FINISH

CARPENTRY

THE BEST OF

r
I
I
.4
 FINISH
CARPENTRY
THE BEST OF
Fme
 ____The

FINISH
CARPENTRY
THE BEST OF
Fme Homebuilding

'ihunton Press
Cover photo: Kevin Ireton
Back-cover photos: Charles Miller (top),
Jeff Kolle (bottom)

Taunton
BOOKS & VIDEOS
for fellow enthusiasts

©1997 by The Taunton Press, Inc.

All rights reserved.

First printing: 1997

Printed in the United States of America
Library of Congress Cataloging-In-Publication Data
A Fine Homebuilding Book Finish carpentry : the best of Fine homebuilding.
p. cm.
Fine Homebuliding® is a trademark of The Taunton Press, Inc., "A Fine homebuilding book"—T.p. verso.
registered in the U.S. Patent and Trademark Office. Includes index.
ISBN 1-56158-183-6
The Taunton Press, Inc. L Finish carpentry. L Taunton Press. II. Fine homebuilding.
63 South Main Street TH5640.F56 1997
P.O. Box 5506 694—dc2O 9641123
Newtown, Connecticut 06470-5506 CIP
 CONTENTS
7 Introduction 48 Table-Saw Molding 90 Building a
The secret is in the order Fireplace Mantel
8 Basic Scribing of cuts Fluted pilasters and
Techniques decorative trim make for a
A finish carpenter shares 50 Installing Crown mantel that appears
his secrets for fitting trim to Molding complicated but really isn't
uneven, unlevel or unpiumb Upside down and
surfaces backwards is the secret 94 Building a
Federal-Style Mantel
14 Hand Planes for 54 Installing Two-Piece Neoclassical detailing in
Trim Carpentry Crown poplar and ceramic tile
Tuned and adjusted right, A method for running wide,
these planes will save time paint-grade crown moldings 98 Baronial Inglenook
and improve your work Laminated mahogany and
57 Making Curved Crown antique English details
20 Plate Joinery on the Molding surround a fireplace
Job Site Glue-laminated trim can be
Quick and easy insurance shaped on a table saw 104 A Built-In
against joints opening up Hardwood Hutch
60 Making Classical When working with solid
24 Survey of Finish Nailers Columns wood, joinery techniques
The new trim nailers are Interior detailing with a must accommodate
lighter, smaller and more router and lathe seasonal movement
powerful than ever
64 Retrofitting a 108 Simple Closet Wardrobe
30 Installing Baseboard Threshold A biscuit joiner and a layout
There's a tad more to it than A three-piece threshold
jig make construction quick
coping the joints provides extra weather and easy
protection, especially in
32 Curved Baseboard exposed locations 112 Bed Alcove
Corners Convert wasted attic space
Biscuit joinery and a simple 68 More Than One Way to into a bed that has drawers,
jig solve a perplexing trim Case a Window bookshelves and a vanity
problem You can vary the look with
simple combinations of flat 116 Bookshelf Basics
34 Running Baseboard boards and stock moldings A guide to support systems,
Efficiently shelving designs and
Simple steps help you make 74 Making Curved Casing materials
the most of time and Strip-laminating arches to
materials match straight casing 122 A Home Library
profiles Maple, purpleheart and an
37 Molding Character enthusiasm for small details
Using a molder/planer on 78 Raised Paneling
site to create a formal 18th- Made Easy 126 Index
century interior A combination of built-up
moldings and layers of
42 Making Plaster plywood duplicates a
Molding traditional look
Sometimes complex shapes
and profiles are easier to 82 Raised-Panel
make in plaster than in Wainscot
wood Traditional results with
table saw and router
46 Cutting Crown Molding
Calculating miter and bevel 86 Drywall Detailing
angles so you can cut crown An alternative to wood trim
on compound-miter saws around doors, windows and
skylights
 INTRODUCTION

T'S AXIOMATIC IN BUILDING thatwhatevertrade follows yours in the

I construction process will cover up the mistakes you leave behind. Hence,
the foundation crew assumes the framers will compensate for their basement
walls being out of level. The framing crew assures themselves that drywall will
hide their bowed studs and out-of-square walls. And the drywall contractor tells
anyone who'll listen that "trim will cover it."

Now, I've heard some trim carpenters say "the painter will fix it," but you don't
want any of them working on your house. For the most part, the buck stops with
the trim carpentry. It's the piping on the cowboy's shirt: everybody's going to
see it, and it has to be right.

In this book you'll find articles about trim carpentry collected from back issues
of Fine Homebuilding magazine. Covering everything from baseboard and crown
molding to built-in furniture, these articles were written by experienced trim
carpenters—none of whom rely on painters to hide their mistakes.

—Kevin Ireton, editor

 Basic Scribing Techniques
A finish carpenter shares his secrets for fitting trim to
uneven, unlevel or unpiumb surfaces

by Jim 101pm

he first thing I learned as a finish carpenter Also, don't assume that you can simply press measure it, then reset your bevel square from the
was that square corners, plumb walls and level the square against converging surfaces to get an block to mark your trim. If you need to quantify
floors and don't exist on this planet. And accurate reading. Say, for instance, that you want an angle in degrees, measure it on the block with
because that's just the way it is, it was up to me to to fit a baseboard to a door casing (top photo, a Speed Square.
learn how to work with these unfortunate diver- facing page). To measure the angle of the end
gencies from the way it ought to be. As the finish cut, set the baseboard where you want it on the The base hook—Another homemade tool,
man, my job was to fit the pretty stuff to the struc- floor, then place the body of your bevel square called a base hook, eliminates the need for a
tures that framers and rockers left behind, no on top of the baseboard to measure the angle of bevel square in some applications. Similar in
matter how crooked they were. the casing. If you simply lay the body of the concept to a siding gauge (see the cover of FHB
In my quest for perfect fits, I learned how to square on the floor, any bumps or dips in the #47), it's simply an L-shaped piece of a stable,
use bevel squares and base hooks, among other floor next to the joint will fool the square into split-resistant wood used primarily for laying out
tools, and became proficient in the use of a measuring a false angle. An alternative is to set a the end cut of baseboard where it butts against
slightly customized pencil compass. I learned level or a straightedge on the floor and to mea- standing moldings such as door casings (bottom
from legendary boat builder Bud Macintosh sure the angle off of that. right photo, facing page). To use the hook, lap it
how to use something called a spiling batten to Once you measure an angle, be careful not to over the baseboard and hold it hard against the
solve certain awkward scribing problems, such jar the bevel before you scribe the workpiece. standing molding while scribing a cutline across
as fitting the last ceiling board. I even paid hom- Fortunately, there's an easy way to ensure against the baseboard. Be sure the faces of your base
age to the linoleum trade and learned the inge- the loss of an angle setting on a bevel: Record it hook are perfectly square to the edges, or you'll
niously simple "Joe Frogger" method of creating with the help of a boat-builder's bevel board. introduce a margin of error.
a template that can produce dead-accurate fits
every time. Saving angles—Boat builders, who confront Scribing to irregular surfaces—Shortly after I
compound angles on nearly every piece they fit, became a finish carpenter, I bought a $5 pencil
Using the bevel square—A bevel square is a have developed a simple, shop-made accessory compass like the kind my kids tote in their school
layout tool with a wood, metal or plastic body that makes it easy to measure and record a se- bags. It has two adjustable arms, with a metal
having an adjustable metal blade attached to ries of angles for future reference at the saw feeler point at the end of one arm and a pencil at
one end. The square is used mostly for deter- table. Called a bevel board, it's a board with a the end of the other (bottom photo, p. 12). For
mining the angle at which a piece of trim needs bunch of lines drawn across it at angles ranging improved accuracy, I heated and bent out the
to be cut to fit tightly against a surface. from 00 to 45° (bottom left photo, facing page). feeler point of my compass slightly so that the
My first bevel square came from my grandfa- The bevel board allows you to measure an an- point, rather than a portion of its side, contacts
ther. it's a nice rosewood-bodied job with a 6-in. gle with your bevel square and then read the de- the meeting surface. (The meeting surface is
long blade. It's pretty and has sentimental value, grees of the angle directly from the board. If the whatever is being scribed to; I'll call the piece to
but like many contemporary bevel squares, it's angle scale on your bandsaw, table saw or chop- be cut the workpiece.)
not the best tool for taking angles. This is because saw is calibrated to the bevel board, you need Although I haven't tried it yet, I recently learned
its locking lever, which is located at the pivot only to set the saw to the appropriate degree a tip from Gary Katz, a contractor in Encino,
point of the tool, often sticks beyond the edge of mark and cut away. if more than one angle is be- California. To ensure that he can always scribe a
the body and gets in the way. Also, the body is ing taken at once, the angles are simply recorded fine line, Katz fits his compass with a Cross #3503
quite thick, which holds the blade away from the on a scrap of wood or paper that represents a mechanical pencil (A. T. Cross Co./ATX Market-
stock. This can throw off the angle measurement. story board of the piece or pieces to be cut. ing, One Albion Road, Lincoln, R. I. 02865; 401-
What's more, the body is relatively short, which The bevel board should be made of a stable 333-1200). This pencil is expensive ($15.50), but it
can also produce inaccurate readings. wood, such as mahogany or teak, that has an in- scribes a very fine line, is well made and has a
I like my all-metal Japanese bevel square better terlocking, split-resistant grain. You can also use a wonderful warranty. No matter how you damage
(bottom left photo, facing page). It's much thin- scrap of 3/8-in, plywood. Using a protractor, scribe it and regardless of its age, you can return it to
ner than a conventional bevel square; the lock is the lines to the board with an awl and then fill Cross, and they'll send you a new one.
a knurled knob that's out of the way; and it can them in with an indelible ink. Keep the board When scribing a line with a compass, you are
be held and locked with one hand. thin so that it will be lightweight thick is actually transferring the pattern of the meeting
Although the use of a bevel square may seem sufficient); leave room at both ends of the board surface onto the workpiece. It is very important,
straightforward, it's not. Always extend the blade for indexing the body of the bevel against it; and as you scribe the line, that the feeler point on
fully before pressing the outside edge of the body radius or chamfer the top edge of the board so one side of the compass not get ahead of or lag
against a surface to measure an angle (such as that you can orient it at a glance. An alternative behind the pencil point on the other side.
when measuring an inside corner where two board, sans the romance (and not quite as easy Throughout the scribing process, these two
walls meet). Any protrusion of the blade beyond to read), is made by scratching the lines deeply points must align parallel to the direction the
the outside edge of the body will hold the body into a piece of Lexan plastic. workpiece will move to contact the meeting
away from the surface it's resting against, throw- In lieu of a bevel board, you can scribe and la- surface. If they don't, the result will be an inac-
ing off the angle reading. bel each angle on a wood block right after you curate pattern and, ultimately, a sloppy fit. (top

8 The Best of Fine Honiebuilding

 a

Fitting baseboard. Author Jim Tolpin uses

either a bevel square (photo above) or a home-
made base hook (photo below right) to lay out
baseboards for a fight fit against door casings.
Bumpy floors can fool a bevel square, so To!-
pin always measures the bevel off of a straight-
edge or a piece of the baseboard Itself. The
base hook is used by holding it hard against
the casing (or plinth block in this case) while
scribing a cutline directly on the baseboard.
Boat builder's bevel board. Etched wIth 46
labeled lines spaced 10 apart, the homemade
bevel board (photo below) makes It easy to
read an angle off a bevel square, then adjust a
saw to that angle.

_._ —

finish Carpentry 9
 drawing, left). Chinkless-log-home builders, who
Scribing with a compass Right routinely scribe logs to fit together tightly, have
developed a homemade compass with an ad-
When scribing a line with a corn pass, an accurate pattern results from keeping the feeler
point and the pencil point aligned in the proper direction throughout the process. These just- able bubble level on it that makes it easy to
points should always align parallel to the direction the workpiece will move to contact the keep the compass oriented properly while scrib-
meeting surface. ing. For more information about this compass,
- pOult
Feeler see FHB #53, pp. 80-84.
Workpiece It's amazing what the pencil compass allows
you to do. For instance, it really comes into its
Scribed
own for fitting a wall panel or a vertical siding
board to a bumpy surface, such as a fireplace
(bottom drawing). The procedure is straightfor-
ward. First, plumb the panel or board and tack it
to the wall about '/2 in. away from the closest spot
on the meeting surface. Then set the compass to
distance "X" between the edge of the panel and
the bottom of the deepest valley on the meeting
surface, plus 1/4 in. so that the scribed line won't
fall off the edge of the workpiece. Hold the com-
An inaccurate pattern results from letting the feeler point of the compass get ahead of or pass level along the entire vertical run and trace
lag behind the pencil point while scribing. the meeting surface with the feeling point so that
the pencil transfers the profile to the workpiece.
(If the workpiece is dark, a strip of wide painter's
masking tape applied to the panel will make the
line more legible.) Finally, remove the workpiece
from the wall, back-cut it (bevel it back) a few
degrees along the scribed line, then test fit it
against the meeting surface, lithe fit is good in
some areas and way off in others, you probably
let the compass wander from level during scrib-
ing. If this happens, try again. If necessary, final
fitting is achieved through hand planing, sand-
ing, rasping and filing (more on that later).
Sometimes the closing (last) board or panel on
a wall must be scribed. This is tougher to do be-
cause the board has to fit into an existing gap.
Scribing a panel For one solution to this problem, see Tom Law's
Scribe panel with compass method on the facing page.
level. set to distance "X." Another common scribing problem is fitting
stair treads between a pair of skirtboards (or sim-
surface
ilarly, closet shelves between two walls). This is
accomplished by cutting the tread ¾ in. longer
than its final length; dropping it into place with
one end riding high on a skirtboard; scribing and
cutting the low end; marking the final length of
the tread by measuring off the scribed end; drop-
ping the tread back into place with its scribed
end riding high; and then scribing and cutting
the low end to the measurement mark (for an il-
Panel
lustration of this trick, see FHB #68, p. 61).
¾ in.
1/ Some fitting jobs are accomplished using a
pencil compass in concert with a bevel square
and a combination square. Laying out a window
Panel Back-cutting detail stool is a good example (photos, p. 12). In this
case, use a combination square to locate the out-
Face of panel
side corners of the window opening (where the
wall meets the side jambs); a bevel square to lay
out the angles of the side jambs relative to the
front edge of the stool; and a compass to scribe
the stool horns to the wall. For added conve-
Cutting the scribed edge at slightly nience, a couple of sticks tacked to the sill (per-
less than 90' ensures a snug fit.
pendicular to the window) will support the stool
while you lay it out.
To scribe a wall panel to an uneven surface, plumb the panel and tack it up about ½ in.
away from the closest spot on the meeting surface. Then scribe the panel with the Scribing with a spiling batten—Sometimes
compass points set to distance "X" (the distance between the edge of the panel and the it's awkward to hold a workpiece in position for
bottom of the deepest valley on the meeting surface plus ¾ in.). Back-cutting the panel
will ensure a snug fit. scribing. A perfect example is scribing the clos-
ing board of a wood-strip ceiling. Not only do
you have to hold up the board while scribing it,

10 The Best of Fine Homebuilding

 remove all of the tacked-up unchanged, to scribe the board to
Scribing the closing board boards, hold the closing board the meeting surface.
by Tom Law hard against the meeting surface I almost always make the cut
and mark the top and bottom of with a handsaw, cleaning up to the
The installation of wainscoting simply measure the gap and rip its trailing edge on the wall line with a block plane if
and the Installation of vertical the board to width. lithe meeting (points C and D). necessary. A handsaw cuts on the
siding both have the same surface is straight but not plumb, I Setting the closing board aside, downstroke, ensuring that any
problem—fitting the closing (or measure across the top and bottom I now adjust my compass to span tearout will occur on the backside
last) board. I approach this of the gap, transfer the either the top or the bottom two of the board where it won't show. I
problem by nailing up all but the measurements to the board, marks, whichever are the farthest undercut the board slightly so that
last few boards. Then I tack up the connect the marks with a straight apart (points A and C in the when it's sprung into place it
rest of the boards except for the line and rip the board with a example). I then use the compass makes a nice, neat joint.
last one. I mark and cut this circular saw. to make a new mark (point E) at When installing wall paneling
closing board, remove the tacked If the meeting surface is the opposite end so that the top instead of boards, only the
ones, then spring the whole group irregular, the board needs to be and bottom pairs of marks are the second-to-last panel is tacked
Into place at once and nail them to scnbed as shown in the drawings same distance apart (if they up and removed to fit the last
the wall. below. In this case, after I've weren't to begin with). Finally, I piece. Otherwise, the scribing
How I fit the closing board tacked up the next-to-last board, I replace the closing board so that and cutting procedure is the same.
depends on the nature of the mark the top and bottom of its its trailing edge falls on the —Tom Law is a consulting
surface it meets. If the meeting leading edge on the underlying appropriate marks (C and E), then editor of Fine Homebuilding and
surface is straight and plumb, I wail (points A and B). Then I use the compass, with its setting lives in Westminster, Md.

Here's a four-step method for fitting the last vertical board Meeting surface
on a wall to a bumpy surface. Closing board
Underlying wall

Last few boards tacked

temporarily to wall.

Point D
Underlying wall
Step 1: Step 2:
Install al/but the last board on the wall, tacking up the last few Remove the tacked-up boards, hold the closing board hard against
boards for easy removal. Mark the leading edge of the second-to- the meeting surface and mark the top and the bottom of the board
last board on the wall (points A & B). along its trailing edge (points C & D).

Point A
Point

Underlying
wall

Point
Point F

Point Step 4:
Step 3: Align the trailing edge of the closing board with points C & E, then
Adjust a pencil compass to span either the top or the bottom two scribe the board off of the meeting surface with the compass
marks, whichever are the farthest apart (points A and C this time). setting unchanged. Once the board is cut to fit, spring it and the
Use the compass at this setting to mark point E on the wall. remaining boards in place and nail them to the wall.

Clapp
Finish Carpentry 11
 Fitting the last ceiling board (Looking up at the ceiling)
½-in, maximum gap
Ceiling joists batten is tacked to joists.

½-in, gap Next-to-last ceiling board

-.... ..

Spiling batten is scribed with corn pass.

Insurance circle records Set compass to

the proper compass setting. maximum gap plus ¼ in.

.T III.

Fitting a window stool The author

uses a combination square to mark the
outside corner of the window opening Clamp Align unscribed edge of batten
on the stool (top photo above). Then he flush with trailing edge of stock.
uses a bevel square to measure the angle
between the window jamb and the front
edge of the stool. This angle is scribed The best way to fit the closing board batten, remove it and clamp it to the
(last board) of a wood-strip ceiling is to ceiling-board stock. Now scribe the
on the stool through the corner point use a spiling batten. Cut the batten ceiling board off the batten with the
(middle photo above). Finally, he sets his (typically a ¼-in, thick strip of softwood) compass setting unchanged. Before
pencil compass to span the distance be- about I in. shorter and 1/2 in. narrower scribing the batten, it's a good idea to
tween the corner of the opening and the than the ceiling gap, then tack it into the draw an insurance circle on it with the
corner mark on the stool, then scribes gap tight against the next-to-last ceiling compass so that you can double-check
the stool's horn to the wall (bottom pho- board and centered end to end. Scribe the compass setting against the circle
to above). the leading edge and both ends of the before scribing the ceiling board.

but the oversize board tilts into the opening. This board, spaced ½-in, shy of the wall at either end it easy to plane the board to fit if necessary.
tilt can throw off the scribed line. (drawing above). I then set my compass to the
Boat builders confront this exact situation maximum gap between the batten and the wall, Fitting floors to posts—The bottom right photo
when planking a wooden hull, and they've come plus ¼ in. to make sure that the scribed line on the facing page shows a wide floorboard that
up with a nifty device to cope with it: the spiling doesn't veer off the edge of the batten. Before fits tightly around a post. If the post had been
batten. The spiling batten is simply a thin strip of scribing the batten, I draw a circle on the batten square and its faces flat, I would have laid out
wood (¼-in, thick softwood is standard) that's with the compass to serve as a reference if the the floorboard using a combination square. But,
tacked into the opening that the last plank (or compass is bumped inadvertently. Once the bat- of course, the post isn't perfect, and the combi-
shutter) will have to fill. The batten is scribed (or ten is scribed, I remove it and clamp it to the clos- nation square stayed in the toolbox.
spiled, as boat builders would say) to the meet- ing board, positioned with its trailing edge (the Instead, I called on the Joe Frogger method,
ing surfaces along its leading edges and ends, edge that meets the second-to-last ceiling board) as it's known in the linoleum trade, to make a
then removed and clamped to the workpiece. flush with the trailing edge of the board. The template that works like a spiling batten. You'll
The scribe is then reproduced in reverse, from board is then scribed off the batten with the com- need a pencil, a utility knife, a piece of heavy felt
the batten back to the work. pass setting unchanged (double-checked against paper or noncorrugated cardboard for the tem-
For ceilings, I cut the batten about 1-in, shorter the insurance circle on the batten). plate and a small block of wood measuring
and Y2-in. narrower than the ceiling gap, then I back-cut the ceiling board about 50 to allow about 11/2 in. square by about V2 in. thick (the
tack it up snug against the next-to-last ceiling the board to swing into place. And this cut makes block is the frog).

12 The Best of Fine Homebuilding

Photos this page Jim Tolpin
Joe Frogger. A trick of the linoleum trade called Joe Frogger makes It easy to fit a floorboard to
a post. First, Tolpin cuts a cardboard template to fit around the post and tapes it to the subfloor,
tight against the previously installed floorboard. Next, he holds a small block of wood (the frog)
against the post in several spots while marking the frog's outside edge on the template (photo
above). After removing the template, he tapes it over the next floorboard and uses the frog to trans-
fer each mark from the template to the board (photo above right). The marks are joined using a
straightedge and a pencil. The reward is a perfect fit (photo right).

The procedure is simple (photos this page). each mark on the template while you mark the reach into dips. Fine-tuning is accomplished with
use the utility knife to cut an opening in the
First, opposite end on the floorboard. Finally, join the flat and round files.
template that matches the profile of the post, marks using a pencil and a straightedge, then I've worked with a guy who insists that a belt
adding about 3/4 in. clearance all around. Slip the back-cut slightly along the cutlines. If you're care- sander is faster and more controllable than a jig-
template around the post, tight against the last ful, you'll be rewarded with a perfect fit. saw for wasting stock to a wiggly line. Another
installed floor board, and attach it to the subfloor guy uses an angle grinder. Still another scribes
with doublestick tape. Then hold the frog against Cutting it—Once a workpiece is laid out, there with a small bandsaw, which he outfits with a
the post at stations spaced a couple of inches are a number of ways to cut it. Unless the cut- pair of wheels to make it maneuverable on the
apart while you mark along the outside edge of lines are relatively straight, allowing the use of a jobsite. E
the frog on the template with a sharp pencil. circular saw, I always use a Bosch 1581 VS jigsaw
Rabbets at opposite ends of the frog make it easy to cut just to the line. The saw blows dust off the Jim Tolpin is a finish carpenter, cabinetmaker and
to orient the frog in the same way at each station cutline, its reciprocating blade cuts fast, and its writer in Port Townsend, Wash. His book, Working
(scribing is always done off a rabbetted edge). tilting base allows back-cutting. Besides making it at Woodworking, is available from The Taunton
Next, remove the marked template from the easy to trim stock for a tight fit, back-cutting al- Press, Inc.; (800) 888-8286. His manual on finish
subfloor and tape it to the floorboard to be fit, lows the workpiece to be squeezed into place. carpentry was published by Craftsman Book
flush with the board's end and trailing edge. I use a block plane and rasps to remove stock Company Photos by Patrick Cudahy except
Then index a rabbetted end of the frog against up to the cutline, skewing the block plane to where noted.

