1st Edition, 2nd Impression
By Bernal Osborne o/"M%r Cycling"
This practical and informative book deals with most of the

electrical problems likely to confront the motorcyclist.
First principles of electrical theory are explained, together
with the functioning of the ignition and lighting equipment
of the modern motorcYcle. The Manual is written throughout in simple, non7 technical language and the text is
supported with many line drawings and a supplement of
wiring diagrams of post-war BritIsh machines.
CrowlI 8vo (paper boards) Illustrated 5s. net, by post 5s. 4d.
Obtainable from your Bookseller or by post from the Publishers
TEMPLE PRESS LIMITED

Bowling Gl"een Lane, London, E.C.I
290
291

Tining For Speed

Uploaded by

Copyright:

Available Formats

Tining For Speed

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tining For Speed

Uploaded by

Copyright:

Available Formats

What topics does the book cover?

What topics does the book cover?

What modifications are discussed for improving engine performance?

What modifications are discussed for improving engine performance?

THIIID

TUNING FOR SPEED

ALSO PUBLISHED IN ASSOCIATION WITH

"MOTOR CYCLING" ELECTRICAL MANUAL

HolY to increase the performance of a standard

"MOTOR CYCLING" ROAD TESTS

"MOTOR CYCLING" SERVICE SERIES

TEMPLE PRESS LIMITED

FIRST PUDLiSHED 1948

ATTENTION TO O.H.V. CYLINDER HEADS

ADJUSTING COMPRESSION RATIOS

CHECKING PISTON AND RING WEAR

THE CONNECTING-ROD AND BIG-END

TRUING AND BALANCING FLYWHEELS

REASSEMBLING CRANKCASE COMPONENTS

WORK ON THE CRANKCASE

LIGHTENING THE VALVE MECHANISM

ADJUSTMENTS TO VALVE GEAR

IMPROVED CYLINDER FILLING

ASSEMBLING THE ENGINE IN THE FRAME

MAINTENANCE OF RACING CARBURETTERS

MADE IN GREAT BRITAIN

Prill/ed by Purnell alld SOliS, Ltd.

CARBURATION AND INDUCTION FACTORS

"HARD" AND "SOFT" TYPES OF SPARKING PLUG

TESTING THE FINISHED WORK

supplied in this' Manual. Written by a well-known technician

TUNING FOR SPEED

be needed a scribing block and dial gauge reading to .001

if necessary, until it reaches half-way up the hole. Now

and so on for any other capacity (see Chapter IV), The

The method 0/ measur-

I'oiume with a burette

of the head will be somewhat less than the apparent volume

TUNING FOR SPEED

squirted up the inlet and exhaust ports. For checking the

plan to determine for subsequent reference what the actual

plotted on squared paper and the effect of any subsequent

Sometimes a bearing will be slightly pitted in one small

TUNING FOR SPEED

in methanol and, as the mixture in such engines is often

TUNING FOR SPEED

destroying the close adherence between the two metals in

Ri,ngs should always be shrunk in, never pressed in cold,

TUNING FOR SPEED

Should the plug hole be tapped to take the old-fashioned

18 mm. " I.S PITCH

All old-type cylinder head fitted with all 18 mm. sparking

TUNING FOR SPEED

If ),ollr cams give a

indication that cams giving longer valve openings are

The lift must also be considered and should be slightly

CARBURATION AND INDUCTION FACTORS

CARBURATION AND INDUCTION FACTORS