Finish Carpentry 13
 Hand Planes for Trim Carpentry
Tuned and adjusted right, these planes will save time and improve your work

by Scott Wynn

ore than 100 different wood and metal

hand planes are described in R. A. Salaman's
book Dictionary of Woodworking Tools: c. 1700-
Metal block planes.
1970 (published by The Taunton Press, Inc.). The low-angle block
Store-bought or handmade, many of these clever plane (right) excels at
devices were once indispensable to builders. shearing end grain. The
high-angle block plane
Before the advent of power planes and routers, (below) is designed for
a carpenter's repertoire might include assorted general work. Both are
designed for one-
bench planes for preparing and smoothing wood handed use.
stock; molding planes for shaping everything
from stair nosings to door casings; and various
contraptions for plowing dadoes, grooves and
rabbets. A specialized carpenter might even
own a compass plane for cutting convex or con-
cave curves and a "galloping jack" plane for
smoothing floorboards.
Nowadays, most of these planes are prized
more by museum curators and tool collectors
than by carpenters. But some types remain as vi-
tal on the job site as ever. As an architect/builder
who specializes in trim carpentry, I use several
Block-plane anatomy
kinds, primarily for fitting wood trim or casework
against previously installed work, or wherever
the use of a power plane or a router is impracti-
cal. My favorites are the block plane, the shoul-
der plane and the butt mortise plane. I also use
an assortment of specialty planes (I made some Throat-
myself) for cutting roundovers and chamfers. adjustment
Hand planes are available from woodworker's lever
suppliers, mail-order tool outfits and some hard-
ware stores and lumberyards. But don't expect
planes to make smooth cuts straight out of the
box. Properly tuned and adjusted, though, they'll
cut wood like butter and sing while they work.
/
Adjustable Nib engages
The block plane—The typical metal block throat piece Blade seat notch in blade
plane (drawing right) features an adjustable
blade housed in a 6-in, to 7-in, long metal body.
Mounted bevel-side up, the blade is clamped by
a lever cap in two areas: against either one or cleaning up jigsaw cuts, fitting cabinets to walls, hard to tell the difference between them except
two milled plateaus or a lateral adjustment lever flush-trimming screw plugs, planing door jambs under the most demanding circumstances.
at the top end of the blade, and against a narrow flush with adjacent walls before installing cas- All four of these planes have an adjustable
angled seat at the bottom end. The seat is direct- ings and plenty of other routine tasks. throat, an important feature for preventing
ly behind the throat (the opening in the sole of The planes that most carpenters are familiar tearout—especially when making exceptionally
the plane through which the blade projects). with are the Stanley No. 12-020 and No. 12-060 fine cuts. Some block planes don't have an ad-
Depth of cut is controlled by turning a knurled (photo above) and the Record No. 09½ and No. justable throat: Don't bother with them.
nut or knob at the back of the plane. 060½, though similar tools are made by other
Unlike its larger siblings—the jointer plane, the manufacturers (I own Stanleys). The 12-020 and Body work—For a block plane to work right,
jack plane and the smoothing plane—the block the 09½ bed the blade at about 20°. The other its sole must be flat, its blade properly bedded,
plane is designed for one-handed use and will two bed it at 12°. A low-angle plane is best for and the front edge of its throat must be smooth
fit into most tool pouches. These attributes make planing softwoods, hogging off wood and shear- and parallel to the blade's cutting edge. The
it the plane of choice for most carpenters. I use ing end grain; a higher-angle plane cuts hard- blade must be sharpened, with its back flat and
mine for trimming miters, fine-tuning the fit of woods with less tearout. Both types, however, free of imperfections (for more on sharpening,
passage doors and flush-mounted cabinet doors, perform so well when properly tuned that it's see the sidebar on p. 18).

14 The Best of Fine Homebuilding

 Woodworkers have long debated the wisdom
of flattening plane soles. Some argue that planes
come flat enough from the factory, but I think
a few minutes spent flattening a plane sole
can improve performance significantly. Block-
plane soles don't really have to be flat along
their entire length. What matters most is that
three areas of the sole contact a flat surface:
the throat and both ends. lithe throat area is re.
lieved even slightly, the plane performance will
be diminished.
Before flattening the sole, retract the blade but
don't remove it. This way the plane body is
stressed as it would be in use. I flatten the sole by
rubbing it on a dry sheet of 600-grit wet-or-dry
sandpaper laid on plate glass, being very care-
ful not to rock the plane in the process. You can
also use a saw table or a jointer bed instead of
glass if you're sure they're flat (they usually
aren't). The high areas of the sole will develop a
dull, gray color that's easily distinguishable from The rabbet plane. The Stanley No. 12-07b plane has two blaae for regular rabbeting
(above) or bulinose work. It comes with a cutting spur for cross-grain work, an adjustable fence and
the low spots. When the throat area and both a depth gauge. Photo by Vivian Olson.
ends of the sole turn this color, you're done. If at
first the throat doesn't touch the sandpaper, I The 3-in-I plane. The interchangeable nosepleces of the Record No. 311 "3-In-i" plane allow it to
switch to 220-grit sandpaper to speed up the pro- be used as a shoulder plane for rabbeting (below), a builnose plane for workinif In confined
cess, then to 320-grit, 400-grit and finally 600-grit spac r a chisel plane for cuffing - - rabbets.
paper once the throat makes contact. Finally, I
smooth the edges of the sole with a file to remove
any burrs or imperfections.
Next, inspect the blade seat to make sure that
no burrs or bumps remain from incomplete mill-
ing. High spots can be leveled by removing the
adjustable throat piece from the plane and flat-
tening the bumps carefully with a fine file. If you
don't see any bumps, don't touch the seat: You'll
have a tough time restoring it if you mess it up.
Now mount the blade (and the throat piece if
you've removed it) in the plane, sight down the
sole and adjust the blade so that it protrudes
'/32 in. or so, with the cutting edge parallel to
the sole. Then adjust the throat piece so that it
almost touches the cutting edge of the blade.
Hold the plane up to a light and sight through
the throat to make sure that the cutting edge is
parallel to the edge of the throat piece. If it isn't,
or if the edge of the throat piece isn't smooth and
sharp-edged, remove the throat piece and file it
where necessary. The throat edge must be
straight and sharp. Do not round the edge of the
throat piece, or the edge won't bear effectively
on the workpiece to help prevent tearout.
Lastly, if you plan to use the plane with a miter-
shooting board (see sidebar, p. 16), use a square
to check that the sides of the plane body are rel-
atively flat and are perpendicular to the sole. (If plane so that the entire cutting edge appears at
you plan to buy a new plane, check it for square the throat as a black hairline. Now hold up the
in the store first so that you don't get stuck with a plane to a light source and adjust the throat piece
lemon). Carefully file off any high spots on both so that the throat opening (the distance between
sides. Now the plane is ready for action. the throat piece and the blade) is about in.
for planing hardwoods or in. for planing soft-
Using the block plane—To use the block woods. To combat tearout, the throat opening
plane, mount the blade bevel up in the body and should be no wider than the thickness of the
clamp down the lever cap, making sure that the shaving. Test your settings by taking a few shav-
lever cap's adjusting screw is tight enough to pre- ings from a wood scrap. For fine work, the shav-
vent the blade from being pushed around easily ings should be straight or rippled and thin
(but no tighter or you risk damaging the plane). enough to read through. lithe throat jams, the
Next, set the depth of cut and the throat open- opening is too narrow or the blade is set too The Japanese block plane. Designed for
maximum control, the author's Japanese block
ing according to the work you are doing. Flip the deep. Adjust the plane and try again, repeating plane features a laminated-steel blade that
plane over, sight down the sole and adjust the the process until you get the shavings you want. holds an edge longer than western blades do.

Finish Carpentry 15
Making a miter-shooting board Block planes will also hog off wood. To do this,
open the throat about '/8 in. wide to prevent over-
Trimming small, short pieces of wood with a during use. I also rub a little candle wax to heating of the throat piece and the blade. Adjust
power miter saw is dangerous. A hand miter reduce friction where the plane will contact the blade downward incremently until you get
box won't trim less than a saw kerfs width the board. shavings of the desired thickness.
(if that), and I'm not ready to buy a miter Before using the board, make sure your To preserve the cutting edge, don't bang it
trimmer, a pricey tool that resembles a plane's sides are square to the sole. If not, against the workpiece when beginning a cut, and
paper cutter. The solution to this dilemma is file the sides until they are (mine only don't drag the plane backwards along the work
very old: the miter-shooting board, also needed a touch-up in a few spots). Make surface between strokes. Also, I always set down
known as a bench hook (drawings below). sure the blade is sharp, and set It for a my planes on their sides, not on their soles.
It's cheap, portable, safe and is less likely very fine cut with the throat open wide
than expensive tools to walk away when (tearout isn't a factor when planing across The Japanese plane—Despite their versatility,
your back is turned. Better yet, it can be the grain). Then lay the plane on its side on my metal block planes have one drawback: lim-
used with a block plane, which lives in most the shooting board, making sure that it rests ited durability of the cutting edge. Nowadays,
carpenters' and cabinetmakers' tool kits. fiat against the base of the jig and the there are high-quality aftermarket blades avail-
I made my shooting board from scraps. shooting edge. Move the plane to engage the able that hold an edge longer than my stock
The shooting edge should be made out of a workpiece and then, with one firm stroke, blades do. One company, Hock Handmade
durable material (such as '/4-in, tempered remove a continuous shaving. Don't rock Knives (16650 Mitchell Creek Dr., Fort Bragg,
hardboard) glued to a '/2-in, to 3/4-in, thick the plane during the stroke. Also, don't get Calif. 95437), offers handmade, high-carbon-steel
plywood base. The miter block should be a running start and crash into the piece, replacement blades for under $20.
made out of a 1-in, thick composite material, and don't chop at it. If you have to chop,
Nevertheless, 16 years ago while searching for
such as particleboard or medium-density either your blade is set too deep or it needs
an alternative to my quick-dulling metal block
fiberboard (cross-grain movement of a to be sharpened.
solid-wood block would affect its accuracy). I usually take the board right to the area planes, I bought a small Japanese plane (bottom
I cut the miter block using a power miter I'm working on so that I don't have to walk photo, p. 15). Designed to be pulled instead of
saw, then screw it to the base so that the around after every stroke or two to check pushed, this plane has a 13/4-in, wide, laminated
block can be easily replaced if it's damaged the fit. Also, with a little practice, you can steel blade wedged with a laminated-steel chip-
or worn. A hardwood cleat is glued to tilt either end of the workpiece off the breaker into a 7¼-in, long wood body (roughly
the base so that the shooting board can miter block to trim the piece for out-of- the same length as my metal block planes).
be hooked over the edge of a worktable square conditions. —S. W Though I often use my Stanley planes, I actually
prefer using my Japanese plane on the job site.
That's because it's lighter (I think of this every
time I lift my toolbox), it fits comfortably into my
Miter-shooting board
construction small hands and my hip pocket, and it's surpris-
ingly durable, having survived even a 35-ft. fall
off a scaffold. Conversely, a short drop to a hard
surface can crack an iron casting, usually at the
throat, which renders the plane useless. I've also
found that the plane's pull stroke gives me more
control than the usual push stroke does (though
I prefer pushing the plane when hogging off a lot
of wood).
But the biggest reason I like the Japanese plane
Y4-Ifl. is edge durability. The secret to this durability is
hardboard the marriage of a thin, extremely hard layer of
glued to base high-carbon steel to a thick, strong layer of soft
steel. The hard steel provides the cutting edge;
the soft steel supports it. The cutting edge on my
Japanese plane has actually shaved the very tops
off nails (though the nail usually wins). I can use
the plane all day, sharpen it that night and be
ready for the next day.
Candle wax Like their metal counterparts, Japanese planes
lx hardwood applied here must be tuned before use. The principles are sim-
cleat allows jig reduces friction. ilar—the blade must be sharpened and bedded
to be hooked over edge of ½in to 3/4-ifl. properly, and the bottom must be flat—but the
work table during use.
plywood base execution is a bit trickier. One excellent source
of information on conditioning these planes is
Japanese Woodworking Tools: Their Tradition,
Spirit Qnd Use by Toshio Odate (published by The
Taunton Press, Inc.). Another wellspring of infor-
Workpiece The miter-shooting mation on tuning and using hand planes, in-
board is used in cluding Japanese ones, is The Best of Fine
tandem with a Woodworking: Bench Tools (also published by
hand plane for
fine-tuning miters. The Taunton Press, Inc.).
Japanese planes like mine cost about $45, com-
Work table parable to the cost of metal block planes. They're
available from a number of suppliers, including
Hida Tool and Hardware Company, Inc. (1333
Block plane San Pablo Ave., Berkeley, Calif. 94702; 800-443-

16 The Best of Fine Homebuilding Drawings: Scott Wynn

Butt mortise planes. Author butt mortise plane (photo above left) cuts level moruses tor hinge teaves and other flusn-mourneu hard-
ware. Lie-Nielsen Toolworks makes the more common metal version (photo above right). Right photo courtesy of manufacturer.

5512) and The Japan Woodworker (1731 Clement

Ave., Alameda, Calif. 94501; 800-537-7820).

The shoulder plane—Block planes are the

workhorses of trim carpentry, but a few other
planes are worth having. I carry a shoulder plane
for trimming rabbets because, although I rarely
use it, sometimes nothing else will do. This is es-
pecially true when fitting new work to old.
Rabbets are easily cut with a router or a table
saw. But all too often new work is plumb or
square, and the old work is not, so the rabbet
needs a custom taper. This is easily accom-
plished using a shoulder plane.
My shoulder plane is an old Record No. 311
"3-in-i" plane (middle photo, p. 15). The 3-in-i
designation refers to three configurations ac-
complished through the use of interchangeable
nosepieces. This allows me to install a long nose
for shooting straight rabbets, a short nose for bull-
nose work in restricted areas, or to remove the
nose altogether for chisel-planing to the end of
stopped rabbets (rabbets that dead-end).
The 3-in-I plane has become expensive since
I bought mine and now costs about $150 (Clif- Chamfering and rounding over. The aumor's collection of chamfer planes and rotiiiuiug-over
planes includes from left to right: a Radi Plane, which cuts roundovers having radii ranging from
ton makes a similarly priced model). If I had 1116 in. to 1/4 in.; a ¼-in, radius rounding-over plane; a 1-in, radius rounding-over plane; and an ad-
to choose an alternative, I'd pick a Record No. justable chamfer plane.
778 or a Stanley No. 12-078 rabbet plane (top
photo, p. 15), which sells for about $65 to $75. With these types of planes, it's especially im- own a rather obscure German wood model (top
Though it doesn't have a chisel-plane mode, it portant to sharpen the edge of the blade square left photo) that I bought for $10 at a closeout sale.
has a bullnose mode and a standard rabbeting with the sides because there is virtually no al- Like a metal plane, its long throat lets the chips
mode. It also has a cross-grain cutting spur, a lowance for lateral adjustment of the blades to pass through and allows you to watch what
depth gauge and an adjustable fence, making it compensate for an out-of-square cutting edge. you're doing. Given the rather rough nature of
probably more versatile than the 3-in-i plane. The blade should protrude in. or slightly less mortising, there is no need to tune these planes
However, it's too large to fit easily into a toolbox, from either side of the plane body. Otherwise the beyond sharpening the blade.
and it usually takes two hands to use, requiring plane will slowly step out from the shoulder of When using my plane, I first lay out the mor-
the use of some clamping system to hold down the rabbet as you plane. tise by outlining it with a chisel, then I make suc-
the work. cessive cuts with the chisel to the approximate
Shoulder planes are generally machined more The butt mortise plane—The butt mortise depth required. At this point, the chips would
accurately than most brands of block planes, so plane is used to cut level mortises for letting in normally be cleared out and the mortise leveled
unless you're having performance problems I hardware, such as hinge leaves, strike plates and with the chisel. But I use the mortising plane. The
wouldn't attempt to tune them. Trying to level dead bolts. Costing $45, the metal version made blade depth is set to the thickness of the hard-
the sole on a shoulder plane may tilt the sole out by Lie-Nielsen Toolworks, Inc. (Route 1, Warren, ware (top right photo) and then the plane is
of square with the sides, which is a hassle to cor- Maine 04864; 207-273-2520) resembles a standard pushed over the chisel cuts, popping out the
rect. Likewise, unless the blade obviously does plane except that it has a handle at each end, chips much like a router plane. Then the plane
not sit flat, I wouldn't touch the blade seat. and its throat is wide open (top right photo). I blade is passed over the whole mortise again to

Finish Carpentry 17
remove any high spots. Lastly, the edges of the over tool called a Radi Plane (which cuts round- the chamfer plane, these planes allow me to
mortise are squared with the chisel. overs having radii ranging from Y16 in. to 1/4 in.); a avoid fussing with a router when I have to pro-
The narrow body of this plane allows it to reach small Japanese-style, 1/8-in, radius, rounding-over duce a simple edge detail or an extra piece of
confined areas, such as mortises for hinges or plane that gives an exceptionally smooth finish; trim. The 1-in. roundover plane is pretty versatile
strike plates in installed jambs with rabbeted and a similar 1-in, radius rounding-over plane. in shaping a variety of radii that you might find
stops. A router is certainly faster for production The chamfer plane allows me to match the on, say, stair nosings or door casings.
work, but if you have to cut a variety of mortises, chamfers that I machine in my shop on the edges The Radi Plane costs about $22. Adjustable
hang a door in an existing opening or install a of deck parts or trim, a boon if I need to produce chamfer planes cost about $50. The rest are usu-
dead bolt in an existing door, the mortising plane an extra part on site. The Radi Plane and the ally priced somewhere in between.
will help speed things up. small roundover plane duplicate the roundovers
produced by some of my router bits, as well as Scott Wynn is an architect/contractor in San
Specialty planes—I cany other planes that can those found on a variety of common moldings. Francisco, Calif He also designs and builds (urn i-
be time-savers (bottom photo, p. 17). I have an The planes also allow me to match the slightly ture. Photos and drawings by the author except
adjustable Japanese chamfer plane: a rounding- rounded edges typically found on flat stock. Like where noted.

I know carpenters who hone their edge

tools by rubbing them on two or three Sharpening plane blades the water-stone grinder. Bench grinders
work okay, but you have to be very
progressively finer sheets of soppy wet/dry careful with them or they'll overheat
sandpaper (ranging from 240 grit to 600 grit), taped or tacked to a scrap the blade and draw its temper, destroying the blade's ability to hold a
of plywood. I've heard of others who sharpen on their belt sanders. My cuffing edge. Bench grinders also hollow-grind the cuffing edge, which
sharpening system is more sophisticated than either of these methods, leaves less metal on the blade than a flat bevel does for supporting the
costs more and takes some time to master, but it produces a superb, cutting edge. Japanese blades need a fiat bevel because their hard,
long-lasting cutting edge that allows me to do top-of-the-line finish work. brittle steel at the edge requires the support of the softer, shock-
Whatever sharpening system you use, I strongly discourage the use of absorbing steel laminated to It. However, any cuffing edge subjected to
honing guides. Feel the blade resting on its bevel and develop the body hard use will benefit from a flat bevel.
mechanics necessary to maintain that angle while sharpening. You may The water-stone grinder (photo below) overcomes all of the
get frustrated at first, but you'll soon learn to get an adequate edge. As shortcomings of the bench grinder. Water-cooled, It never overheats the
your woodworking skills improve, your sharpening skills will, too. blade. And because the revolving water stone is flat, you don't end up
with a hollow grind. The water-stone grinder is ideal for beveling
Sharpening stones—When it comes to producing a sharp, durable nicked or damaged blades, Its only drawback is that the rotating water
cutting edge with a minimum of effort, sharpening stones beat stone wears out of flat quickly, requiring frequent trueing (I do this
sandpaper every time. The selection of sharpening stones on the using my diamond stone). Water-stone grinders cost up to $300, but the
market is overwhelming. Oilstones have for generations been the ones I've seen will handle everything from chisels to planer knives.
mainstay in the West. Recently developed ceramic and diamond stones
promise to combine many of the attributes of other types of stones with The work area—Successful sharpening also depends on the nature of
few of the drawbacks. For serious sharpening, though, I use Japanese the tool-sharpening station itself. If you sharpen at a workbench, the
water stones. Though they wear faster than other types of stones and stone should be 4 In. to 5 in. below your belly button. Unfortunately, the
must be flattened frequently, they cut very fast and produce an typical 3-ft. high bench Is much too high for the average person. If the
incomparable cutting edge. stone is too high, your wrist and elbows will be overly bent, and you'll
Both synthetic and natural water stones are available. Synthetic water have trouble maintaining a constant bevel on a blade. Also, your arms
stones are less expensive and less fragile than natural water stones, but will do all the work without any help from your body weight.
good-quality natural stones produce sharper and longer-lasting edges I was taught to sharpen on the floor. Kneeling on a pad is pretty
than the synthetic ones do. I use synthetic 1200-grit and 6000-grit water comfortable and often brings respite to a back tired from standing for
stones to sharpen American and European plane blades. For Japanese long periods. The floor may be your only alternative on the job site,
blades, I use the 1200-grit stone, an intermediate natural stone called an anyway. If you sharpen on the floor, elevate the stones about 6 in. Mine
"Aoto Toishi" (or blue stone) and a deluxe 8000-grit synthetic finishing sit on a homemade redwood water trough (bottom photos, facing page),
stone, On the job site, I use a diamond stone to touch up all my blades but you can also use a scrap of 6x6.
so that I don't have to deal with water. Whatever surface you work on, mount stops on it so that the stones
Whatever stones you use, buy don't move around during use, or
the best that you can afford. This use one of the manufactured
is especially important for systems that hold and store
finishing stones, where price does stones. You can keep synthetic
equal quality. My water stones water stones in a lidded plastic
range in price from about $15 for tub filled with water so that
the course stones to $50 for the they'll be ready to go. Don't,
fine ones. My fine-grit diamond however, store natural water
stone cost $56. Most fine- stones in water or they'll
woodworking suppliers carry a disintegrate. They also may
full line of sharpening equipment, crack when frozen, even if
including Japanese water stones. they're dry.
I got mine from The Japan
Woodworker and Hida Tools (see flattening the back—The first
addresses in text). step in sharpening a new blade
is to flatten and polish the back.
Grinders—Though you can get Don't worry about polishing
by without a grinder, if you use the entire back, however, just
edge tools a lot you'll eventually The water-stone grinder. Water-stone grinders hone chipped blades at minimum a narrow flat along
want one. The two most common quickly. An attached reservoir continuously dribbles water onto the the cuffing edge (top left photo,
types are the bench grinder and rotating stone to eliminate the risk of overheating the blade. facing page). I usually accomplish

18 The Best of Fine Homebuilding

this using a steel lapping plate and silicon-carbide abrasive powders. the blade, move on to the blue stone (if you're using one) or to your
To use a lapping plate, pour teaspoon of 220-grit abrasive powder finish stone. Don't exert as much pressure on these stones as you did on
on the center of the plate and moisten the powder with a few drops of the 1200-grit stone; they polish more with the slurry formed than by
water (photo 1 below). Lay the blade backside down on the plate, direct contact with the stone. On the finish stone, back off (remove) the
perpendicular to the length of the steel, and rub the blade back and burr by laying the blade flat on the stone and rubbing it back and forth
forth. Try to work all of the powder, including the piles that form at (photo 4 below). Then flip the blade over and polish the bevel.
each end of the plate. The powder will eventually get very dry and fine, Alternate between the bevel and the back, shortening the number of
and the high spots on the back of the blade will start to get shiny (as strokes per turn until you finish with two or three light strokes on each
opposed to the dull gray finish elsewhere). Continue rubbing until all of side. There's no need to polish the edge further with a strop or a buffer.
the silicon carbide is a fine paste (you may have to add a few drops A 30° bevel works best for most planing. This angle is easy to gauge:
of water now and then) and the back has a mirror polish along the The length of the bevel is twice the thickness of the blade. If you're
entire cutting edge. Maintain even pressure at all times, and be careful planing softwoods, a 25° bevel will cut cleaner and easier. Some people
not to lift the blade and round the edge. Backing up the blade with a like to hone a secondary 5° mlcrobevel on the cutting edge. I think this
stick helps (photo 2 below). is self-defeating because the microbevel increases friction at the cuffing
If the back of the blade is reasonably fiat to begin with, I substitute a edge and shortens its life. Besides, after the second or third sharpening,
diamond stone, a 1200-grit water stone and a 6000-grit water stone for a microbevel becomes a macrobevel that requires nearly as much effort
the lapping plate. The water stones must be dead fiat, though. If a gray to sharpen as a full bevel.
oval or large dot appears In the center of the stone while rubbing the Try to create a convex curve across the width of the blade while
blade on it, the stone needs flattening. honing. This feathers the cut, eliminating steps or ridges across the
surface. The curvature of the edge should be virtually Indiscernable,
On the bevel—Now sharpen the bevel. Soak all but the finish water equaling the thickness of the shaving you expect the plane to make.
stones in advance until they stop bubbling (this takes just a few This way the blade will cut across its full width for maximum efficiency.
minutes). Sprinkle Just enough water on the finish stones to create a An easy way to achieve curvature is by alternately applying pressure on
slurry during sharpening. one corner of the blade and then the other while sharpening.
To sharpen, grip the blade between the thumb and forefinger of the While sharpening, check your water stones from time to time to make
right hand (if you are right-handed), wrapping the other three fingers sure they're flat. One way to flatten them is to rub them on wet 220-grit
underneath the blade for support (photo 3 below). Holding the bevel or 320-grit wet-or-dry sandpaper laid on a flat surface (such as plate
fiat on the 1200-grit water stone, glass laid on a Jointer table so
press down on the edge of the that the glass doesn't flex). This
blade with one or two fingers of technique tends to glaze the
the left hand and move the blade stones, however, reducing the
up and down the full length of the cuffing action until the top
stone, gradually working from left particles are worn away. I prefer
to right and back as you stroke. to flatten my water stones with a
Ideally, the cutting edge should be diamond stone; it's quick and
perpendicular to the length of the doesn't glaze the surface.
stone; in practice, it's easier to Before using the water-stone
hold the edge diagonally. Keep grinder, I saturate it with water. I
the stone wet but not flooded. As don't use the bevel guide on the
you stroke, bend your arms and grinder. Instead, I simply feel the
wrists to maintain the blade at the bevel, grinding perpendicular to
proper angle. the edge and moving the blade
Check your progress by holding from side to side to wear the
up the bevel to a light. The honed water stone evenly. Be careful
portion will be shinier than the that the grinder doesn't grab the
unhoned portion. Also, check for blade and throw it, particularly
a burr by brushing your finger when you first set the blade
away from the cutting edge. Once down. I don't use this grinder to
the bevel reflects light evenly A sharp blaae. The of the well-tuned blade (above, left) is flat flatten the backs of blades
(photo far right) and you can feel and polished along its entire cuffing edge. The bevel (above, right) is because it grinds too fast and
a burr along the entire width of honed to a mirror finish. Photos by Bruce Greenlaw. may gouge the blade. —S. W.

2. flattening the blade back. the bevel. 4. Backin off.

Finish Carpentry 19
 Plate Joinery on the Job Site
Quick and easy insurance against joints opening up

by Kevin Ireton

ike most people when they first buy a kerf they fit into, which not only allows room
plate joiner, David Mader, a carpenter in Yel- for excess glue but also provides some play
low Springs, Ohio, wanted to find out how for aligning a joint along its length. This gives
strong plate joints really are. Mader crosscut a biscuit joinely a distinct advantage over dow-
2x4 and reassembled it with a pair of no. 20 eling as an indexing technique.
plates (the largest size available), one over the Plate joinery works in hardwood, softwood,
other. After letting the glue diy, Mader tried to plywood, particleboard and even in solid-sur-
break the 2x4 over his knee. He couldn't do it. face countertops (using Lamello's clear plas-
Convinced of plate joinery's strength, Mader tic C-20 biscuits). Plates can be used in edge-
proceeded to use his plate joiner to butt-join to-edge joints, butt joints and miter joints.
custom flooring that wasn't end-matched.
Often considered the province of shop-bound Joinery comes to the job site—Over the
woodworkers, plate joinely, it turns out, is be- past 15 years, plate joinely has proven itself
ing used more and more by carpenters on the strong enough and accurate enough to earn a
job site (photos at right). Plate joinery and bis- place in many woodworking shops, where it
cuit joinely are the same thing, and in this ar- competes with other joinely methods such as
ticle I'll use the terms interchangeably. doweling, splining or mortise-and-tenon join-
The basic idea behind plate joinely is sim- ery. The merits of plate joinery relative to
ple: plunge a 4-in, circular sawblade into a these other methods can and have been de-
piece of wood, and you get a crescent-shaped Slotting in pLace. 10 aeai a small bated. But most carpenters in the field don't
slot. Make a series of these slots along the piece, this carpenter installed the plinth block
first and slotted it in place (above). In the pho- enjoy the luxury of a fully equipped shop, and
edges of two boards that you want to join. In- tos below, he adds a biscuit, then the casing. often their only joinery options are whether to
sert glue and a football-shaped wooden spline use nails or screws. That's why plate joinely
into each slot on one board. Insert more glue adds a valuable technique to a carpenter's
into each slot on the other board, then press repertoire. After all, a cabinetmaker can suc-
the two boards together. Water from the glue cessfully argue that a biscuit-joined face frame
causes the splines to swell, making a strong, is not as strong as one joined with mortises
tight joint. and tenons. But no one will argue that adding
a biscuit between two pieces of mitered cas-
Biscuit-joiner basics — The typical biscuit ing (photos, p. 22) won't strengthen the joint or
joiner is a cylindrical machine (drawing facing greatly improve its chances of weathering
page), about 10 in. long, and weighs between changes in humidity without opening up.
6 lb. and 7 lb. It has a D-shaped handle on top Joint strength isn't the whole story, though.
and a spring-loaded faceplate in front with an A biscuit joiner is very portable, taking up less
adjustable fence. When the tool is pressed room in a toolbox than a circular saw. And it's
against the workpiece, a 4-in, carbide-tipped extremely fast. Cutting slots and adding bis-
blade extends through a slot in the faceplate cuits to a mitered door casing might require
and scoops out a kerf in the workpiece. You 30 seconds. Admittedly, even that little time
can adjust the distance between the kerf and can be significant when multiplied by a house-
the face of the workpiece, but any closer than ful of doors and windows. You might consider
in., and the biscuit, or plate, may pucker it worthwhile, though, if you've ever been dis-
the surface of the wood when it swells. You appointed when returning to a job to discover
can also adjust the depth of the kerf to fit the gaps in joints that fit perfectly when you nailed
size of biscuit you're using. them up.
Biscuits come in three basic sizes (all three
are arcs of same circle): #0 is about in. Who's using them where?—Stephen Sewall,
wide and in. long, #10 is 3/4 in. wide and a builder in Portland, Maine, feels so strongly
2'/8 in, long, and #20 is 1 in. wide and 21/2 in. about the advantages of biscuit-joined trim
long. Biscuits are made of beech with the that he seldom installs trim without biscuits.
grain oriented diagonally to the length, mak- On a recent job where he didn't have his bis-
ing them very strong across their width. Bis- cuit joiner, Sewall nailed up the side casings,
cuits are also compressed so they'll fit easily but left the head casings loose so that he
in the kerf and then swell once the glue hits could add the biscuits later.
them. All biscuits are slightly shorter than the Sewall also says that biscuit joinety has

20 The Best of Fine Homebuiidzng Drawings: Bob Qoodlellow

Biscuit joiner and setup for slotting narrow stock Front view

4-in, circular
sawbiade.

Faceplate

Scrap

Base

Quick-set stop mechanism

to adjust depth of cut for
different biscuit sizes Dust-ejection slot
Anti-kickback pin

When slotting stock narrower than

the anti-slip points, clamp or nail
scrap blocks on either side ot the
stock to engage points.

After marking biscuit

locations on both shelf
and upright, the shelf can
be clamped flat to the In places where the top or bottom of the face frame will be
upright and serve as a hidden, stock narrower than 2¼ in. can be used by offsetting
fence against which to the slot toward the hidden side and trimming the biscuit flush
register the joiner. after glue-up.

flmsh Carpentry 21
It only takes a With the tool ana me trim registering against makes short work of cutting slots. Icy then adding a
biscuit spline between two pieces of mitered casing, he prevents the joint from opening as a result of wood shrinkage.

Biscuit joints in cabinetry Flooring applications for biscuit joinery

Miter

7
joints

7
Butt joints where flooring Butt joints - - -
isn't end-matched where flooring
meets border

Casing for round-top windows

Screw or nail segments

to plywood.

Position slots so biscuits won't show when radius is cut.

Assemble with 5-minute epoxy. Can be worked 1 hr. after
final segment is glued.

22 The Best of Fine Homebuilding

made building cabinets on site a lot easier Kimberly Rd., Davenport, Iowa 52806; 3 19-391- p. 21). These points keep the tool from slipping
(drawing facing page). When installing a fixed 2386) makes biscuits out of particleboard that during a cut (some joiners employ rubber
shelf in a cabinet, which he used to do by are 1-in, wide by in. long and for which bumpers or pads rather than steel points).
routing dados in the sides to house the ends you cut kerfs with a 6mm slot-cutting router
of the shelf, Sewall can now biscuit-join the bit (available from Woodhaven and from MLCS, But wait, there's more—It's easy to think of
shelf to the carcase faster than he can change Ltd, P. 0. Box 4053, Rydal, Pa. 19046; 800-533- other job-site uses for biscuit joinery: mitered
bits in his router. When biscuit joining 9298). With these same bits you can also use ceiling beams, jamb extensions on doors and
shelves, Sewall often clamps the shelf flat your router to perform conventional plate join- windows, return nosings on stair treads. You
against the upright and registers the biscuit ery, but the cutter is exposed and you don't could even use biscuit joinely where two clos-
joiner against it as he cuts the slots (see drawing, have a faceplate, so you lose some of the safe- et shelves meet at a corner and avoid having
p. 21). Biscuits will also work in Y2-in. stock, ty and convenience of a plate joiner. to screw a cleat to the underside of one shelf
like the 1/2-in. Baltic-birch plywood that Sewall As it turns out, though, you can often get to support the other (drawing facing page).
uses to make cabinet drawers. away with using a standard biscuit in a narrow Do be careful, though, if you decide to buy
Foster Jones, a partner in Maine Coast Build- rail by offsetting the slot to the outside of the a biscuit joiner. Beware the "Law of the Instru-
ers of York, Maine, admits that using biscuit frame and trimming the biscuit flush (draw- ment." This is a theory in psychology that
joinety adds to the cost of a job and says his ing, p. 21). In most cases the exposed ken states: if you give a small boy a hammer, ev-
company usually decides before starting a proj- and biscuit are either pointing down at the etything he encounters will need hammering.
ect whether to use biscuits. If they do use floor and can't be seen, or are covered by a There may be some things that really don't
them, though, they don't just use them on countertop. When cutting slots in narrow stock, need to be joined with biscuits.
miter joints. They use biscuits to join inside it's best to clamp or nail scrap blocks on
and outside corners of baseboard, to join base- either side so the steel points on the joiner's Kevin Ireton is editor of Fine Homebuilding.
board to door casing and to join door casing to faceplate have something to grip (drawing, Photos by the author.
plinth blocks (photos, p. 20).
When laying a hardwood floor, the carpen-
ters at Maine Coast Builders use biscuits to
join the picture-frame border around a fire-
place hearth (drawing facing page) and to join Biscuit-joiner comes from the German word lamelle,
the border to the flooring that butts into it. manufacturers meaning "thin plate"), and despite the
They even use the biscuit joiner as a trim saw Looking for a faster and more accurate fact that nine other companies now
to trim the bottoms of door casing so that alternative to doweling, a Swiss make plate joiners, the Lamello
cabinetmaker named Henry Steiner machine is still considered by many
flooring will fit beneath it. woodworkers to be the "Cadillac" of
developed plate joinery (also called
Jones also uses biscuits when fabricating biscuit joinery) back in the 1950s. plate joiners. The following list
trim for round-top windows. Using biscuits Steiner founded a company to includes all the companies that make
and five-minute epoxy, he joins mitered seg- manufacture slot-cutting machines, plate joiners. Prices range from under
ments of straight stock end to end in a rough called plate joiners, and the plates (or $150 to over $600. So If you decide to
semicircle (drawing facing page). He usually biscuits) to go with them. The company buy a plate joiner, be prepared to shop
screws the segments to a piece of plywood is called Steiner Lamello, Ltd. ("lamello" around. —K. i
rather than clamping them. The five-minute
epoxy lets Jones work with the piece after less Delta 32-100 (bench-top model) Porter-Cable 555
than an hour of diying time. Delta International Machinery Corp. Porter-Cable Corp.
Because biscuit joinely relies in part on the 246 Alpha Dr. P. 0. Box 2468
biscuits' capacity to absorb water from the Pittsburgh, Pa. 15238 Jackson, Tenn. 38302
glue and swell up to form a tight joint, you (800) 438-2486 (901) 668-8600
may be wondering how well the system works
Elu Joiner/Spllner 3380 Ryobi JM100K
with epoxy, which isn't a water-based glue. Ryobi America Corp.
Black & Decker, Inc.
Jones and Sewall wondered, too, and broke 5201 Pearinan Dairy Rd.
P. 0. Box 798
apart joints that they had assembled with ep- Hunt Valley, Md. 21030 Suite 1
oxy. Both found the joint to be just as strong (800) 762-6672 Anderson, S. C. 29625
as those made with yellow glue, which has a (800) 226-6511
water base. In fact, Bob Jardinico at Colonial Freud JS 100
Saw, sole U. S. distributor for Lamello joiners Freud, Inc. Skil 1605:02
(see the sidebar at right for address), recom- P. 0. Box 7187 Skil Corp.
mends epoxy for biscuit joinely used outdoors. High Point, N. C. 27264 (subsidiary of
(919) 434-3171 Emerson Electric Co.)
Makes you wonder if biscuits and epoxy aren't
4300 West Peterson Ave.
the way to keep mitered handrails on exterior Chicago, II. 60646
Kaiser Mini 3-D
decks from opening up. (dist. by W. S. Jenks & Son) (312) 286-7330
1933 Montana Ave. NE
Plate-joining face frames—A common com- Washington, D. C. 20002 Virutex 0-81
plaint when assembling face frames (or cabi- (202) 529-6020 Rudolf Bass Inc.
net doors) with biscuit joints is that the rail 45 Halladay St.
must be wider than the slot for the smallest Lamello (3 models) Jersey City, N. J. 07304
biscuit. Otherwise the biscuit will show. This (dist. by Colonial Saw Co., Inc.) (201) 433-3800
limits you to a rail that's at least 2'/4 in. wide. P. 0. Box A
Responding to this complaint, Lamello recent- Kingston, Mass. 02364
ly introduced face-frame biscuits (H-9) that (617) 585-4364
are '/2 in. wide by 11/2 in. long. But such a
short biscuit means you have to switch to a
3-in. sawblade (also available from Lamello).
A company called Woodhaven (5323 West

Finish Carpentry 23
 Survey of Finish Nailers
The new trim nailers are lighter, smaller and more powerful than ever

by Jim Britton

hen I broke into the trades in 1973, I could count the different brands
of pneumatic trim nailers on one hand. Back then, plenty of professional
carpenters hung doors, trimmed windows and affixed baseboards with a
claw hammer and a pouch full of 8d finish nails. Not anymore. Today, a
contractor simply cannot compete without a decent complement of air
nailers. That's because a pneumatic nailer not only speeds the work, it also
makes for a better job. To tell the truth, I don't even keep hand-driven nails
on my truck anymore.
To fill the demand for trim nailers, tool companies new and old, domestic
and import, have offered up more than two dozen finish nailers for the
pneumatically enhanced carpenter to choose from. And the tools aren't
just for pros. The weekend builder/woodworker can also take advantage of
the increased quality and convenience of a finish nailer without necessarily
forking over the $400 or so that the most expensive nailers fetch. Imported
tools that sell for half of that amount can let anybody join the ranks of the
pros and leave the hammer tracks behind.
In this article I'll talk about the pros and cons of the details that are a part of
every finish nailer. And I'll also give you my impressions of the 23 nailers I
evaluated in the course of my work on various job sites earlier this year.

First, choose between 15-ga. or 16-ga. nails—Pneumatically driven fin-

ish nails come in two sizes: 15 ga. and 16 ga. Both kinds are available in
lengths ranging from 3/4 in. to
Angled for compactness. Because of their angled magazines, 15-ga.
in. The 15-ga. nails are slightly larger in nailers fit into the same confined spaces as smaller 16-ga. nailers. This
dia.—about 0.069 in. to 0.072 in. The 16-ga. nails are about 0.057 in. to Fasco BA-65 loads from the side; a magnetic strip helps to retain the nails.
0.060 in. Which should you choose? I use my 16-ga. nailer for installing the
thinner trim materials, such as 3/8-in, thick baseboards. I'll also use the 16-ga.
nailer for affixing 14-in, trim, but only when the nails aren't asked to span a
gap or a soft material, such as drywall, in order to get to the substrate. For
example, if I'm hanging a door and the nails have to span a gap between the
jamb and the trimmer, I'll use the 15-ga. nailer. The extra beef in the larger
nail is better-suited to the cantilever action of spanning a gap, especially
when the nails have the variable loads imposed by a door on them.
The thinner the nail, the less likely it is to split wood. Therefore, it makes
sense to use the 16-ga. nailer when running casings or window aprons. But
there's a catch: Thinner nails are more likely to be deflected off course by
the grain pattern in the wood. This result usually happens with chisel-point
nails when the cutting edge of the nail is parallel with the grain. Solution:
Rotate the nailer 90° so that the cutting edge is perpendicular to the grain.
In most cases, 16-ga. nailers have straight magazines (bottom photo), and
15-ga. nailers have angled magazines (top photo). The angled magazine
allows the generally larger 15-ga. nailer to behave more like a smaller tool,
poking into corners where it otherwise might not fit.

Two basic types of magazines—The strips of nails are held in the tool's
magazine, the long, thin compartment below the handle. Magazines are ei-
ther closed or open, rear load or top load. Most of the variations take place
in the 16-ga. nailers because the rectangular strips of nails are easy to load 16-ga. nallers are use smallest in this class. Because they fire a
smaller nail than their 15-ga. brothers, 16-ga. nailers can be packed into a
any which way. The angled, 15-ga. strips of nails, on the other hand, are al- smaller package. The Paslode 3250 F16 (above) has a top-load magazine.

24 The Best of Fine Homebuilding

most always fed into the magazine from the rear. To load nails into a tool The job determines the contact type. On profiled trim, a rounded
with a closed magazine, you pop the latch on the magazine's cover, pull loop contact such as that on the Airy ATF 0350 (photo 1) slides along
back on the magazine's spring, load the nails in from the side and close the work without denting the trim. A cushioned contact (photo 2) is
the cover. The magazine's spring is now ready to advance the nails into the best used on flat stock that has to be pressed against the substrate be-
firing position. This system works well if the magazine has a strip magnet fore nailing. The sharp prongs on a bent sheet-metal contact (photo 3)
built into it to retain the nails as they are loaded (top photo, facing page). will ding soft trim, such as pine. A flip-open noseplece (photo 4)
But without the magnet, nails can sometimes bunch up and fall out as the makes for easy jam removal.
spring is returned to position. My favorite magazine style is the top-load de-
sign (bottom photo, facing page). This type allows for quick loading be-
cause there are fewer parts to fiddle with, and the nails are held in their
track as the spring is returned to position.

Some contact points are better than others—A pneumatic nailer en-
gages the work by way of the contact point at the business end of the tool. As
you push down on the nailer, the contact point disengages the tool's safety
mechanism so that pulling the trigger fires the nail.
When I'm running trim that has a profile to it, I prefer a contact point like
the one on the 16-ga. Airy ATF 0350 (photo 1, right). This rounded, horse-
shoe-shaped loop tucks into milled details well, allowing the nail to be dri-
ven all of the way below the surface of the trim. When I do production work
with this type of contact, I slide the nose of the nailer along the trim, indexing
the contact against the milled edges. The method works because the round-
ed metal loop rarely mars the wood.
Cushioned contacts (photo 2, right) also reduce marring, but on profiled
stock a cushioned contact may make it difficult to drive the nails all of the
way. Cushioned contacts are best used for installing flat stock, such as base-
boards, where you want to be able to press the workpiece tight to the dry-
wall. Remember, all cushions are removable, and this feature may allow for
better nailing of detailed moldings. My advice is to avoid the bent sheet-met-
al contacts that some manufacturers favor (photo 3, right). The prongs on
this type of contact gouge and ding all but the hardest of woods.

Consider the firing sequence—The nailer's contact is part of a safety

mechanism that prevents accidental firing. In order to fire a nail, a sequence
of steps has to be followed. The most common firing sequence is the contact
trip. This arrangement allows the tool to fire anytime that the trigger is pulled
and the contact piece is depressed. So if you keep your finger on the trigger
and simply push the tool down, it will fire. When done repeatedly without re-
leasing the trigger, this process is called bounce nailing, and it is often used
by framing crews to install sheathing.
My view is that bounce nailing is of dubious value, especially for trim
work where accurate nail placement is critical. Most nail jams occur during
bounce nailing. It should be noted that contact-trip nailers don't have to
be bounce nailed. They will drive a single nail. Another firing sequence is
called the sequential trip. With this method, the contact must be depressed
first, and then the trigger can be pulled to drive a single nail. The trigger
must then be released before another nail can be fired. Of the tools sur-
veyed here, most have contract-trip sequences. Only the Stanley-Bostitch
N6OFN and the Duo-Fast LFN-764 use the sequential-trip firing sequence.

Nailer-survey ground rules

As an interior-trim contractor, my job site is my laboratory. The tools evalu- Adjustable exHaust ports are a convenience. If you have to work in
ated in this article were used on a variety of projects. Some were fancy stain- a dusty area, you'll appreciate the exhaust port on this Stanley-Bostitch
N6OFN. It can be directed with a quick twist.
grade installations, and others were paint grade, with plenty of medium-
density fiberboard (MDF) trim.
For the power test, I drove nails through 3/4-in. MDF into 2x Douglas fir. To Cordless nailing
see if the tools left dents in trim, I used them on pine casings. My air source
for pickup jobs.
Paslode's Impulse nail-
was a twin-tank professional compressor rated at 3.5 CFM © 125 psi, with no er is powered by an
more than 75 ft. of hose. With the exception of the two oilless nailers, I put internal-combustion
two drops of 20-weight nondetergent oil into the air intake of each nailer pri- engine, so it needs no
or to use (never use detergent oil in a nailer; it will degrade the 0-rings). compressor. It's ideal
The nailers were tested in a temperature range of 20°F to 75°F, with no for quick jobs that
temperature-related failures. However, I do recommend warming nailers don't warrant schlep-
ping compressors and
before winter use. Most makers will advise a thinner oil in colder weather. air hoses.
Also, it is important to note that I'm a middle-aged man of medium build,
with average strength and medium-size hands. You should make your judg-
ments accordingly.

Tool survey begins on the next page.

Finish Carpentry 25
 Airmark AT-64 800-999-9195 Atro Monza 64 800-284-5347 Duo-Fast LFN 764 800-752-5207
List price* $185 I
List price* $400 List price* $340
Weight 5.1 lb. Weight 4.5 lb. Weight 5.1 lb.
Height 111/sin. Height lOV2in. Height 9in.
Nail Nail Nail
length ill. length 11/4-21/2in. length %-2 in.
Nail Nail Nail
capacity 100 capacity 100 capacity 120
Adjustable Adjustable Adjustable
depth No depth depth No
75 psi Power test° 1 Power test° 70 psi
This medium-weight nailer has decent balance Along with Senco, Atro is the only other compa- The LFN is small and feels a little heavy. But its
and a comfortable grip. Its closed side-load mag- ny to offer oilless technology. This Italian nailer is compact size and comfortable suedelike grip
azine is simple and sturdy, but it lacks a magnet billed as a 16-ga./15-ga. tool, but I couldn't get it make this nailer easy to maneuver. The closed
for retaining nails during loading. At 75 psi this to behave as such. The open rear-load magazine magazine with magnetic retention is as nice as
nailer clobbered 2-in, nails through 3/4-in. MDF works with 16-ga. nails but not with the fatter they come. My only beef with this nailer is that
and into fir. The major drawback to this import is 15-ga. nails. I couldn't find any 15-ga. Atro nails, the nails exit the tool well behind the contact
the bent sheet-steel contact piece. It marred pine so I loaded up the tool with some Stanley- piece. I got familiar with this quickly enough but
and tended to dig into the soft wood if the tool Bostitch nails, which had the same 25° angle. still had some trouble nailing casings adjacent
was slid along the trim to the next nailing point. But the Stanley-Bostitch nails wouldn't advance to hinges. The Duo-Fast brand nails come in
properly. Evidently, their heads are too big. A call strips of 60, for a total load of 120. This tool is sol-
Airy ATF-0350 800-999-9195 to the Atro folks confirmed this. They told me id. It'll outlast your pickup.
Woodtek 864-374 800-645-9292 that only 15-ga. Atro nails would work in the tool.
List price* Unfortunately, Atro doesn't make any 15-ga. nails Fasco FN7O 800-239-8665
Weight 4 at this writing. In light of this Catch-22 circum- List price5 $394
Height stance, consider this to be a 16-ga. nailer. Weight 5.5 lb.
Nail The Monza comes with a large, cushioned-loop Height 111/oin.
length 3/4-2in. contact. It's so large that it obscures the work and Nail
Nail keeps the nailer from getting close to detailed length 3/4-23/41n.
capacity 100 moldings. Atro does offer an optional contact. It Nail
Adjustable provides better visibility, but it increases marring. capacity 100
depth No
The depth adjustment is a hassle requiring Adjustable
Power test° 100 psi wrenches. But the nosepiece has a single clip for depth No
This nailer is compact and light, and has the un- quick jam removal. Power test° 70 psi
canny ability to drive both 16-ga. and 15-ga. nails. This Italian nailer is a heavyweight among the
Used exclusively on a stain-grade job, I found Craftsman Sears stores 16-ga. tools, but terrific balance offsets its weight.
this tool to have the best noncushioned contact List price* $200 It has remarkable power. I used it on one of my
piece of any nailer tested (photo 3, top right, Weight 5.4 lb. stain-grade jobs and had it countersinking 2-in.
p. 25). 1 could slide this nailer down the wood Height 10 in. nails into solid pine at a surprisingly low 65 psi.
without losing contact and have a perfectly Nail The Fasco performed similarly in the MDF power
placed nail almost every time. On the other length 3/4-2in. test. My only gripe is with the too-stiff contact
hand, I did not care for the plastic-composite Nail spring, which complies with tougher European
magazine. Static electricity seems to attract dust capacity 100 safety standards. With the tool overhead, I had
to the magazine, where it then sticks. Adjustable difficulty depressing it every time. This nailer uses
This tool has a single-screw nose for jam re- depth No
a semiclosed magazine with magnetic retention.
moval, which is slower than other flip-open mod- Power test° 90 psi Its sturdy construction and careful machining
els. And the tool comes with a spare piston-driver Sears' nailer uses a closed side-load magazine suggest this tool will last for many years.
assembly, a handy part to have on hand. I wish that required careful alignment of the nails dur-
other manufacturers included extra drivers in ing loading. But once the nailer was loaded, I
their packages. was able to zoom on trim. The cushioned con-
tact prevented marring, but it obscured the view
a little bit. The contact wants a daily drop of oil
on the point where it travels through the nose to
the trigger. At $200, this Taiwanese import is a
good value.

*Refail and mail-order outlet prices are often far less than list price.
°The minimum line pressure required to countersink a 2-In, nail through ¾-in. MDF into Douglas fir.

26 The Best of Fine Homebuilding

 -__________

16-GA. NAILERS:

Grizzly G2413 800-541-5537 Hitachi NT65A 800-706-7337 Paslode 3250 F16 800-323-1303
List price* $225 List price* $749 List price* $360
Weight 5.9 lb. Weight 4.4 lb. Weight 4.5 lb.
Height 10 in. Height 10'/sin. Height 101/sin.

Nail Nail Nail

length length 1-2'/2in. length
Nail Nail Nail
capacity 100 capacity 150 capacity 150

Adjustable Adjustable Adjustable

depth No depth Yes depth Yes

Power test° 110 psi Power test° 90 psi Power test° 95 psi

This chunky 16-ga. nailer is as heavy as it looks. This handsome, light, well-balanced tool has a Light and well-balanced, this nailer feels great in
Compared with other tools in its price range, the good deal of power. I drove nails at as little as your hand. It, too, has a 150-nail, top-load maga-
Grizzly is crude. When I first hooked it up, it 80 psi, and the contact tip is gentle on the work. zine. Paslode has used this design for years, dat-
would not set a 13/4-in. nail through 9/e-in. MDF, The top-load open magazine holds a bonus sup- ing back to the venerable Mustang nailer. The
through drywall and into a fir stud. I filed the ply of 150 nails, but my test nailer had a major Paslodes like a higher line pressure, and I found
nose some, which simulated lengthening the dri- problem. The nail pusher jammed nearly every 100 psi to be the reliable minimum.
ver, but still the nailer had to be held perfectly time that the tool got down to about 50 nails. It A clever quick-release nosepiece allows the
square to the work to drive and set the nail. The would work very well for 100 or so nails. Then nose to be opened in a heartbeat to clear a jam.
Grizzly's slippery enamel finish and poor bal- the pusher would slip off of the nail strip and jam. The only thing that I didn't like about this nailer
ance make it hard to hold. And the bent sheet- It appears that the spacers that set the width of was the side play in the rionmarring contact
steel contact mars work and digs in. There are the nail slot are too thick. With extra space in the piece. It wiggles a bit, making fast arid precise
plenty of better nailers in this price range. magazine, the pusher can leave the nails. nail location a little tricky.
According to Hitachi, they've taken steps to cor-
Haubold SKN 64 AuG 800-437-9818 rect the situation, and they will fix older nailers at Penn State ANK3 800-377-7297
Kihlberg SKN 64 AuG 800-437-9818 no charge. List price* $240
Hilti FBN212A 800-879-8000 Weight 5.6 lb.
List price* $575
Jet JDPN-671.4 Height 10 in.
Weight 4.9 lb. List price* $331 Nail
Weight 5.7 lb. length %-2½in.
Height 9% in.
Height Nail
Nail capacity 100
length 13A6-2'/2in. Nail
length Adjustable
Nail depth No
capacity 105 Nail
capacity 110 Power test0 100 psi
Adjustable
depth No Adjustable To my way of thinking, the Penn State ANK3 is
Power test° 110 psi depth No -
another Stone-Age nailer. It appears to have been
Power test0 100 psi chiseled out of an iron ingot. Like the Grizzly
This beautifully cast nailer is a gorgeous example I

of German workmanship. However, it suffered This tool is a counterfeit of the Flaubold, and G2413, Penn State's 16-ga. nailer has trouble set-
the same malady as the other tools using a bent therefore it suffers from the same maladies as the ting nails consistently. you cant the nailer to
sheet-steel contact. Specifically, the tool did not much prettier German nailer. The bent sheet- the side, the nails sometimes end up protruding
set nails when the nailer was slightly canted. The steel contact mars and digs into wood. The han- from the work. The ANK3 also has a sheet-metal
contact cushion compounded this problem. By dle is too close to the magazine, making loading contact tip, which digs into the work and mars
removing the cushion, nail penetration was bet- inconvenient. Although it is a copy of a German the wood. This nailer is a poor value.
ter, but the contact piece marred the pine. This tool, the casting and machining are not even
nailer has an open, top-loading magazine. It is close. Like the Haubold, the tool is compact and
nicely machined, and the nails slide smoothly a bit heavy but well-balanced. Tool survey continues on the next page.
inside it. But ergonomically, the magazine is too
close to the handle. It's hard to hold the tool by Good nails, bad nails
its handle while loading it. Tool manufacturers claim that you should
The power test revealed another weakness but use only their nails. Nonsense. I had six
more likely confirms my suspicion that the con- brands of 16-ga. nails on hand for my tests,
tact prevents proper nail driving. I had to set the as well as generic nails from Asia, and they
line pressure to 110 psi to drive 2-in, nails into were all Interchangeable with the 16-ga.
the test medium. With the cushion removed, I nallers. But nails are not of equal quality.
was able to reduce the line pressure. The tool is With one exception, the best nails are made
medium weight and well-balanced. It feels great
in America. The only other nails that
in the hand all day long. Shoots all brands of compare are the Fasco nails (or Beck nails),
16-ga. nails. which come from Austria. Steer clear of nails
that have voids in the strips (photo right). A
void will cause a mlsflre, followed by a jam.
Also, make sure the strips are straight. A
curved strip can cause friction, hindering the
nails as they advance. A jam can result.

Finish Carpentry 27
 _________________________

Airy ATH-0565T 800-999-9195

List_price* Penn State ANK5 800-377-7297
Weight Woodtek 832-392 800-645-9292 Weight
rfeight List price* $398 Height
Nail Weight 5.1 lb. Nail
leng leng
Height 103/4in.
Nail Nail
Nail
length
Adjustable Adjustable
depth No Nail
capacity 100 depth No
Power test" 80 psi Power test° 65 psi
Adjustable
The 1640 is a curious nailer: It's both heavy and depth
Holy smoke! This nailer easily won the most-pow-
compact. The firs: things that I noticed when us- Power test° 110 erful honors. It nailed the power test at 65 psi,
ing the tool were the faraway feel of the trigger, This import from Taiwan is a surprisingly good with a 2½-in, nail through a double layer of MDF
and its long travel. At I would be pulling the tool. Its medium weight, excellent balance and into a fir stud. The straight, closed magazine in-
trigger and waiting for the nail to shoot. Then I smallish head make the nailer maneuverable. It cludes magnetic nail retention. It opens conve-
would pull a little farther, the nail would fire, and has an excellent, nonmarring, cushioned con- niently with a nicely located front-release latch.
I'd flinch. After I'd gone through a box of nails, tact piece that provides good visibility. The open, At a whopping 7.7 lb., it is also the heaviest tool of
however, I got used to the tool. rear-load magazine was my favorite of all of the the bunch. But its excellent balance and a suede-
This nailer has a simple, top-loading open mag- angled nailers tested. The magazine is simple like grip make it easy to hold.
azine that's easy to use. I liked the exposed push- and loads easily. The nails almost glide into their As with the Duo-Fast LFN, this tool has the nail
er spring because it's easy to clean. The contact slot. The nose opens with a clip for easy jam re- exit the nose behind the contact piece, which
piece is the common front loop, but it's made moval. My only complaint about this nailer was makes for some awkward nail placement. The
out of flat stock and doesn't slide along the work that its tiny screws would loosen and fall out. cushioned contact, with a built-in hard-rubber
very well. Still, the tool has good power and a This problem could be solved by using a thread- pivot, does mar some. The pivot allows the con-
no-nonsense likability. locking fluid prior to operation. tact to be square to the material, even when the
nailer isn't square to the material. Because of its
Spotnails FBi 632 800-873-2239 DeVilbiss AFN 5 800-888-2468 great power and solid workmanship, I think this
List price* $268 List price* $280 tool is best-suited to a shop situation where its
Weight 3.4 lb. Weight 5.9 lb. lack of maneuverability won't count against it.
Height 81hin. Height llin.
Nail Nail
length in. length 1-2½in.
Nail Nail Weight
capacity 100 capacity 100
Height
Adjustable Adjustable
depth No depth No Nail
leng
Power test" 85 psi Power test0 110 psi
Nail
If you like compact, then you'll love the 1632. The U. S.'s largest maker of air compressors now capacity 100
This nailer is by far the smallest one in the survey. imports the AFN 5. It is a Senco knockoff, and a Adjustable
It's about the size of most 18-ga. brad nailers. But pretty good one. It is medium-heavy, but well- depth No
it still has incredible power. Pulling the nail push- balanced. The mar-free cushion tip tended to fall Dower test0 110 psi
er back and over the loaded nails is awkward at off the contact, which made the nailer jam fre- The Italian penchant for durability ann ciesign is
first, but I found it easier to reload after I'd run quently. I solved the problem by gluing the tip evident throughout this nailer. It is a heavy tool,
about 1,000 nails through the tool. The contact back on with a dab of construction adhesive. If but it is well-balanced. The closed magazine in-
piece is too small, and as a consequence, it mars the tool does jam, it can be easily cleared by f lip- cludes a magnetic retention strip.
the pine trim. Fortunately, the tool has a soft con- ping open the front of the nosepiece. The open, This nailer won't shoot Senco nails, which is a
tact spring. rear-load magazine is made of steel, which is potential drawback because Senco nails are
reminiscent of the older Senco SN1. This nailer widely available. I got out my protractor to see
loads easily and has reasonable power. For hard why. The Fasco nails read approximately 34Y2°.
materials, though, you have to increase the line The Senco nails read about 35'½°. Interestingly,
pressure to the maximum. the Fasco worked fine with the All-Spec nails,
which are also collated at 34½". The BA-65 has
good power, but I had to raise the line pressure to
the maximum for hard materials. The air-supply
port is smartly angled for easy connection. As I
did with its little brother, I found the BA-65's con-
tact spring to be too stiff.

* Retail
and mail-order outlet prices are often far less than list price.
°The minimum line pressure required to countersink a 2-in, nail through "/4-In. MDF into Douglas fir.

28 The Best of Fine Homebuilding

 _____________________

Makita AF 631 714-522-8088 Senco SFN 40 Paslode Impulse 800-323-1303

List price* $425 List price* $580 List price* $625
Weight 5.6 lb. Weight 4.9 lb. Weight 61b.
Height 121/8in. Height Height 121/2 in.

Nail Nail Nail

length length in. length
Nail Nail Nail
capacity 100 capacity 104 capacity 100
Adjustable Adjustable Adjustable
depth Yes depth Yes depth Yes

Power test° 90 psi Power test" 80 psi Power test° n/a

The new Makita nailer works beautifully and The SFN 40 replaces the cumbersome SFN 2. This The Impulse IM25OF, the only cordless nailer, is in
looks great. It is similar to the Stanley-Bostitch new nailer has all of the features of the SFN 1 a league of its own. This 16-ga, nailer uses a liquid
N6OFN and uses the same nail strips. Makita plus greater nail-length capability at a cost of on- hydrocarbon fuel to drive the engine that pro-
will have its own nails available when it intro- ly an additional half-pound in weight. The depth pels the nail. Although it looks bulky, the nailer
duces the tool in the fall. While a bit heavier than adjustment is a brilliant self-locking cam. The weighs only 6 lb. with fuel, battery and nails.
the Bostitch, the Makita is well-balanced. It has a SFN 40 has a quick-change exhaust port for di- As this tool evolves, it is sure to get smaller. But
comfortable handle, a nice trigger and a nine- recting the exhaust. For all of the performance, in its present configuration, it does not get into
setting depth adjustment. This nailer has the best this tool is quiet and has almost no recoil. It's corners very well. This tool's niche is the quick
quick-open nosepiece that I have seen. It has no among the top tools. job, the callback or a short punch list—any time
screws or clips. To open, simply push the cover that packing the compressor out, running some
up and open. Stanley-Bostitch N6OFN 800-556-6696 hose and running the cord would get to be too
A clever new feature is in the magazine. With Stan-Tech SDN 15 BR 800-343-1234 much of a hassle (bottom photo, p. 25).
an open, rear-load magazine, it looks pretty ba- List price* $ I

sic. But hidden away is an override that will not Weight 4.4 lb.
allow the tool to fire when it is empty. This fea- Height 113/8in.
ture will save on tool wear. The feature also saves Nail
unneeded puttying. length 11/4-21/2in.
Choosing a nailer
Makita has paid great attention to detail. The Nail Selecting the right nailer depends on the
workmanship is as high as I have seen. I enjoyed capacity 100 type of work you plan to do with it and
the testing so much that I trimmed three houses Adjustable how often it will be used. Will you use it
depth Yes occasionally in a home shop, or every day
with this prototype.
Power test" 100 psi ] as a professional? If you plan on part-time
Senco SFN 1 800-543-4596 This tool is well-made; however, it seems to lack use, my suggestion is to get one of the im-
List price* $514 punch. To make it reliably pass the power test, I ported tools. They cost about half of top-
Weight 4.4 lb. had to take the line pressure beyond the tool rat- of-the-line domestic models, and they will
Height 95/8in.
ing. This problem could be related to the tool's give years of service to the part-time user.
light weight. It is the lightest nailer to handle a Choose a 15-ga. nailer if you plan to nail
Nail
length 1-2 in. 2Y2-in. nail. Beyond this glitch, I found the N60 to mostly hardwoods. A 16-ga. nailer is fine
Nail be beautiful, light and well-balanced. The tool's forsoftwoods. If you've got a cabinet shop,
capacity 104 exhaust port can be easily redirected (center stick with the 15-ga. nailer. Larger nails are
Adjustable photo, p. 25) for those times when you're work- less likely to deflect and ruin a cabinet.
depth Yes ing along a floor that's covered with drywall dust.
For the professional trim carpenter, reli-
Power test° 90 psi This tool, along with the Atro and the new
ability is the most important issue. Get a
Forget about squirting oil into this nailer. After Makita nailer, is the only 15-ga. nailer that used
nails collated at 25°. Also, these nails are a tad
nailer that has a local supplier
introducing the first oilless nailer in 1985, Senco
fragile. They tend to break apart in a tool pouch.
for both parts and nails (some manufac-
has the technology perfected. This tool is light
turers even offer job-site service). If you're
and easy to maneuver, with great balance. The The Stan-Tech is the same tool, but it is painted
going to get only one finish nailer, make it
plastic-composite magazine easily loads from the bright blue. It's available through contractor-sup-
a 15-ga. model. And no matter what the
rear, but dust tends to cling to it. This nailer now ply outlets.
application, always use the best nails you
includes the excellent depth-adjustment dial
can find.
from the newer SFN 40. The SFN 1 has good pow-
er and is quiet. The quick-open front and the
cushioned contact with "cross hairs" make this Jim Britton is a trim carpenter and contrac-
tool one of the best. tor living in Fairfield, California. Photos by
Charles Miller.

Finish Carpentry 29
 Installing Baseboard
There's a tad more to it than coping the joints

by Bob Syvanen

aseboard installation is often done badly. shaper. Another way of getting a unique base- tact with the square-cut board when the two are
Why? Probably because it comes at the end of board profile is to assemble it from combina- brought together. If the baseboard has a flat top
the job, after the crowns and casings, and car- tions of standard moldings, as shown in the edge Qike the one in the drawing), this edge
penters are anxious to wind things up so they drawing below. There's really no end to the should be square cut—an angle would show as
can get on to new projects. Or it may be be- shapes that can be achieved when two or three- a gap. If you've made the cut correctly, the end
cause it's uncomfortable work done on hands piece combination baseboards are used. of the coped baseboard will slip right over the
and knees, with a lot of getting up and kneeling square-cut end of the one you installed earlier.
down. But maybe it's just because a lot of car- Coping the joint—Many of the techniques for This technique will work on just about any base-
penters don't know how to do it right. cutting, fitting, nailing and finishing baseboard board, and can also be used to fit ceiling mold-
are similar to those required for casing. But the ing. It sometimes requires a little adjustment
Shapes and styles—Baseboards are used to miter joint used so frequently for casing tends to with a sharp chisel or utility knife.
cover any gaps that occur at the juncture of open up on the inside corners of baseboards. A
walls and floors, and they also protect the lower much better baseboard joint for inside corners Installing standard baseboard—If there's a
wall from dings arid scrapes. Visually they give is the coped joint. simple choice between a long, unbroken wall
weight, definition and presence to the wall, A coped joint requires a different cut on each and a short one, I start the installation with the
working with the crown molding and corners to of the two boards to be joined. It involves some long one. It's easier to get a good fit with a long
frame the wall." Baseboards are usually made miter cutting, so a backsaw and miter box or a piece of baseboard than with a short one, and
of the same wood that's used for trim elsewhere power miter box are required. Though I've cut you'll see why in a moment. I also try to mini-
in the house, and they can be either hardwood miters by hand for years, I like the power miter mize any possibility that people will see a poorly
or softwood. The central part of the back face box because it's fast, but doesn't sacrifice quality. fitted corner joint (if one happens to slip into
of baseboard stock is partially relieved, or You'll also need a coping saw. This small tool the job). To do this, I like to install the first
plowed away, like casings; this helps it lie with a spring-steel frame looks like a C-clamp length of baseboard on the side of a room that's
better against the wall. Baseboards, however, with a wood handle, and has a slender, fine- opposite the door. The baseboard on the adja-
are usually thinner than casing stock. This is be- tooth blade stretched across the mouth of the C. cent wall will conceal the imperfect joint so that
cause casing is frequently made with a rounded Cutting the first board in a coped joint is it won't be visible when someone first enters.
outside edge, and a somewhat thinner base- easy—just cut it square so it fits tight into the Let's assume that you've chosen to start on an
board can be butted against this edge without corner of the wall. The second board is coped. unbroken wall that can be fitted with a single
its looking awkward. To begin a coped joint, miter the board vertical- length of baseboard. Begin by measuring the
Standard baseboard comes in a variety of ly, as if you were going to make an inside mi- wall, making sure to take your measurements at
shapes and sizes, and custom shapes can be tered corner. When you're done, look closely at the floor level. Walls aren't usually in perfect
made in the shop with a table saw, router or the front edge of the cut—it will reveal the base- plumb, so the measurement will vary depending
board's profile, and will serve as a guideline for
making the second cut on the board. I rub the
edge of the cut with the edge of a pencil lead to
make it more visible.
To complete the cope, support the board,
front face up, so that the end to be cut hangs
just beyond some solid support—a workbench,
sawhorse or cricket (a cricket is a portable step
turned mini-workbench). Then, with the saw-
blade nearly perpendicular to the bottom edge
of the board, cut along the pencil line, following
whatever curve is indicated (drawing, facing
page, left). While cutting, incline the saw slightly
to put an angle on the cut. The angle should
slope away from the front surface of the board,
and will help the lead edge to make good con-

When baseboard must be fit to door casing, a

measuring block (photo left) makes it easy to
mark the baseboard to length. Built from scrap
wood to fit the particular baseboard being In-
stalled, it is placed against the outside edge of
the casing and over the baseboard, and a cut
line is then scribed on the baseboard.

30 The Best of Fine Homebuilding Photo: Pat Syvanen

on where you take it. Select a straight length of ing your way around the room until all the base- Outside corners—Outside corners on base-
baseboard and clean up one end by cutting it board is in place. There may be times when you board are always mitered. Since walls rarely
square. Use the wall measure you took to find have to fit a baseboard between baseboards on make a perfect corner, I a'ways make trial cuts
the other end and cut it square also. The cut opposite walls. If you do, just cope both ends to find the right angle for the miter. For a good
should allow a snug fit, particularly at its top and snap it into place. fit, the angle cut on both pieces should be the
edge where it will be visible. Test the fit on the For fitting baseboard to door casing, I use an same. I make my first cut long and gradually trim
wall, trim off a tad if necessary and nail the L-shaped block @hoto facing page) to help me it to perfection using the miter box or a block
piece in place. If you have to splice lengths of mark an accurate cutline. By holding the top of plane. (With a miter box, shim cardboard or
baseboard on long walls, a vertical scarf joint the block against the edge of the casing and even plane shavings between the miter-box fence
(overlapping 45° miters) is the best to use. marking down the leg onto the baseboard, I get and the back of the baseboard; make fine adjust-
Nail through the baseboard and into the a precise measurement. With one end coped, ments to the saw angle by moving the shim away
studs, using 8d finish nails top and bottom. Use slip the baseboard into place, allowing it to run from or closer to the saw blade.) Outside corners
as many nails as necessary to get the baseboard slightly long past the casing on the other end. are cross-nailed to lock the joint in place.
to pull tight to the wall. Studs can be located Place the block against the outside edge of the
before setting the base in place by probing with casing and over the baseboard, and mark a ver- More tips on nailing—Occasionally you will
a nail driven through the finished wall (the base- tical line for the square cut. A cut made a tad be faced with a situation where the baseboard
board will cover the holes). Mark the locations outside this line will give you just enough extra has to be pulled in against the wall, but there
above where the top edge of the baseboard will length to spnng the board into place. isn't any stud to nail into behind it. As an alter-
be with a light pencil mark so they won't be ob- native, drive a lGd finish nail through the base-
scured when you set the baseboard in place. Other joints—Coping doesn't work on all board, angling it down and out so that it catches
With this done, you can begin work on the baseboard profiles. Simple rectangular base- the 2x4 bottom plate in the wall. When the nail
adjacent length of baseboard. This one gets a board, the kind with a rounded upper corner, is set, the baseboard will be pulled snug against
coped cut at one end and a square cut at the just doesn't look good with coped joints be- the wall.
other (this same combination of cuts is used cause the abrupt curve makes for a very fragile In similar situations at inside corners and
when a baseboard has to be fit between a door overlap on the adjacent baseboard. Instead, I door openings, 16d finish nails again come to
casing and another baseboard). First, cope the use a combination butt and miter joint for inside the rescue. Just angle the nail until hit some-
end that will butt against the board you just corners (drawing, below right). thing solid. Keep it a few inches from the end of
nailed in place, and check it for a good fit. Then The trick here is to make a mitered lap joint at the baseboard and predrill the hole before you
measure to locate the square cut at the other the top edge of butting baseboards. Begin with a nail to prevent splitting. If you are framing up a
end. The fit should be snug, but not so tight that square cut at the end of one baseboard. Miter new house, it's a good idea to install short
it cracks the plaster or drywall when you nail it the top edge with a finish saw, cutting to the lengths of 2x4 baseboard blocking (offcuts or
in place. point where the rounded corner ends, and fol- scrap pieces) at all inside corners and at each
To get just the right amount of snugness on low this with a cut at 90° to the first. This will side of door openings.
long boards, cut them a tad long so that the release a triangular piece of wood. At the end ol If you're installing a two or three-piece base-
middle of the board will be bowed out from the the second baseboard, make a miter cut that board, the lower molding should be nailed to
wall when you put it in place. The amount of corresponds to the first, again just to the bottom the floor, not to the baseboard. If it any
bow depends on the length of the baseboard of the rounded edge. Then turn the baseboard shrinkage of the baseboard will pull the molding
you're installing, but it's usually about a finger's over. Working from the bottom edge, make a away from the floor and expose an unsightly
width, measured midway between the ends. But square cut that's angled just a tad away from the crack. When it's nailed to the floor, molding
you shouldn't have to force it—a gentle push front of the workpiece. This cut should be serves as a slip joint, concealing any shrinkage
and the baseboard should snap into place as it stopped at a line scribed off the baseboard al- cracks. Paint the baseboard before installing this
nears the wall. Take particular care when you're ready in place, because there isn't much room molding so an unpainted strip won't show up
fitting baseboard to the casing of a door. As you for adjusting this joint. This is a much better- after the baseboard has shrunk. D
spring the baseboard into place you don't want looking joint than a butt joint or a coped joint,
to push the casing out of position. If the fit looks though a butt joint can be okay when molding is Bob Syvanen, of Brewster, Mass., is a consult-
good at both ends, nail it in place. Keep work- used on top of it. ing editor of Fine HomebuLlding magazine.

Mitered lap joint

-8d nails as needed

For the best corner on simple round-

edged baseboard, a ed lap joint ss used. -
A coped joint is a two step process that begins with a 45 miter cut With a coping The board on the left is cut first, and nailed
the inside edge of the Cut as shown Angle the saw slightly and back bevel the cut Thi ill ensure in place. The second board receives
that the visible front edge of the coped baseboard wi/I fit tightly against the adjoining baseboard. corresponding cuts.

Finish Carpentry 31
 A jig for joining rounded baseboard corners. Rounded corner
pieces help baseboard fit plaster walls with bullnosed corners. Corner pieces are
biscuited to straight runs of baseboard, and a simple jig ensures a perfect fit.
The jig, made from 3/4-in, plywood scrap, holds a rounded corner piece in
position so that the biscuit joiner can plunge the cuts for two #10 biscuits. The
fence on the biscuit joiner should be set at 90°. The thickness of the scrap used
to make the jig should equal the radius of the corner piece.

Curved Baseboard Corners

Biscuit joinery arid a simple jig solve a perplexing trim problem

by Eric Blomberg

hin-coat plaster is increasingly popular for It took trial and error to get the technique right. the corner pieces. After cutting the pieces to
interior walls in northern California where I work, We realized that rounded corner stock was part match the height of the baseboard (drawing
but the variable textures and the bullnose cor- of the answer; the trick was learning how to join above), we put them in our jig and mark the lo-
ners of these walls can raise havoc when it's time the baseboard and the corner pieces together cations for two #10 biscuits (biscuits come in
to apply trim. I discovered this a couple of years cleanly. We resolved the problem with biscuit three sizes, ranging from #0, the smallest, to #20,
ago when the crew I was working on started to in- joinery, using a jig we devised. The technique is the largest). Using #10 biscuits with 1x6 base-
stall baseboard in a house near Santa Rosa, Cali- fast and effective. board means we can get two biscuits at each
fornia. Each outside corner on these coarse-tex- joint, and the slots aren't deep enough to break
tured walls had a 3/4-in, radius. Mitering the Corner pieces from a mill shop—The corner through the face of the corner pieces. (We cut
baseboard around these corners was out of the pieces were created with a single pass through a the biscuit slots in the baseboard later, after all
question. Square corners would look bad juxta- multiple-head molder, producing 10-ft. and 12-ft. of the pieces have been cut to length.)
posed with the buLlnose walls, and mitered cor- long pieces that we cut to length on the job. Be- The trick to cutting the slots accurately is the
ners would leave an awkward gap between the cause the shop grinds knives for each job, it jig; it holds the corner piece and aligns the cutter
wall and the baseboard. Pondering paradigms could easily make radiused corner pieces. You in the biscuit joiner. The jig is simply two pieces
led us to our current solution. can also make the pieces yourself (see sidebar of 3/4-in, scrap plywood screwed to a base. The
Our technique requires only a biscuit joiner, a facing page). corner piece fits between the two scrap pieces.
simple jig and roinded corner stock that we For baseboard that will be painted, grain direc- The jig aligns the base plate of the biscuit joiner
have run off at a local mill shop. The back of the tion in the corner pieces is irrelevant. That was with the inside edge of the corner piece. If the
rounded corner pieces are concave, so they fit the case on this job. But the mill shop we use al- corner piece is aligned correctly in the jig, the
snugly against the radius of the plaster walls at so can produce corner pieces with the grain run- joint between the baseboard and the corner
the corner. That eliminates the triangular gap that ning horizontally, just like the base. That would piece will be flush. It's important to cut the bis-
would result ii the baseboard were mitered at be helpful if the trim were going to get stained, cuit slots perpendicularly to the end of each cor-
45°. In this house we butted the 1x6 maple and the grain direction had to match. ner piece; if the slots are skewed, gaps at the
baseboard into the rounded corner pieces, so joints are inevitable. After marking the corner
the baseboard follows the wall cleanly, even Corner pieces first—We have a three-step pro- piece for slot locations, the cuts can be made on
if the corner isn't exactly 90°. cedure for fitting baseboard, and we start with both edges. With that done, we thy-fit the corner

32 The Best of Fine Honiebuilding Drawings: Bob La Pointe

 Tack to wall if
necessary.

2#lObiscuits
To install baseboard, dry-fit corners first. To
of baseboard, corner pieces are dry-fit to scraps of baseboard and tacked in
place temporarily. Baseboard between two outside corners is cut and squared at
each end. For coped inside corners, the first piece of baseboard is cut to length
and held in place before the second piece of baseboard is measured, cut to
length and coped at one end. In this installation, the author blocked up
baseboard to allow room for the finish floor later.

pieces to scrap pieces of baseboard and then

use a router and a roundover bit on the top edges
of the corner pieces so that they will match the Making corner pieces on a shaper
top-edge profile of the baseboard. Once this If there's not a mill shop in your area that be joined with the straight runs
profile has been cut, the corner pieces are ready will make the corner pieces, you might of baseboard.
for installation. try to make them yourself. The required Next, I would cut the inside curve on a
tools are a table saw, a shaper and, of shaper. The radius depends on variables
Measuring and installing the baseboard— course, the necessary shaper knives. I like drywall or plaster thickness and the
The second step in the process is to set corner wouldn't recommend a router because thickness of the base itself. Then I would
pieces in place temporarily and measure the the bits required for the cuts would be go back to the table saw to rip the
straight runs of baseboard. In the Santa Rosa very large. opposite (outside) corner to remove most
house, a tile floor was to be installed after the A mill shop in my area makes all the of the waste material.
baseboard was installed, so we raised the base- corner stock I need. But if I were going Finally, I'd use the shaper to finish up.
board 11/2 in. off the subfloor and took our mea- to make my own corner stock, I'd use Again, the radius required may vary. I
surements for the base at this height (drawing a three-step process on the table saw think two passes would do it, each pass
above). To hold corner pieces in place while we and shaper. First I'd rip the stock to cutting half the outside radius. A little
measured straight runs of baseboard, we thy-fit the appropriate dimensions (drawing sanding will finish the job up nicely. If
scraps of base to corner pieces with biscuits. The below). These cuts would establish the you try this, use material that's long
assemblies could be tacked to the wall to hold two faces of the corner piece that will enough to be machined safely. —E. B.
corner pieces in the correct position while we
took measurements.
With the straight runs of baseboard all cut, we Using a table saw and a shaper. After the stock has been dimensioned
could install all of the pieces. At each outside and squared, use a shaper to remove material on the inside corner (1). Next, use
corner, we dry-fit both pieces of base to the cor- the table saw to remove waste on the outside corner (2). The last cuts on the
ner piece to check the joints. If the fit looked shaper (3) produce the outside radius.
good, we glued the slots, inserted the biscuits
and then stuck the pieces together. If the dry fit
had been perfect, we would have nailed the
pieces in place right after the glue was applied.
If extra pressure were needed because one of
the joints was slightly off, we let the joint set up
off to one side before installing the pieces. When
the glue was set, we nailed the assembly in place.
It was sometimes possible, especially with short-
er pieces, to assemble a three-piece corner or a
five-piece U-shaped section, let the glue dry and
then install it in one piece. The process may
sound tedious, but it's not. Once a rhythm is es-
tablished, the work flows smoothly.

Eric Blomberg is a carpenter with Jim Murphy &

Associates of Santa Rosa, Calif

Finish Carpentry 33
 Running Baseboard Efficiently
Simple steps help you make the most of time and materials

by Greg Smith
L et's face it. There is little or no glory in the
installation of baseboard, If you want, for in-
stance, to talk about hanging doors, you can
probably find plenty of guys who are happy to
grant you their expert opinions on the best tools
and the most elaborate techniques. But when it
comes to installing baseboard, we're back to
grabbing a scrap of lumber or an unspent napkin
from lunch to record measurements.
The job may go something like this: enter
room, plop saw on floor, measure, cut, nail; mea-
sure, cut, nail; measure, cut, nail. And you won-
der if you'll ever get to the last piece, because it
seems like there is always another little piece in
some nook or cranny or some space that was
missed. It is a job that brings screaming protest
from the knees and a hacking voice of discon-
tent from the lungs of the person who fires the
nailer that kicks up the dust from the floor adja-
cent to the workpiece. That may be why, when a
team of carpenters is finishing a house, running
baseboard is often relegated to the least-experi-
enced person of the group or the low man on
the totem pole.
The best way to deal with an unpleasant,
though necessary, task is to get it done as quickly
as possible. I have seen many carpenters ap-
proach the installation of baseboard in many dif-
ferent ways, but I had never seen a system that
works very efficiently. That's why I developed a
methodical approach that makes baseboard in-
stallation fast and efficient.

1. Plan your strategy—The time to run base-

board is before the painter has hidden the loca-
tion of studs (assuming that you are dealing with
drywall) and after the door casing, the built-ins
and cabinets and the hardwood or tile floors are
installed. In the areas that will be carpeted, hold
the baseboard off the floor with a piece of hard-
wood flooring or other scrap of 3/4-in, material—
you won't want your beautiful work hidden by
the carpet. Though like to leave a wake of com-
pleted baseboard behind me when I am work-
ing, that's not always possible. If the bathroom
floors have yet to be tiled, for example, I cut my
baseboard for the room and set it aside.

2. Set up the saw—I usually set up my 10-in.

power miter saw across extra-tall sawhorses (so
that I don't have to bend far to see a close-up of
the cut) in the biggest room on the floor level on
which I'll be working. I'm not as concerned At the wall. To run baseboard efficiently, Greg Smith measures several rooms at a time. He pre-
about how close I am to the area to be worked pares a cut list as he goes and marks lengths and cut angles on a preprinted form.

34 The Best of Fine Homebuilding

 blanks is sequentially numbered, and the mea-
surements are recorded in order (drawings
next page). The sequernial numbers are impor-
tant. They will later be recorded on the boards
and used to guide you toward correct board
placement. You can increase your ability to keep
track of what you're doing by drawing a line be-
tween each set of measurements when you
change rooms. Write the name of the room in
the vertical space to the ieft of the column. In the
boards #1 to #4 go in the master bed-
room, boards #5 and #6 go in the closet.
The solid line to the right of the bianks repre-
sents the baseboard as you are looking at it on
the wall. On the left and right ends of this line,
you will write a symbol to represent the kind of
cut required on each end. Use whatever symbols
make sense to you—but be consistent. In the
small example on the facing page, the straight,
vertical line on baseboard #1 means the cut is
to be a straight, square cut This end of the board
will butt against the casing around a door. The
"2" on the right-hand side of the notation repre-
sents an inside 22Y2° cut. Baseboard #2 starts
with a 22½° cut; the slash at the end of the base-
board line indicatesa45° inside cut. Baseboard
#4 shows a square cut and a 45° outside cut
(represented by the 0 at the end of the line).
When working with a stain-grade wood, such
as oak, most carpenters like to cope one end
to get a tighter fit. I use a C to indicate the end to
be coped. You won't find many different cuts
in baseboarding, so it won't be hard to memo-
rize the symbols you'll need. If you work with
others, you might want to define your symbols
on the form so that everyone will be singing from
the same song sheet. As for measuring, I simply
Al tue situ'. With a pile of stock nearby and his list of baseboard dimensions at the ready, hook the tape where it's most convenient.
parks himself at the saw and cuts down the list. Each piece gets a number that corresponds to
list; later on, he'll be able to mate each length with the correct wall.
5. Cut each closing board—After you have
compiled measurements from several rooms,
on as I am about having enough room to extend board to be spliced somewhere along its leng h take your clipboard to the saw and start cutting
long lengths of material on either side of the saw. (drawing next page, bottom). Start by cutti g baseboard (photo lefi). Each time you cut a
With my system, I don't spend a lot of time walk- lengths of stock with a 45° inside miter on,t e board, mark the backside near an end or in the
ing back and forth between my saw and the left-hand side and an inside miter e middle (be consistent so that you will know
work area. If there are no large rooms, or if for right-hand side (assuming you're working fr rn where to look for it later), using the sequential
some reason I cannot use one, I set up instead left to right). The splice will occur at the second number on your list. Then cross that length off
where I can extend stock out through a door or a cut. Later on, you'll be able to take one m4a- your list. There is no need to write exact lengths
window. When all of the baseboard material is surement from the 22Y2° end to the corner of 4ie on the board, as many carpenters do. For ex-
spread out near the saw, I'm ready to start work. wall to complete the wall. Go ahead and ample, if you cut a inside cut on the left-
By the way, no matter how clean the subfloor these lengths of baseboard. hand side and in. to the right, make a 45°
is, when you're shooting in baseboard with a inside cut; turn the board over, write the num-
nailer, the dust is going to be flying, and your 4. Take closing measurements—Now ta ber 3 and cross it off yoJr list. Then the board
mouth and nose, being close to the ground, are all remaining measurements for three or Ic ur aside and begin work on board #4.
going to scoop up a lot of it. You might want to rooms at a time. Starting at the door, measi. re You need not cut in any particular order. One
use a dust mask for this part of the job. In situa- each and every length in the room. Work yr ur additional advantage of this method is that you
tions where I can free up my left hand, I put it way around the room in a clockwise or count can make very efficient use of your material by
under the place where air is released from the clockwise direction, whichever you prefer. T taking a little extra time here to avoid waste. Start
nailer. This keeps the dust from being kicked into direction you choose is not as important as with your longest pieces, then see what you can
the air. being consistent. Use a mechanical pencil (t get out of the offcuts. You'll have a long list from
As for what joints to cut, it's up to you whether kind you can get in any grocery or drug sto e) which to choose.
you miter or cope. The system I use to organize both for recording your measurements an or
the process won't change the way you work. For marking cutlines later on. The point will al ys 6. Distribute the boards—Now that you have
purposes of explanation, however, I'll use the ex- be sharp and consistent, and you won't w ii tIe cut and numbered all of the pieces on your list,
ample of mitered baseboard. away time carving up a carpenter's pencil wit] ia you can distribute the pieces. Because you have
razor knife. numbered them in the order in which you mea-
3. Install the long boards—Tackle the longest Each measurement is recorded on a v ry sured the walls, it is easy to pick up any random
walls first—the ones that are longer than the stock simple form that I've designed and carry oi a piece and qwckly find the place where it be-
you are using. These walls will require the base- clipboard (photo facing page). The column of longs. Let's say that the first board you pick up is

Finish Carpentry 35
 Baseboard euthst. The sample at left shows how to use the cutlist (below). Marks at the end of
i.b.51z [ Z each solid line indicate the cuts to be made on each baseboard.

2/ Baseboard Cutlist

/ 1
1

2
1

2 2
1

1
0 3 3 3

4 4 4

5 5 — 5

6 6 6

marked 18, and you see that it is the third board 7 7 — 7

cut in the dining room. You can go to the dining
room immediately and set the board alongside 8 8 8
the third wall from where you started measuring
in that room. 9 9 9

7. Nail 'em up—When all the pieces are lying in 10 10 — 10

front of their respective locations, it is time to
crawl around the room and nail them up. You 11 11 11
may occasionally find that a board or two will
need to be recut or trimmed because of walls 12 12 12
being out of square or because you measured
wrong. This would happen regardless of your ap- 13 13 13
proach to the job and will not affect your other
cuts. You can either recut as you come to them 14 14 14
or write the adjustment needed on the back of
the board; e. g., "-'/8" indicates that you need to 15 15 15
off of the length.
cut 1/8 in.
This system also works well if you like working 16 16 16
with a partner. One person measures and installs,
the other does all of the cutting. Work this way 17 17 17
with a partner only if you like and respect each
other and if each of you has a sense of humor. 18 18 18
Inevitably there will be discussions about who
can't measure right and who can't cut right when 19 19 19
boards occasionally come up short or long.
20 20. 20.
Nothing fancy, but it works—It's a simple sys- —
tern, really, but it does work. You don't have to
use a preprinted form, of course, though it does
eliminate the need to write out the sequential
numbers and draw lines. Besides, it's a heck of a
lot easier to write on and read than some of the
things that you see carpenters writing on at con-
struction sites.
To save you a bit of setup time, you can photo-
copy the sample form at right for your own use.
Photocopy enough to keep yourself supplied for
a few months and keep a few extras in your
truck. The columns will allow you to work
through the "measure, cut, nail" process 60 times
per page. You can also use the form to record
measurements for other types of molding and for
closet poles. Baseboarding may still be the lowli-
est job going, but with a bit of organization, you
won't be down there quite as long.

Greg Smith is a general contractor in West Los

Angeles, Calif., who specializes in custom home
building and remodeling. Photos by Marilyn Ray

36 The Best of Fine Homebuilding Drawings: Bob Goodfellow

 Molding Character
Using a molder/planer on site
to create a formal 18th-century interior

by Douglas Honychurch

s a Connecticut native and real-estate ap- brother asked me to help with a new house that toured plinth block at the floor. The formal man-
praiser, I've gradually come to appreciate the he was building. didn't need much coaxing.
I
tel surrounding the fireplace would be flanked
examples of early Colonial architecture that re- The house is a two-story Dutch Colonial by fluted pilasters (photo below). There would
main in our area, and particularly value the
I flared eaves, cedar clapboards and a corbeled also be an elaborate crown mold all the way
craftsmanship that made them worth preserving. end chimney (top right photo, next page). My around the room. For all this work, we wanted
My interest began with old houses that I noticed brother gave me free rein in the design and con- to forego stock moldings and make our own trim
while driving through the New England country- struction of the architectural details. wherever possible.
side, and eventually led me to look for historic The designs came primarily from three books:
houses that were open to the public. I began Design—With a slate-faced fireplace on one Southern Interiors of Charleston, South Caroli-
buying books about Colonial houses and en- wall and a large, multi-pane window on na by Samuel and Narcissa Chamberlain (Has-
joyed studying their wonderful architectural de- a designer's para- tings House Publishers, 1956), Architectural
tails. Some of the books had cross-sectional dise. We decided to build twin bookcases on Treasures of Early America, vols. 3 and 7 (re-
drawings of walls, showing wainscoting and each side of the large window. A raised-panel printed in 1977 by Arno Press Inc.) and Early
crown molds. These especially interested me. wainscot would enclose the room, with a two- Domestic Architecture of Connecticut by J. Fred-
The desire to do some of this work myself led piece chair rail above and a two-piece base- erick Kelly (Dover Publications, 1963). The first
me to purchase a Williams & Hussey molder! board below. The door and window casing two books are out of print (I found them in
planer, and use it to run moldings for a remodel- would have a beaded edge toward the opening used-book stores), but the third is still available.
ing project in my own house. But my real chance and a small band mold around the outside. On My brother and I wanted to create a room
to design and mill Colonial trim came when my the doors, this casing would die into a that felt like the late 18th century. Decisions
about what moldings to use where were based
on my taste. I liked the wainscoting and fire-
place designs in Architectural Treasures of Early
America. But they were too elaborate. We sim-
plified the moldings and eliminated the carvings.
These decisions were also influenced by the
limitations of my molder/planer, which can cut
only ¾ in. deep. And of course, trial and error
played a part in the design—tack on a molding,
stand back and look—a process that led to
many changes.

Construction—Unfortunately, the carpenter

who had done such a fine job framing the house
was behind schedule for his next job and Surroundea ny nutea puasters and a pair of raised panels, the fireplace mantel (photos above left
and previous page) commands attention in a room full of architectural details. Both site-made and
couldn't be persuaded to do the complex trim stock moldings were used for the trim work. To cut costs and avoid problems with wood movement
work on the first floor. Our search for an equally due to moisture, the raised panels were made in one piece, using medium-density particleboard.
skilled carpenter took some time, but on the The Interior detailing, inspired by that of the 18th century, was a natural choice to complement the
recommendation of our painter, we met Ed Dutch-Colonial styling of the house (above right).
Rockwell and looked at some of his recent
work. We soon learned how fortunate we were
to find him. His experience with this type of fin-
ish carpentry was extensive, since his family had
been in the business for several generations. Un-
like some tradesmen who are steeped in tradi-
tional lore, Rockwell was willing to try new de-
signs and methods.
With our carpenter on the job, we were ready
to make the moldings. The molder/planer we
used was a Williams & Hussey model W7S
(photo right), with power infeed and outfeed
rollers (Williams & Hussey Machine Co., Elm St.,
Dept. 1361P, Milford, N. H. 03055). When I got
the machine in 1984, 1 also bought a 2-hp motor
to go with it. The whole outfit cost about $800.
We set up the molder/planer in the living
room alongside Ed Rockwell's table saw, which
was mounted on a rolling stand with a plywood
extension added to the back of the table, both
of which I found quite helpful. We used kiln-
dried, select eastern white pine from New
Hampshire. Ultimately we made seven different
moldings (drawing, bottom right) and used
three of them in the living room. The first step
was to rip the boards to width on the table saw.
We used a high-quality carbide-tipped blade, Using a Williams & Hussey model W7S molder/planer on the site, Honychurch made many of the
moldings used to trim the first-floor rooms. Molding profiles are shown in the drawing below.
which made cleaning up the edges unnecessary.
Williams & Hussey offers a range of molding
knives in standard patterns, and the company
will also custom-grind special patterns. All you
have to do is send them a drawing or a sample
of the molding you want to cut. The cost of the
knives is based on the overall width of the high-
speed steel blanks (from 1 in. to 7 in.), not on
the profile of the molding. I had custom sets of
knives made, and they ranged in price from $72
to $120 per set.
To install these knives (two per set) on the
molder/planer required only tightening some
bolts. The alignment and registration of the
knives were perfect without any adjustment. We
used two plywood fences on the molder/planer,
and secured them with small C-clamps that come
with the machine The knives had to be lowered
into position and the fences adjusted and posi-
tioned so that the dimensioned boards would
run through without any slop. We put the board
between the two fences and lowered the knives
to see where they would hit. If the alignment
was right, we next adjusted the tension on the

38 The Best of Fine Homebuilding

 feed rollers, which pressed the wood firmly onto
the table while cutting, in addition to forcing it
through the cutterhead. We learned how impor-
tant feed-roller tension was when we broke a
blade because the rollers weren't tensioned
enough, and the stock began to chatter. Adjust-
1
ing them was a real challenge—too little tension
and the stock would flap violently and get
chewed up; too much and it wouldn't feed.
Overall, the machine performed well for our
needs and allowed us to make custom molding
at less than custom prices. The blades stayed
sharp; we used one set to make door and win-
dow casing for the entire house. These did need
sharpening at one point, for which Williams &
Hussey charged $15. To keep the knives sharp,
it was important to prevent resin, pitch and sap
from building up on them. We cid this by care-
fully scrapng them with an old chisel.

The mockup of the Greek-key fretwork on the

I
The crown mold in tile living room (above) Is
just one 5-in, wide piece, though it looks built
The fireplace—In the center of the 30-ft. long
exterior wall is the fireplace. We wanted it to be
the focal point of the living room. So one of our
mantel (above) was rejected because it would
have been too complicated to make and apply. up from three pieces. It's a discontinued stock first decisions was to extend the 8-ft. section of
Saw kerfs were made in the back of the pine design called Curtis mold, and was ordered wall that encompassed the fireplace into the
board to relieve cupping and were later covered from a local mill. Around the fireplace (below),
beaded cove mold was added under the crown. room about 4 in., thereby up the long
by a thin band.
expanse. This 8-ft. section is f,anked by fluted
pilasters, which we made on the using a
router and a fluting bit. The area between the
pilasters is finished with an elaborate mantel
and two large raised panels above it.
Our mantel design was suggested by one that
we found in Architectural Treasures of Early
America, vol. 7, and was a popular style in the
eastern United States during the late 1700s. In
deference to cost, we simplified the design by
omitting the carved elements and using stock
moldings to supplement the ones we made our-
selves. We built the mantel directly onto the
wall, furring out the main section with 2x4s
nailed flat and ccvered with ¾-in, plywood to
provide plenty of nailing surface.
At one point in the course of my trial-and-error
design for the mantel I considered using a
Greek-key fretwork under the mantel shelf. I ex-
perimented on my table saw, cutting alternating
grooves in a 1-in, piece of pine (photo top left),
but realized that running it around the various
offsets on the mantel would be tough. At this
point, a trim carpenter who was working up-
stairs walked by and saw what I was ttying to
do. He didn't like the molding I'd made or how
difficult it would be to apply, even though it
wouldn't even be his job. He told rue "You real-
ly know how to try a man's patience." He then
built a jig for hts router and made an alternative
dentil molding that was good-looking and
simpler to apply. It's the one we used.
Rockwell suggested hat we embellish the
crown mold in this 8-ft. section, and so we
added a beaded cove at the bottom (photo bot-
tom left). Also, at the extreme outside corners
of this section, rather than a butt or mitered
joint, we used a nearly circular beaded joint. Al-
together, it took Rockwell and me a full week to
complete the mantel.

Wainscoting—We decided to run a raised-

panel wainscot around the entire living room,
dining rcom and hall. The 34-in, chair-rail height

Finish Carpentry 39
Photos at top of page: Dougiaa I: I
After some debate over wnether me cnair rail
should protrude beyond the casing, the detail
above was chosen as the appropriate way to
end the molding at doors and windows.

was worked out to accommodate panels of

pleasing proportion, separated by 3½-in, wide
rails (horizontal border) and stiles (vertical bor-
der). In laying out each wall and determining
panel sizes, we tried to keep the room as sym-
metrical as possible.
We chose not to use the traditional method of
constructing raised panels—with stiles and rails
mortised together, the raised panel let into a
groove within them and the finished frames and
panels applied to the wall. We wanted a method
that would be less costly. Before sheetrocking,
we had nailed plenty of blocking between all
the 2x4 studs at the anticipated heights.
We started with the rails, nailing 1x6 pine
around the room at chair-rail height. Then we
ran the bottom rail so that once the 1x6 base-
board and base cap were installed, 3½ in. would
be exposed above it. After the rails were up, we
cut the stiles to fit between them. We used
butt joints with a little glue and just nailed the
pieces to the wall. No half-lapping was done.
The pine wasn't a uniform thickness so we had
to shim some of the joints to make them flush.
We then measured all the openings and cal-
culated how much material we'd need for the
raised panels. To save time and money, we de-
cided to experiment with a medium-density par-
ticleboard called Medite (Medford Corp., Me-
dite Division, P.O. Box 550, Medford, Ore.
97501) to make the raised panels. It is extreme- Traditionally styled bookcases [lank the large, multi-pane window. Before they were attached, all
ly dense and smooth and takes a beveled edge
well, in addition to being very stable. Medite is raised-panel cutter, and arranged with a local the paneling, we were able to remove only that
available in 4x8 (actually 49-in, by 97-in.) and woodworking shop to use their large shaper one panel with ease.
5x8 (61-in, by 97-in.) sheets and a range of (1¼-in, spindle) to bevel the panels' edges. The chair rail, instead of being flat on top, has
thicknesses from ¼ in to in.1
We were extremely pleased with the way the a slight wave in it, which then steps down be-
By using sheet goods for the panels, we were raised panels turned out. We took the numbered fore protruding out in a rounded edge. Its
able to cut each one from a solid piece, as op- panels back to the house and found that most of downscaled profile resembles a traditional 18th-
posed to traditional panels, which are usually them fit without any need for adjustment because century chair rail. This was one of the few mold-
made from glued-up pine boards. Also, by using we had intentionally undersized them in. all ings we ordered from a local mill. We ripped
Medite, which is moisture resistant, we were around. Then, using brads inserted with a brad the chair rail to a width of about 3 in., then
able to reduce the chance that the panels would driver, we ran %-in. quarter-round around the scribed it to the wall with a pencil compass. We
warp or check. openings to hold the panels in place and cover sat it directly on top of the IxG and nailed it into
Using a table saw equipped with a carbide- the gaps. We later discovered another advan- the studs. Underneath it, we ran the same band
tipped blade, we cut the Medite into panels to fit tage of our paneling method. When the electri- molding as on the door and window casings.
the 63 openings. We bought a carbide-tipped cian needed access at a particular spot behind Rockwell and I disagreed about the treatment

40 The Best of Fine Homebuilding

 The same wainscoting design used in tile living
room also covers the walls of the foyer and
runs up the stairs. But the crown molding in
this area is a simple beaded design and was
made on the site.

¾-in, deep cut, which the manufacturer says is

the maximum. We made the molding in one
pass, and it did test the machine's limits. We
could hear the motor strain; in fact, the power
demanded to cut this molding was so great that
the lights dimmed when we ran it. A much
simpler molding than the Curtis mold, it was
easier to install, especially when coping the in-
side corners.

Bookcases—The bookshelves and cabinets

that flank the multi-pane window (photo left)
were designed by Rockwell, based on ones he'd
done in the past. He built them in place, with
the facings were beaded along their inside edges—a simple detail that adds depth and richness. 10-in, deep adjustable shelves above and 16-in.
deep cabinets below. The upper portion of each
of the chair rail where it met the door and win- Crown molding—The crown molding we bookcase has two bays made from lxlO pine.
dow casing. As I had designed it, the chair rail used in the living room is actually one piece but Rockwell beaded the inside edge of :he facings
protruded about I in. beyond the casing. Rock- looks like it was built up with at least three with his router. The exposed parts of the base
well strongly objected to this, believing that the pieces (top right photo, p. 39). It is a discon- were made with birch plywood, including the
profile of the chair rail should be shallow tinued stock design called Curtis molding and is counter, which was then nosed with window
enough to die into the casing without sticking made from a single piece of 5/4 by 5-in, pine. stool ripped to 1¼ in. Under the nosing we ran
out. I insisted on my design and he relented, This molding has been made on the Williams & the same band mold as below the chair rail.
notching the chair rail so that it lapped onto the Hussey molder/planer, but we felt that it would Rockwell made the raised-panel cabinet doors
casing as much as it protruded beyond it. When be pushing the limits of the machine, and we on site, using his router and table saw. The
he applied the band mold to the door casing, he therefore had it made for us at a mill. edges of the door openings are also beaded,
decided to cope it around the chair rail. He did For the crown molding in the center hall which adds a satisfying depth to the design. E
this 26 times throughout the house, and even he (çthoto above right) and dining room, we used a
seemed pleased with the result (photo facing 4-in, beaded cove that I designed. This we did Douglas Honychurch is a real-estate appraiser
page, left). make on our machine. The profile required a in Trumbull, Coon.

Finish Carpentry 41
 hen my wife and I added on to our mod- reading, has been limited to patching. The key to using the molding template suc-
est-size Cape, we installed three exterior doors, Nevertheless, armed with reference books, limit- cessfully is the leg that rides on the jamb (bot-
each of which had an elliptical transom. We both ed experience and lots of motivation from wife tom photo, p. 45). This leg ensures proper posi-
love the look of these transoms, but trimming out and wallet, I decided to trim my transoms in plas- tioning of the plaster molding and determines
these exotic shapes can be difficult, even to an ter (photo above). the size of the reveal. I started with a piece of
experienced contractor such as myself. steel a couple of inches larger in each direction
We needed more than 50 ft. of door trim, in- Making the molding template—The basic than the molding I was matching. I positioned a
cluding the straight sections and ellipses. My first process of making plaster molding involves scrap of this molding on the steel and scribed
choice was to have the trim custom-fabricated, building up layers of plaster where trim is going the shape so that a I-in, by %-in. leg was left with
but the prices I was quoted were astronomical. and dragging a molding template across each a reveal. I cut the rough shape in the steel
Next, I considered making the trim myself. I've layer until the shape of the trim is created. My with a jigsaw, staying back a little from my
had some experience laminating wood for odd- first step was to fashion a molding template out scribed line. I used fine files and the grinding
shaped applications. But between the cost, the of steel stiff enough not to need the wood back- stone on a Dremel tool to shape the mold. When
set up and the lengthy time involved, I decided ing that normally would give the metal the nec- I was satisfied, I added a wood handle to make
against this alternative. essary rigidity (top photo, facing page). 1 used the tool easier to use and to protect the jamb
Looking for another solution, I realized that 14-ga. steel I had on hand, but I recommend thin- from damage from the metal leg.
much of the interior-trim work I've seen over the ner steel, perhaps an old taping knife, because
years was done in plaster. But most of my plas- heavier-gauge steel is more likely to transmit file Prepping the walls—I cut back the drywall at
tering experience, beyond asking questions and marks to the plaster. the edge of the door jamb to provide a keyway

42 The Best of Fine Homebuilding t'hot,, this page Roe A Osborn

 The molding template. The template was shaped to match the molding in
the rest of the house. The author added a wooden handle covering the leg to
make the tool easier to use and to protect the jamb from being scratched. The
trim (left photo) is fashioned of plaster using the template (above).

for the plaster arid to give the molding more

body (bottom photo, right). I then primed the
diywall and the jamb with two coats of primer to
protect the paper face of the drywall from tearout
due to the plaster's wetness. Next, I ran my mold-
ing template along the jamb and penciled in the
outer edge of the new trim. The area inside my
line was covered with a bonding agent that al-
lows plaster to adhere chemically to a subsur-
face. The bonding agent also kept the plaster
from drying too rapidly.
In retrospect I should have applied the bonding
agent 3 in. or 4 in. beyond the outside boundary
of the trim because there was a tendency for a
small amount of excess plaster to build up be-
yond the trim and collect in irregularities in the
wall. Unless this excess plaster is cleaned as you
work, which you will have little time for, it even- Akeyway locks the molding in place. The wallboard has been cut bacK from the uoorjamu
tually flakes off, creating future headaches. ating a void that fills with plaster and keeps the molding securely anchored.

Top photo this page: Robert Marsala

Finish Carpentry 43
The first layer oi plaster is put on. After me primer and bonding agent have been applied to the molding area, the author uses a taping knife to ap-
ply a stiff mixture of plaster. The molding template then is dragged over this first layer of plaster, and the basic shape of the trim Is created.

Although I skipped this next step, I strongly rec- evenly and helped to eliminate lumps in the mix. the size of the job and the ambient temperature.
ommend the addition of expanded galvanized I filled a 5-gal. pail about three-fourths full of Clean water is a must. Water that has been cont-
metal lath mechanically fastened to the jamb clean water (14 qt. per 25 lb. of lime) and added aminated with lime or plaster scrapings from
and to the wall to strengthen the trim. the lime a little at a time. I stirred the mix occa- tools accelerates a batch of plaster beyond a us-
The final step in preparation was protecting my sionally as I added the lime, and I let it sit able setting time.
work area with plastic, 4 mu or better (especially overnight. The next day, I stirred it again with a Next, I made a ring of lime putty on a plywood
if you are working over hardwood floors—lime is stick (a heavy-duty drill and paddle also would scrap about 30-in, square and filled the ring with
caustic and will discolor hardwood). work well). The consistency of this lime putty my water/retarder mix (the ring of lime putty cre-
should be almost the same as it is for drywall ates a kind of mixing bowl). Then I added the
Mix the lime putty ahead of time—In addi- joint compound. A mix that is too wet can cause molding plaster, sifting in a little at a time. I was
tion to my molding template, 1 needed other small holes to develop in the finished molding, aiming for a 1:1 proportion of plaster to lime. I
equipment, including a hawk (a small metal while a mixture that contains too little water can let the plaster sit until all the water was absorbed
board on a handle used to hold the plaster while produce small, hard lumps that shrink and leave and until there were no dry spots. Next, I mixed
working it); a trowel; and a 6-in, taping knife, all little voids in the finished molding. the lime putty and plaster together until they
available from the local mason-supply yard. I al- were blended thoroughly. This initial mix of put-
so picked up a bag of hydrated finish lime, a bag Mixing the plaster—When I was ready to begin ty and plaster should be somewhat stiff so that it
of molding plaster (or gauging plaster) and a my trim work, I mixed the retarder in a bucket of stays put on the wall.
small quantity of commercial retarder. clean, potable water. The recommended ratio is
I made the lime putty the day before I needed 24 oz. of retarder to 4 gal. of water. The exact Applying the plaster—Using a taping knife, I
it. This allowed the lime to soak up the water amount of retarder varies based on your ability, applied the plaster to the area being trimmed

44 The Best of Fine Homebuilding

(photo facing page). This step is called blocking
out. Next, with the handle of the molding tem-
plate riding on the inside of the jamb, I ran the
template over the area I just blocked, working
up from the top of the plinth block. It is important
to keep the molding template perpendicular to
the wall at all times to make sure that the shape
of the molding is consistent.
The blocking operation created the basic
shape of the trim. Once the initial blocking was
done, I immediately went back and filled in any
voids with the same material and passed the tem-
plate over the area again. I repeated this process
until the rough shape of the mold had been trans-
ferred to the plaster or until the material had be-
come too stiff to use. Never add water to soften
your mixture! Always discard any used material,
and take the time to clean the excess plaster
from your tools.
The next step required a looser mix (more wa-
ter/retarder) and a smaller batch of plaster. I
smeared a handful of the mix over the trim (top
photo, right). Then I ran the molding template
over that area, forcing the plaster into the voids
that were left by the earlier blocking-out process.
I repeated this procedure until all of the imper-
fections had disappeared.
Instead of smearing this layer of plaster, another
option would have been to hold a handful of the A second layer Is smeared on top. A spread over the first layer of wet plaster. As the
loose plaster mix against the molding template as template Is drawn over the molding, the boo Is forced into the voids left from the initial shap-
it passed over the trim. This operation is known lag to complete the profile of the trim.
as stuffing the mold. An extra person would be
helpful with this activity because it normally
takes two hands to hold the molding template in
the correct position. In any case, you shouldn't
continue the shaping process if the initial layer of
plaster has swollen, causing the molding tem-
plate to bind and chatter. Remember that plaster
expands as it sets.
When I reached the top center of the transom, I
repeated the entire process beginning at the bot-
tom of the other side and blending the two sides
at the top (bottom photo, right). I always worked
against gravity. Working down would have left
the trim full of cracks and voids.
The final step was splashing water on the trim
with a mason's brush and running my mold tem-
plate over the trim one last time to remove any
excess material or drips. Some plasterers use a
handful of lime putty at this point and stuff the
mold to fill in any minor blemishes and to give
the trim a shine. If it looks good after the initial
stuffing, however, I suggest leaving it alone. After
the plaster dried, I did final refinements with dry-
wall joint compound and sandpaper.
The entire process took me around four hours
per door. This was probably twice the time it
would have taken a professional, but it was cer-
tainly quicker than the time to fabricate and in-
stall wooden moldings, and at a fraction of the
cost. In addition to the time advantage, this
process gave me moldings that exactly match
the curve of the door because the molding tem-
plate rode on the actual door jamb. LI

Frank C. Freyvogel is a contractor who lives in

North Beilmore New York and works on Long
Is/and and in New York City. Photos by Tern Fine-tuning the trim. With the template le on the jamb, the author makes the final pass, fill-
Freyvogel except where noted. lag In the last of the voids as he blends the I slues together In the middle.

Finish 45
 Cutting Crown Molding
Calculating miter and bevel angles
so you can cut crown on compound-miter saws

by Stephen Nuding

few years ago, I purchased an 8½-in. makes a 38° angle to the wall. Wait a min- So, why use a compound-miter saw?—You
compound-miter saw. It was light and com- ute... what if my crown doesn't make a 38° are probably wondering why anyone would
pact, but had the same capacity for cutting angle to the wall? want to calculate angle settings for a com-
large crown moldings as a regular 10-in. Fortunately, my daughter's protractor was pound-miter cut when crown molding can
miter saw. Remodeling Victorian homes, I in- in the car, so I was able to measure the angle easily be cut on a regular miter saw with no
stall a lot of crown so this seemed to be the the crown made to the wall by holding it math at all. With a regular miter saw, the
perfect power tool for me. against the inside of a framing square. The crown is positioned at an angle between the
I eagerly brought the saw to the job and set angle was more like 43° or checked all fence and the table (photo below left), but is
the miter and bevel angles for 900 corners, as the crowns I was installing only to find that turned upside down so that the wall face of
indicated by the instruction manual. When I none were the same, varying from 35° to 45°. the crown is against the saw's fence and the
cut my first lengths of crown, the joints I finished the day's work as best I could ceiling face of the crown lies on the saw's
weren't perfect, but I figured that the walls and went home determined to calculate the table. The crown is then cut at 45° to create a
and ceiling weren't perfect either, so with a angle settings for each of the crowns. Using 90° corner, 22.5° for a 45° corner, and so on.
little shaving here and there I was in business. my wife's high school math text to brush up Very simple. (For more on cutting crown mold-
The next crown molding I had to install, on some trigonometry, I wrote down equations ing, see the article on pp. 50-53.)
however, was a larger one, and when I cut it and measurements. I worked late into the Most 10-in, miter saws, however, can only
and held it up to the ceiling, I was looking at night, but couldn't come up with a formula. cut crown molding up to about 4'/2 in. wide.
a pie-shaped gap % in. wide. What's more, I finished the crown job eventually by trial Five and one half-in, crowns are readily avail-
this room had two corners that were 1350, and error, playing with the angles on the saw able, though, and cutting these requires a 14-
not 90°, and the saw's instruction manual gave until they were right. Still, the problem gnawed in. or 15-in, miter saw—a large, heavy tool.
no miter or bevel angles for this situation. I at me. I spent a lot of late nights scribbling Cutting large crowns on any of these saws
soon discovered that throwing miscut pieces and thinking, but I just couldn't get it. also requires the extra step of constructing a
around the room in rage and frustration is a Fortunately, I had hung some French doors jig or fence extension, preferably both.
veiy slow and expensive way to complete a job. in the home of Roger Pinltham, professor of Even a 15-in, miter saw is not big enough
By now I was ready to return the saw to the mathematics at the Stevens Institute of Tech- to cut crown molding more than 6½-in, wide,
dealer and demand a refund. But in despera- nology. So one Saturday morning, at my re- and larger crowns are also available. For in-
tion I grabbed the instruction manual one quest, he graciously came to the house and stance, the Empire Molding Co., Inc. (721-733
last time. According to the manual, the miter we pored over my notes. Several hours later, Monroe St., Hoboken, N. J. 07030; 201-659-
and bevel angle settings were correct for 900 we had it. We could calculate the miter and 3222) makes an 831s-in. crown that I often use.
corners when using a standard crown, which bevel angles for any crown and for any angle. So unless you want to make a king-size
miter box and cut the molding with a hand- The bevel angle (B in our equation) is the where and save a couple of minutes the next
saw, you'll have to use one of the new slide inverse sine of D divided by (the square root time you run that crown.
compound-miter saws or a radial-arm saw to of 2) times C. What if you have a wall corner ttjat is not
cut these wide crown moldings (for more on 90°? To make this calculation you'll need a de-
these saws see FHB #57, pp. 58-62). With a ) I vice for measuring the angle of the wall cor-
compound-miter saw, crown molding is laid To calculate this, multiply the square root of ner. I use the Angle Devisor (manufactured by
flat on the saw table (photo right, previous two (done on the calculator) times C. Then dif- Leichtung Workshops, 4944 Commerce Park-
page). No jig or fence extension is necessary. vide that into D and hit the inverse sine but- way, Cleveland, Ohio 44128; 800-321-6840).
The saws can be smaller for cutting the same ton, or arc sine (same thing) button on the Whether you are installing inside corners or
size crown, resulting in a lighter tool with a calculator. Using the values from drawing outside corners, be sure to use angle of
smaller blade, which is therefore cheaper to the calculations would go like this: the square the inside corner (the angle less than 180°) for
buy and costs less to sharpen. root of 2 = 1.41, times 4.8125 (C) equals the equation.
6.8059, divided into 3.875 (D) equals .5694, the Here's how the equation looks:
Figuring the angles—To calculate the miter inverse sine of which is 34.7°. This is the bevel A
M = tan C x tan (F±2)
and bevel angles for any crown molding, you'll angle at which to set your saw for a 90° corner.
need a framing square and a calculator that's Once you have calculated the miter and If we were to use our crown from A,
capable of doing trigonometric calculations. bevel angles for a particular molding, you nev- we would have 135 (F) divided by 2 = 67.5.
These calculators are usually called "scienti- er have to calculate them again as long as Hit the tangent button and you get 2.4142.
fic calculators." No cause for alarm, though, have 90° corners. Jot down the angles some- That times 4.8125 (C) = 11.6184. Divide 11.6184
just think of yourself as a carpentry scientist. into 2.875 (A), then hit the but-
I use a Radio Shack model that is out of pro- ton, and you get 13.9° (the miter angle).
duction now, but a Radio Shack EC 4008 will Figuring the angles For the bevel angle:
do nicely and retails for only $13.95. A Dx cos (F÷2)
So, here we go. First let's consider the B=sin C
most common case, the 90° corner. Hold r Plugging in some real numbers we get: 135
whatever crown molding you're using up to (F) divided by 2 = 67.5, the cosine of which
the inside of a framing square as in the draw- is .3827. Multiply .3827 times 3.875 (D) and
ing right. Measure lines A, D and C to the you get 1.4829. Divide that by 4.8125 (C), then
nearest 16th of an inch. (To convert fractions hit the inverse sine button, and
of an inch to decimals, simply divide the de- That's your bevel angle.
nominator into the numerator. To convert 3 875 Finally, because the difference of one de-
for instance, divide 8 into 7 and you get .875.) E ang'e of gree in the miter angle or bevel can be
The miter-table setting (M in our equation), is - crown to wall the difference between acceptable and unac-
the inverse tangent of (A divided by C). — F = angle of corner ceptable joints, you must set the angles on
M = tan' (A÷C) M = miter angle your compound-miter saw carefully. Math may
To calculate this, divide A by C, and then hit B = bevel angle for be perfect, but measurements and the real
the inverse tangent button (tan1), or arc tan- 90 corner world aren't, so slight may be
gent button (same thing). In our example, Crown molding varies not only in size needed to get an acceptable joint. But by using
2.875 (A) divided by 4.8125 (C) = .5974. With but also in the angle that it makes with these equations you will avoid the fu$s-and-fid-
the wall. So the first step in calculating dle approach I first used.
.5794 still on the calculator screen, hit the in- miter and bevel angles is to measure
verse tangent button and you get 30.9° (round- the crown with a framing square and
ing to the nearest tenth of a degree). This is determine the measurements shown in Stephen Nuding is a carpenter in ir-loboken,
the drawing above. Then plug those New Jersey. Photos by Susan Kahn.
the miter angle at which to set your saw. figures into the formulas shown below.

j D
90° corner B —— sin C 90 corner M = tan1 (A—C>
Bevel angles Miter angles A
(D x Odd angle corner M = tan
Odd-angle corner: B = sin1 tan (F — 2))
C

Finish Carr.eItry 47
 Table-Saw Molding
The secret is in the order of cuts

by Bruce Andrews

when the landmark Winooski Block was we checked our cutters for sharpness. Our stock
finished in 1862, the builders festooned it with was as straight and as square as we could make
all manner of ornamental moldings and wooden it; we were ready to begin shaping.
filigree. But by time we (Moose Creek Res- 2

torations Ltd.) got the repair contract, 117 Ver- Setting up—Milling complex moldings on a
mont winters had weathered, cracked and split table saw requires precision. Begin with an accu-
all of its remaining woodwork. Three-fourths of rate template of the molding, to which you can
the building's cornice moldings were either rot- adjust the sawblade's settings and against which
ten or missing. We were to replace 10,000 linear 4 you can compare results. The best template is a
feet of various moldings, not one of them a type short piece of the molding you want to copy. If
manufactured today, and we didn't even own you must create a template from molding in
the usual tool for milling moldings, the spindle
shaper. We still were able to complete the job, ( place, you'll have to use a profile gauge. (See
Figure 2 on the next page.) Many exterior mold-
relying on our table saw and a lot of careful plan- ( ings are too large to be handled with one appli-
ning. We found tc,at the table saw could handle cation of the gauge. If this is the case with your
most any profile—it could even scoop out con- trim, you'll have to take a series of readings,
cave curves—but we also learned that every transfer them to paper and combine them for
profile required its own sequence of cuts. Figur- the complete profile. In fact, it's a good idea to
ing out that sequence is the heart of our method. sketch all molding profiles on site, for the gauges
The first thing we worried about was getting may get distorted before you return to the shop.
enough good stock. Molding stock must be the Fashion your template out of a rigid material
highest quality, close grained and knot free. We 5 6 such as Masonite or plywood.
were still short of stock after several deals to ob- Before any cutting, even before setting the
tain a couple thousand board feet of Vermont sawblade, scrutinize the template or molding
pine in varying widths, thicknesses and lengths—
all rough cut and in need of finish planing, di- ) ( cross section. The question is how to determine
the order of the cuts. You don't want to take out a
mensioning, and in some cases, drying. We were piece of stock you'll need later to run against the
bemoaning our plight when two young entre- fence for making another cut. Think things
preneurs wandered into our office. They asked if through on a piece of paper. Certain cuts simply
/
N
we knew anyone who could use several thou- have to be made before others.
sand board feet of redwood and cypress beer-vat In Figure 1, for example, cut 1 is crucial be-
staves from the old Rheingold Beer brewery that cause it is a dividing line between two curves: If
was being dismantled in Brooklyn, N.Y. Well, its angle is incorrect or its cut misaligned, the
yes, we probably knew someone. The wood proportions of both curves will suffer. If it is too
reeked of stale beer, but it was superb for our 7 deep, it undercuts the convex curve; if too shal-
purposes. It was straight, close grained and of low, material in the notch will have to be
course, well seasoned. cleaned out later—a waste of time.
Before any shaping could be done, we had to Cut 2, which creates the concave curve, must
prepare our stock. We thought that the wood meet precisely the high point of cut 1. Because
might have nails hidden in it, but we found none. the stock is fed into the sawblade at an angle,
We did find metal flecks where the vat bands had this is a delicate cut.
deteriorated, but with wire brushes and large Cuts 3 through 6, creating the convex curve,
paint scrapers we removed almost all the rust. must be made after 2. If they had been made be-
On our 16-in, radial-arm saw, we ripped the lum- fore 2, the convex curve would have made sub-
ber to the rough sizes we needed, about in. sequent cuts a problem. (The stock could easily
thicker and '/2 in. wider than the dimensions of roll on that curve as it is fed into the sawblade.)
the finished moldings. Next we prepared the Cut 7 is delayed so that the point it creates with
stock on a jointer and a thickness planer. Once cut 2 won't be battered as the stock is maneu-
we had dressed down the old surfaces '/., in., the vered over the saw. Cuts 8 and 9 are made last,
wood was perfect and unmarked. As we worked,
Figure 1: Sequence of cuts
because leaving the corners of the stock square
Illustrations: Carol Hubbard

48 The Best of Fine Homebuilding

 The Winooski Block (left) is capped by
a cornice assembly over 6 ft. wide; it
consists of 14 elements, including 7
moldings that were reproduced on a
table saw. The milling of the molding is
described on the facing page.

You need a template to mill new

moldings. Use a piece of the original,
or transfer readings from profile
gauge to paper on site and cut a
template later in the shop.
Straight 2x8 serving
as diagonal fence

Stock supported by board bolted to bon't push stock across

posts at exact height of saw table. the blade at an angle
Board braced at angle to direct stock. greater than 60°

Concave curves may require several passes, starting

Figure 3 Fingerboards to hold with the blade set low. On the last cut.
Cutting setup stock against fence saw points should just touch curve outline.

ensures the stability and accuracy of preceding rip fence running parallel to the blade on one would push wood across the blade is 60°; with
cuts. (Cuts 3 through 6 would have been almost side or the other. As shown in Figure 1, cuts 1 wider angles, not enough of the sawteeth are
impossible if cut 9 had preceded them.) and 2 were made with the stock face down on the gripping and the blade will bind. (I'm not sure
To save time, pass all molding stock through a table, while cuts 7, 8 and 9 were made with it why, but a 48-tooth carbide blade binds up less
given'saw setting; be fastidious about such set- face up. The stock stood on edge for cuts 3 than an 82-tooth one. It may be that the chips
tings, making practice cuts on scrap work. Cut through 5. (When cutting some symmetrical co9- clear more easily.) If the blade is binding, make
more molding stock than you'll need at each set- vex shapes, you can leave the sawblade at the several passes to get the curve, starting the
ting, so you'll always have waste stock with the same angle, and after one pass, turn the stock blade low and crankng it up 1/4 rn. for each pass.
necessary previous cuts. In other words, to get 180° to get the cut whose angle mirrors the Don't get so wrapped up in your calculations
an accurate setting for cut 5, you'll need stock first.) Each cut was preceded by carefully adjus- that you become careless. Keep fingers clear of
with cuts 1 through 4 already made. ting blade height and angle against the template. the blade. The speed at wltich you feed the stock
The last cuts on a molding (cuts 8 and 9) must be determined on the job: Too fast and the
Cutting—We used a 10-in. Rockwell Unisaw be slightly larger than 45°—if your sawblade blade will bind, too slowly and the wood will
with a 48-point carbide-tipped blade for all mold- will tilt just a little more—to avoid gaps where burn. The greater the angle of feed, the more
ing cuts. For most cuts we used the rip fence pro- the building surfaces are not quite often you should clean the blade.
vided by the manufacturer. For cut 2 however, We cut the concave shape (cut 2) into the mold- The quality of the wood greatly affects the
we needed a diagonal fence, so we trued a 2x8, ing by passing the stock diagonally across the complexity of cuts you can make. Hardwoods
used a template to carefully set it at the proper table-saw blade. (See Figure 3, at right.)To set are more difficult to mill without proper equip-
angle for the desired cove, and clamped it to the the blade and fence correctly, you'll need a piece ment. If concave curves are possible at all on
table with 10-in. Jorgensen C-clamps (Figure 3). of the old molding. (A template is less hardwood, you'll have to make many gradually
To reduce stock flutter we used fingerboards, Holding the high point of the curve over the increasing cuts; the angle of the stock to the
pieces of wood with a series of parallel kerfs cut blade, slowly crank up the blade so that the tip of blade will be limited. Fortunately the grain in
in one end. Two fingerboards clamped to the the highest tooth just grazes that curve's apex; our cypress varied less than in. in 15-ft.
table held the stock against the fence, while one lock the setting and try a few cuts. To create the To refine the shape of our convex curves, we
fingerboard clamped to the fence held the stock width of the curve, angle the piece of molding used many tools, including jack planes, curved
down. The kerfs allowed enough play to let the until all the teeth of the exposed blade lightly shavehooks and spokeshaves. Among power
wood slide through, but maintained enough touch the arc of the molding. Another person sanders, Rockwell's Speed-block was the
pressure to ensure a straight cut. Using finger- should snug the fence against the angled mold- favorite; we clamped the finished molding to
boards and extension tables, you could cut all ing and then clamp the fence to the table while benches and sanded it using 50-grit pads.
the molding unassisted, but you may prefer to you hold the molding in place. You'll have to Using these techniques we milled 1,000 linear
have a helper to pull the stock gently through tinker a bit to get the exact angle you need. feet for each of nine molding types, some more
the last few inches of a cut. Several times a day, You can create almost any symmetrical curve complex than the one described above.
wipe the tabletop and sawblade clean with tur- with this method. Pushing the stock across the
pentine, to minimize binding. blade at a wider angle will result in a wider Bruce Andrews is a partner of Moose Creek Res-
Except for cut 2, all cuts were made with the curve. However, the widest angle at which we torations Ltd., in Burlington, Vi.

Finish Carpentry 49
 Installing Crown Molding
Upside down and backwards is the secret

by Tom Law

The first piece of molding (left) is cut square and run into the corner. The second piece (right) is cur to me snape of the moldings profile (coped) and
will butt neatly into the face of the other piece. The paper-thin point on the bottom of the coped piece will make the finished joint look like a miter.

he old-time carpenters I learned from used face is visible, so much of the solid back has cuts are made through the face of the molding,
to amuse themselves by quizzing young ap- been eliminated to save material. Also, by and a sharp handsaw will do a better job than a
prentices about the trade. If you could answer eliminating part of the back, only two small dull circular-saw blade will do.
the easy questions, the last question would al- portions of the molding bear on the wall and For this kind of work, I prefer a workbench to
ways be: "How do you cut crown molding?" ceiling surfaces, which makes crown easier to a sawhorse. Mine is just a simple frame of 2x4s
And when you looked puzzled, they'd go off, fit to walls and ceilings that aren't straight or and lx4s with a 2x12 top. It stands 34 in. high,
chuckling to themselves something about "up- that don't form perfect right angles. which is a more convenient height to work on
side down and backwards." Of all the different than a sawhorse provides. You don't need to
moldings, crown molding is the most difficult The crown tools—When I cut crown, I like to deliver a lot of power to cut trim. A broad
to install, largely because of how confusing it work right in the room where the molding will bench top is also convenient for holding tools.
can be to cope an inside-corner joint. go so I can orient myself to the wall I'm work- Although you can still find deep-throated
In classical architecture, crown molding ing on. If a room is finished, however, I may coping saws in the mail-order catalogs, most
(sometimes called cornice molding) is the up- have to do the cutting somewhere else. Then I coping saws nowadays are 5 in. deep and
permost element n the cornice, literally crown- have to imagine the molding in place when have a 6-in, blade. The blades come with dif-
ing the frieze and architrave. These moldings I'm positioning it in the miter box (and believe ferent numbers of teeth. try each kind of
I

were functional parts of the building exterior me, this can get tricky with crown molding). blade to see which works best with the wood
when the ancient Greeks used them, but they I cut and install crown molding with hand I'm cutting. Generally, finer teeth work best
have been used icr centuries on interiors purely tools. I use a wood miter box (top left photo, with hardwood and coarse teeth with soft
as decoration. p. 52) because it's the kind I learned on, but wood, but not always. For the job shown here,
Crown molding is installed at the intersec- also because my view is not obstructed by the I used a fine-tooth blade to cut soft wood. It
tion of the wall and ceiling. Originally crown electric motor of a power miter box. Installing works more slowly, but makes a smooth cut.
molding was triangular in cross section—the crown molding is slow and calls for careful The blade of a coping saw can be inserted
portions abutting the wall and ceiling formed work, so the production speed of an electric with the teeth directed toward you to cut on
two sides of a right triangle, and the molded miter box is not required. I cut miters with a the pull stroke, or away from you to cut on the
face was the hypotenuse. But only the molded standard 26-in, handsaw (10 or I I point). Miter push stroke. Although it's strictly a matter of

50 The Best of Fine Homebuilding

personal preference, orient mine to cut on
I wall opposite the door (bottom drawing). Un- Occasionally I need to pull the Lop edge of
the push stroke because it acts in the same less it's perfect, a coped joint looks better the crown tight against the ceiling, but there's
manner as a handsaw. from one side (looking toward the piece that no ceiling joist or blocking to nail into. When
was butted) than it does from the other that happens, I use lGd finish nails to reach all
Measuring and marking— Crown molding (looking toward the piece that was coped). the way to the double top plate on the wall. Or
can be used by itself or combined with other By first installing the crown molding on the sometimes I put a little glue on the molding
moldings, but it should always be in propor- wall opposite the door, and coping the mo!d- and drive a pair of 6d finish nails at converging
tion to the size and height of the room in ing into it on both ends, the two most visible angles into the drywall about 1/2 in. apart
which it's installed. Too much molding at the joints show their best side to anyone entering (drawing next page). This pins the molding to
ceiling line tends to lower the ceiling visually. the room. the drywall while the glue dries. Another trick
Three or four inches of molding at the ceiling I put the first piece up full length with for nailiEg up crown when you don't have ade-
line is about right for an average-size room. square cuts on both ends. I cut it for a close quate framing is to nail up triangular blocks as
In the rooms shown in this article, I used fit, but if it's a little short I don't worry. Any shown in the drawing on p. 53.
3%-in, crown molding, which is the most small gap will be covered by the coped end of I usually work around the room from right
common size available. This dimension is the intersecting piece. I hold the molding in to left because I'm right-handed, and making
the total width of the molding, but it's not the place, lining up the top edge on the 2'/,s-in. coped joints on the right is a little easier than
critical dimension that you use when install- mark, and nail it. I use the shortest finish nails on the left. The second piece of crown to go
ing crown. You need to know the distance that will reach the framing, usually 6d or 8d. I up needs to be coped on the right and
from the intersection of the wall and ceiling nail into the wall studs through the flat square cut on the left,.
to the front of the molding, measured along of the molding near the bottom, and I nail
the ceiling. Because the back part of the the ceiling joists or blocking through the end Coping with crown—Finish carpentry can be
molding has been eliminated, you can't mea- of the curve near the top. I don't nail too harder at times than cabinet work because you
sure this directly. Instead, I put the molding to the ends when I first put the piece up. I often have to make perfect joints against im-
inside a framing square to form a triangle leave them loose to allow for a little alignri Lent perfect surfaces. Coping inside joints rather
and read the distance (top drawing, right). with the intersecting piece. than mitering them is one way to deal with that
The molding shown here measures 2'/16 in. I problem. If you miter the inside corner with
mark that distance on the ceiling at each cor- crown molding, the joint will often open up
ner of the room and in several places along when you nail the pieces because the wall
the walls. These marks will serve as a guide gives a little. A coped joint on crown molding
when I install the molding. won't open up and will be tight even when the
It's frustrating to drive a nail through a piece walls are not exactly 900 to each other.
of molding and not hit anything more solid A coped joint is like a butt joint, with one
than drywall, so I locate the framing members piece cut to fit the profile of the other (photo
ahead of time—when I can still make probe facing page). The first piece of molding is cut
holes in the wall that will be hidden by the square and run into the corner. The second, or
molding. If the room hasn't been painted, you coped piece is made by cutting a compound
can spot the studs and ceiling joists from the miter on the end to expose the profile of the
lines of joint compound and make pencil molding, then sawing away the back part of
marks on the wall to guide the nailing. If the the stock with a coping saw, leaving only the
room has been painted, you can find the studs profile. The end of the coped piece will then
and joists by tapping with a hammer and test- butt neatly into the face of the first piece.
ing with a nail. Electrical outlets and switches Because I don't always get the cope right
are nailed into the sides of studs the first time, I start with a piece
and offer a clue to stud locations. of molding longer than I need
Running crown molding should and cope the end before cutting
be one of the final jobs on a new it to length. The phrase "upside
house. Walls and molding should down and backwards" refers to
be primed and first-coated, the the position of the crown molding
molding installed and then finish in the miter box when you're cop-
coats applied. If you're retrofit- ing it (photo next page, top left).
ting crown molding, it should be Crown molding isn't laid flat
prefinished entirely so that all you against the side or bottom of the
need to do is touch up the paint miter box; it's propped at an an-
or stain after installation. gle between the two, just as it
On this job the walls were will be when installed. But the
painted and the molding was edge that will go against the ceil-
prestained, so I marked the stud ing is placed on the bottom of
locations right on the crown the miter box and is therefore "up-
molding. Rather than use a pen- side down." The right-hand side
cil to mark the wood, I made a (if that's the coped end) is placed
slight hole with the point of a on the left and is "backwards."
nail, which was easier to find in The crown has to be posi-
the dark stain and which I later tioned so that the narrow flat
nailed through. sections on the back of the
molding, which will bear against
Getting started—When I run the wall and ceiling, are square
crown molding in a typical against the side and bottom of
room—four walls, no outside the miter box. When they are, the
corners—I usually start with the bottom should measure out the

Drawings: Michael Mandarario Finish Carpentry 51

 Before the crown molding can be coped, the end must be mitered to
expose the profile. Positioned "upside down and backwards" in the
miter box, the molding rests against small nails that hold it at the prop-
er angle (top left). After exposing the profile of the crown molding, the
back part of the stock is cut away with a coping saw, which must be held
at a severe angle or the coped joint will not be tight (left). Even with the
coping saw cutting at a severe angle, it's tough to remove enough wood
through the S-curve in crown molding. Additional stock often has to be
pared away with a utility knife (above).

required 2'/16 in. Once I find this position, I I always test the cope against a scrap piece wall, but I find it easier to measure from
usually draw a pencil line on the bottom of of molding to make sure I'm in the ballpark either of the two vertical flat sections on the
the miter box to help me position subse- before actually trying it in place. Despite my molding that the coped piece will butt into. If
quent pieces. Sometimes I'll even put a few best efforts to undercut the curved section, I I'm working alone, I either step off the mea-
nails on the line or glue a strip of wood to it. usually have to pare away some more wood surement with a measuring stick (a 12-ft. rip-
Even with the molding positioned correctly with my utility knife (photo top right). ping, for instance), or I'll drive a nail into the
in the miter box, it's still easy to cut it wrong. I cut the piece just a little long and test it wall (above the line of the crown molding)
When I make the 45° cut to expose the pro- in place before cutting it to final length. If the and hook the end of my tape measure over it.
file of the molding for the cope, I remind my- fit of the coped joint is close, but still a little Wherever I measure from on the wall, I'm
self that I want to cut the piece so the end off, I can sometimes improve the fit by twist- careful to measure to the same place on the
grain will be visible to me as I look at it in ing both pieces either up or down the wall at piece I'm cutting.
the miter box (photo top left). this point—the 2'/16-in, mark on the ceiling When the coped piece is cut to length, I
Coped joints are always undercut slightly, isn't sacred. The buildup of spackle or plas- nail it up just like the first piece, leaving the
but crown molding has to be heavily under- ter in corners can distort the intersection of square-cut end unnailed for the time being.
cut through the S-curve portion of the crown wall and ceiling. Some carpenters carry a If I need to draw the coped joint tighter, I
(called the cyma recta) or it will not fit right. small half-round file with them to fine tune nail through the coped piece into the piece
I start the cope at the top of the molding the fit of the cope. it abuts.
using light, controlled push strokes. If I'm The third piece of crown molding goes up
having trouble going from the straight cut Around the room—Once the coped joint just like the second, but the fourth one needs
to the curve, I back the saw out and come it's time to cut the piece to length. You
fits, to be coped on both ends (bottom drawing
in at a different angle to cut away the waste. can measure the total distance from wall to previous page), assuming the wall is short
I begin the curved line with a heavy under- enough to be covered with a single piece of
cut and hold this angle all way through. I molding. I cut this piece about '/16 in. longer
cut as close to the profile line as I can than the actual measurement, bow out the
(photo above left). middle, fit the ends and snap it into place.
The bottom of the crown molding is made The extra length helps to close the joints.
up of a horizontal flat section, a cove and a Some carpenters don't like having to cope
vertical flat section. 1 cut down to the upper the last piece on both ends because there's
fiat and then take the saw out and start cut- very little margin for error. The way to avoid
ting from the bottom. Some carpenters sim- this goes all the way back to the first piece of
ply square off the bottom, but I try to leave the crown molding that's installed. Rather than
little triangular piece intact (photo, p. 50). 1 put up the first piece with square cuts on
support it with my thumb as I'm coping and both ends, you can temporarily nail up a
slice it paper thin. This little piece makes the short piece of crown molding and cope the
coped joint look like a miter and helps close first piece into it (photo left, facing page).
any small gap if the first piece didn't fit tight- Then take down the short piece, work on
ly to the wall. around the room and slip the butt end of the

52 The Best of Fine Homebuilding

If you install the first piece of crown molding in a room by cutting both
ends square, the last piece will have to be coped on both ends. To avoid
this, you can put up a short piece temporarily and cope the first piece of
crown molding into it (above). If an outside miter is open just slightly,
sometimes you can close it by burnishing the corner with a nail set (top
right). When a line of crown molding has to be neatly terminated on an
open wall, the end should be mitered and "returned" into the wall with
a small piece of molding. To avoid splitting such a delicate piece, it's
best simply to glue it in place (right).

last piece behind the first cope that you Sometimes outside corners will close tight- stair and ended the molding with a return—a
made. This way all four pieces of crown ly but the leading edge of one piece over- mitered piece that caps the end of the mold-
molding in the room will have one square- hangs the other, perhaps because the corner ing. To make a return, I simply cut a miter for
cut end and one coped end. is not exactly 90° or because one piece of an outside corner on the end of a scrap of
When I go into a room that's not a simple molding is thicker than the other (more molding, then lay the piece face down on the
rectangle, the decision about where to start about that in a minute). If the molding hasn't bottom of the miter box and cut off the end. I
is influenced by where I'll end. If there is an been painted or stained, I'll trim the over- glue this in place with white glue so as not to
outside corner in the room, I like to end by hanging edge with a sharp chisel and sand it. take a chance on splitting it by using a nail
installing the shortest piece that has an out- This actually leaves a narrow line of end or brad (photo above).
side miter. That way, there's less wood wast- grain exposed at the outside corner, but once
ed if I cut it too short. If there's not an out- the molding is stained or painted, the end What can go wrong?—Whether because the
side corner, like to work so that the last
I
grain isn't very obtrusive. There are times wood was wet when it was milled, or because
piece is installed on the longest wall that can when the molding has the finish coat already the knives were dull, or because of internal
still be done with a single length of molding. on it, and I can't do this because it would stresses in the wood, the exact dimension and
When I need more than one piece to reach expose raw wood. In that case, I use my nail profile of the pieces in a given bundle of stock
from corner to corner, cut the moldings
I set to burnish the projection smooth (photo molding varies considerably. The differences
square and simply butt them together rather top right). aren't obvious until you try to fit an inside or
than use scarf joints or bevel joints. Butt On this house, I ran crown molding in the an outside corner with two pieces that don't
joints are easier to make for one thing. And foyer and had to terminate the molding at the match. It's best to make joints from the same
for another, although wood isn't supposed to stairwell opening. I ran the molding through piece whenever possible.
shrink in length, the truth is it does. Over the the dining room, turned the corner at the There are times when the wall or ceiling is
years, I've seen a lot of joints that have so crooked that gaps are left along the length
opened up, and of those, the butt joints of the crown. If there is a short hump that
looked better than the others. causes gaps on each side, I scribe the mold-
Nailing tip ing and plane it for a better fit. lithe gaps
Outside corners—These are also mitered When nailing crown molding, walls aren't too bad, it may be best to fill them with
parallel with the coiling joists are often a caulk. Another trick I've used is that of leav-
with the molding upside down and back- problem. A length of 2x4, Cut to fill the
wards in the miter box, but the saw is angled space behind the crown (slightly smaller) ing a small space (usually about in.) be-
to bevel the piece in the opposite direction. and spiked into place, will provide good tween the top of the molding and the ceiling,
nailing. —S.A. Insorra, Jamestown, NY. which makes it harder for the eye to pick up
When you miter for a cope, you expose the
molding's end grain, but with a mitered out- irregularities. If I'm doing this, I put up
side corner, the end grain is behind the fin- blocks to nail to, as shown in the drawing at
ished edge. Sometimes I cut them at an angle left, and use a '/4-in, spacer block to ensure a
slightly greater than 45° to ensure that the uniform reveal. LI

outside edges mate perfectly. I usually add a

little white glue, then nail through the miter, Consulting editor Tom Law is a carpenter and
top and bottom, from both sides. builder in Frizzellburg, Mart/and.

Finish Carpentry 53
 Installing Two-Piece Crown
A method for running wide, paint-grade crown moldings

by Dale F. Mosher

work as a finish carpenter on means no caulks or putties to fill

the San Francisco Peninsula, any gaps that may occur at rniters
where there is a resurgent interest and along uneven walls. Paint-
in formal houses that have a Re- grade material, on the other
naissance European flavor. The hand, will have some sapwood
houses often have a full comple- and uneven grain. On the wall, it
ment of related molding profiles can have caulkable gaps. Obvi-
for base, casings and crown, and ously, the stain-grade material
to be in scale with the rest of the will cost more—how much more
building, these profiles can be depends on the species of wood.
quite wide. In the case of the And it costs a lot more to install.
crowns, I'm talking 10 in. to 12 in. Where I work, we figure four to
wide. In fact, the crown moldings five times more labor is needed
that I sometimes install are so to put up stain-grade crown as
wide they come in two pieces opposed to paint-grade.
(photo at right). The most commonly used
There are several reasons for paint-grade materials in these
making crown in two sections. parts are alder, poplar and pine. I
First, the machines that cut the see more poplar than anything
moldings typically have an 8-in. else because it's relatively inex-
maximum capacity. There is a lot pensive and easy to mill. Even
of waste when wide moldings are paint-grade moldings, however,
carved out of a single piece of don't come cheap, and they
stock. For example, I'd need a should be handled with care. I
3x12 to mill a 10-in, wide piece of have them primed on both sides
crown — an expensive, inefficient as soon as they are delivered, and
use of the resource. Two-piece I store them on racks with sup-
crowns are also a little more for- Joined at the shadowline. Wide crowns, which appear to be made of ports no farther apart than 3 ft.
giving during installation. The a single molding, can be made by running related profiles adjacent to
one another. Any gaps between the two are caulked prior to painting.
type I used on the job shown here Miter-box station—I've used the
can be overlapped in and out a new sliding compound-miter saws
bit, allowing the width of the crown to grow to cut crown, and I've decided to stick with
and shrink as needed to account for dips and my 15-in. Hitachi chopsaw. Here's why: When
wows in the walls and ceilings. you're running crown, you've got to make both
All the two-piece crown moldings I've en- back cuts and bevel cuts. It takes time to ad-
countered have been custom-made. The de- just the saw back and forth, and the constant
signer or architect comes up with section changes multiply the chances for error. Also,
drawings; then the mill shop has the molding- the crown has to lie flat on the table with a
cutter knives cut accordingly. Here, an aver- sliding saw, which makes it harder to see the
age set of custom knives costs $35 per in., path of the blade and the cut line. None of
plus there is a $75 setup charge for each pro- these is a problem when using a chopsaw.
file. So before the wood starts to pass over the The key to cutting crown accurately is hav-
cutters you've already spent a fair amount of ing a good station for the miter box (top
money. But to create a certain look, it can be photo, next page). Mine has a pair of wing ta-
money well spent. bles that flank the saw, connected by a 3/4-in.
MDF (medium density fiberboard) drop table
Stain-grade or paint-grade trim—If you've that supports the saw. Each wing table is 6 ft.
got one, your architect or designer will decide long and 2 ft. wide. They can easily be stood
what grade the trim should be. If you don't, on end and carried through a standard door-
your checkbook will decide, At the mill, stain- way. I can also put a 2-ft. deep table against a
grade means the stock is clear and virtually wall and have enough clearance behind the
free of knots. On the wall, stain-grade means Test fit. After affixing the top portion of the saw to swing it through its settings.
no opaque finishes will be applied. The mold- crown to its backing blocks, Mosher checks The wing tables have fences that support
ings are individually scribe-fitted, and that the fit of its corresponding bottom half. backer plates for the crown during a cut. To

54 The Best of Fine Homebuilding

install them, I begin by snapping keel (a carpenter's crayon) along
a line on the floor where the as- the ceiling/wall intersection. A
sembled station will sit. Then I keel will bleed through most
put the outside legs of the tables primers. Omission of this step
on the line, and anchor each leg gets you a one-way ticket to the
to the subfloor. Next I put the planet of frustration, where you
miter box in the drop section of poke nails into the walls and ceil-
the table. I removed the stock ings, looking for the lumber.
fences from the miter box, allow- Backing blocks are installed on
ing me to bring the arbor slightly layout lines snapped on the ceil-
forward of the original fence line, ing and wall. Using a torpedo lev-
thereby increasing the width of el with a 45° bubble, I position a
the cut. The wider of these two block at one end of the wall so
crowns was 7'/4 in—just about the the bubble reads level. The block
limit of what my saw can handle. should be about 2 ft. from the
I make sure the saw's turntable corner to avoid joint-compound
can swing freely from side to side, buildup. I mark its edges on the
and that its blade is square with ceiling and wall, and repeat the
the tables when set at 00. I secure process at the opposite corner.
the saw to the drop table with These points are the registration
four drywall screws run through marks for the chalkline. I use the
holes that I've drilled in the saw's raised chalkline as a straightedge
base. Each wing table has an 8-in. to locate dips or bumps in the
high fence that's square to the ceiling and walls. These problems
blade. align the fences with a
I
are usually due to framing irregu-
string to make sure they're straight. larities, and the backing blocks
During a cut, the crown bears should be kept away from these
against 3/4-in. MDF backer plates places. I wish the framing crew
that are screwed to the tables and could be around during this part
the fences. The backer plates of the job. If they only knew the
should be ½ in. wider than the trouble we go through to make
widest crown section. The crown them look good, they'd be taking
moldings for this job meet the wall trim carpenters to lunch a lot.
and ceiling at 45°, so I ripped the To attach the backing blocks I
backer-plate edges at 45°. Other use a 2¼-rn. finish nailer because
crowns meet the wall and ceiling finish-nail heads are small
at different angles, and the back- enough to be consistently set be-
er-plate edges should be beveled low the face of the block. Nail
accordingly. Miter-box station. A pair of wing tables linked by a heads that stand proud interfere
It took me about a day to build vides angled backer plates to support the crown moldi with the crown. When I've got
cut. The plates are screwed to the table and fences. framing on one end of the block
this setup, and another half day to
fine-tune everything. But the time it for nailing, but none on the other, I
takes to build one will be returned put a bead of glue on the backing
tenfold in a single good-sized installation job. block to help anchor it to the wall.
At inside corners, I run one block into the
Backing blocks—A two-piece crown needs a corner, and then scribe the adjoining one to it.
solid base for nailing and a flat surface to rest The backing blocks are typically a little too
against to ensure correct alignment for the wide to fit between the lines in the corners be-
pieces. Backing blocks serve this purpose (top cause of joint compound on the wall, and they
left photo, facing page). To find the width of need to be trimmed a bit to fit. Outside cor-
the backing block, I assemble a couple of ners are sometimes mitered as though crown
short sections of crown, as they would appear molding, and secured to the wall, ceiling and
when installed, and measure the backside from to each other through their mitered edges.
the point the assembly hits the ceiling to the When the backing blocks are up, I cut my
point it engages the wall. The backing blocks "tester blocks." These are typically 16-in, to
should be in. less than this measurement 24-in, long pieces of the crown molding. They
to ensure that the crown will go together with- need to be long enough to reach from an in-
out leaving gaps between the pieces, at the side corner to the dosest midspan backing
ceiling, or at the wall. Backing blocks can be block. I cut three pairs of tester blocks with
made of solid wood, but prefer 3/4-ifl, ply-
I
inside miters at both ends, and three pairs
wood because it's affordable, doesn't split and with outside miters. One set has 44° miters at
it holds nails well. After ripping a stack of each end, one has 45° miters and the third has
backing-block stock, I cut the blanks into 6-in. 46° miters. You may ask, "why not cope the
to 8-in, lengths. inside corners?" For one, the curved profile of
I prefer to place backing blocks on 16-in.
the widest molding in this job meant that a
centers, and no farther apart than 24 in. They Dropped platform. The wing tables are joined coped corner would be very fragile. I've found
by a saw platform lowered far enough to bring that a glued inside miter on a paint-grade
should be affixed to the framing, so if the the saw's table flush with the wings. The fences
painting crew is about ready to prime the are braced from behind with triangular blocks job—if the pieces are carefully fitted—yields
walls, I mark stud and joist locations with a on 1-ft. centers. first-rate results.

Finish Carpentry 55
Baciung blocks. Crowns this wide need sub-
stantial backing to provide a consistent plane
for aligning the two pieces and for adequate
nailing. In the corner, one block extends to
the wall while the other is scribe-fitted to it.
Here, a test piece of crown is held in place.
The pencil line along its point intersects the
corner formed by the two blocks, marking the
point from which the overall measurement for
the crown will be taken.

Prybar tweaking. As the moldings are nailed Unu assembly. Short sections ot crown are best preassembled into a single piece. The pencil
home, a small prybar is useful for aligning marks on the ceiling show the points from which the crown-length measurements were taken.
the adjoining sections. On the right you can see a fully assembled run of crown.

Running crown—Installing crown is not a folds we typically erect as work platforms. short sections around a wing wall. In this case,
solo operation. The job will go a lot faster and We also use tester blocks at outside miters I usually preassemble the pieces if they're
with greater accuracy if you've got a good to determine the best angles. To get our mea- shorter than about 12 in. I put the parts togeth-
helper. Working on the theoty that a piece of suring points for an outside-to-outside miter, er with glue and a pneumatic brad nailer, let
trim can always be made shorter, we begin we make pencil marks on the ceiling to note the glue set for 20 minutes and then place it
with the longest run in the room by tucking the long points. For an outside-to-inside miter, as a unit (photo above right).
the pair of 45° test blocks into one of the cor- we mark the long point of the outside miter, As we run the upper crown, we make notes
ners, just the way the finished crown will fit. If and the heel cut at the inside miter. in the corners that describe any special angles
the fit isn't acceptable, we try a 44° and a 46° I'll typically put four 11/2-in, finish nails into or back-beveling that it took to get a good fit.
block until the right combination turns up. It each backing block. The nails should be placed Nine times out of ten, the same cuts will work
might be a pair of 44s. It doesn't matter. It is where the painter can easily putty the nail- on the lower section of crown.
very important, however, that the line of the heads. I don't put nails in a tight radius or too After the crown is up, the drywallers can
miter line up with the corner, whether it's an close to an inside corner. A small prybar can float on any necessary topping compound to
inside or outside miter. be useful for aligning the crowns during nail- hide the bumps and bows in the ceiling and
Once we find the best fit, we make a pencil ing (photo above left). I prefer the ones used wall. If the wails are to be textured, this should
mark along the bottom of the block into the by auto mechanics. be done after the crown is installed. Our paint-
corner (photos above). This marks the point Back-beveling the miters can be useful on ers use oil-base putty to fill the nail holes, and
from which the overall measurement is taken. recalcitrant fits. A good tool for this is a 11/8-in. latex-based paintable caulk to make the joint
I don't bother to cut the piece a little long belt sander, Its protruding belt makes it very between the two pieces of crown disappear.E
and then shorten it by degrees to ease into the maneuverable. If I need shims, I use pieces of
fit. My helper and I can measure it accurately, manila folder. At each miter, I run a bead of Dale F. Mosher is a carpenter who specializes
so I cut it to that length. Period. This saves a yellow glue to ensure a sturdy joint. in finish work in Palo Alto, California. Photos
lot of climbing up and down the A-frame scaf- Sometimes the crown has to work its way in by Charles Miller.

56 The Best of Fine Homebuilding

showed

Making Curved Crown Molding

Glue-laminated trim can be shaped on a table saw

by John La Torre, Jr.

s a carpenter, I spend most of my time at laminate it. As it turns out, Kreutzfeldt decided to Making the bending form itself was the easy
work swinging a hammer or wielding a saw. But paint it (photo above). pat t. Back at my shop, I traced the outline of
I'm always looking for a chance to try a new the template onto three pieces of /1-in plywood,
technique. Recently, while I was touring a house Making the bending form—The first step cut them out and nailed them together, placing
under construction, owner Paul Kreutzfeldt was to make a form that matched the curva ure /1-in plywood spacers between layers to pro-
me a straight piece of stock crown mold- of the convex wall. Finished with drywall, the di.ce the proper thickness. Finally, on the back
ing he planned to use for the kitchen ceiling, wall defined a 900 arc having a radius of about edge of the form, I bandsawed a series of steps
then looked up at a curved wall in the room and 241/ in. Unfortunately the curve was far from per- parallel to the front edge to give the c amps good
said, "That piece is going to be a bear." fect, wandering out of round by up to 1/ in. That purchase (photos next page).
"Yup," I answered, excited. "Mind if I give forced me to make a template for the form.
it a try?" To create the template I bandsawed a in. Preparing the stock—The next step was to pre-
"Go right ahead," he said with a smile. radius curve in a sheet of /-in. tempered pare thin strips of wood for lamination. The fac-
There are two basic approaches to making Masonite. I then held this template against the to.'y-made crown molding I wanted to match was
curved trim. You can glue several pieces of wood curved wall 3/4 in. from the ceiling which is made of white pine, but I selected clear sugar
end-to-end and then cut the curve on a band- where the bottom of the crown molding would pine for m molding. Sugar pine has a uniform
saw, or you can laminate thin strips of wood contact the wall. After scribing the template straight grain is easily bent without and
around a curved fonn. If you use the method with a pencil compass, I trimmed it with a jigsaw well that's what I had on hand.
and decide to stain the trim, the separate pieces for a snug fit. I oduced the laminating stock by resawing
may accept stain differently, and the joints usu- /-in by 4in. wide boards, which produced
ally show through. Laminated trim, on the other thin boards that were in. thick. I used my
hand, is stronger than butt-joined trim and usu- Molding in the round. Made with basic shop bandsaw for ripping the boards into strips
tools, the laminated crown molding seen above because its /1 -in. saw kerf wastes less wood
ally looks better when its stained Because wraps around a slightly out-of-round convex
. I

Kreutzfeldt had yet to choose between staining wall, butting at both ends into straight, factory- than my table-saw blade. Then I rar the strips
and painting his crown molding I decided to made crown. Photo by Rich Miller. through a 10-in, bench planer to remDve irregu

Finish Carpentry 57
Glue up. The laminating form consisted of three layers of 3/4-lfl, plywood separated by
3/4-ln, plywood spacers (photo above). A series of steps bandsawn into the back of the
form provided solid footing for an assortment of clamps. Extra strips of pine (photo be.
low) placed against the molding stock helped distribute the clamping pressure.

Low-tech shaping. The molding was contoured by mak-

ing a series of cuts on a table saw to remove the waste up
to the layout line. A mark on the rip fence indexed the
center of the saw blade, indicating the optimal spot for the
molding to contact the fence during the cutting operation.

58 The Best of Fine Homebuilding

larities and to reduce the strips to a uniform that matched my molding. I could have ordered suits (photo below). Each pass left behind a
thickness of 1/8 in. custom-made cutters, but that would have cost small trail of fine dust instead of the curled shav-
Next, I traced the cross section of the factory $300 to $400, difficult to justify for a one-off piece ings a perfectly tuned scraper would produce,
molding on graph paper. Examining this profile, I of trim. but little by little the sharp steps began to disap-
decided to laminate the trim out of three differ- I decided to shape the blank by making a series pear. Scraping the molding down to the layout
ent widths of sugar-pine stock to simplify the re- of table-saw cuts to remove most of the waste line took two hours and a lot of elbow grease.
moval of waste from the laminated blank. (photo facing page). This worked remarkably The scraping left the molding with some torn
With the strips cut to width, I dry-clamped them well. I clamped a 33/4-in, tall board to the fence to fibers and minor irregularities, so I decided to
to the form to identify and eliminate any prob- make it the same height as the molding, then finish the job by sanding. I began by using
lems before glue up. I discovered that keeping marked the top of the fence to index the center- 80-grit sandpaper to work out the irregularities,
the strips aligned would be difficult. My solution line of the saw blade. While making each cut, I then worked my way up to finer-grit paper. I
was to trace a slightly oversized cross section of held the blank against the fence at the index use Wetordry TRI-M-ITE sandpaper (3M Construc-
the three-step assembly on two scraps of ply- mark. Any deviation from this mark was insignif- tion Markets Department, 3M Center, Bldg. 225-
wood, cut the patterns out and then slip the icant, because it merely caused the saw blade 4S-08, St. Paul, Minn. 55144; 612-736-7761) be-
scraps over opposite ends of the assembly. These to wander into the waste area, requiring nothing cause its backing doesn't tear while sanding. The
simple jigs helped to prevent the wood strips more than a second pass across the table saw to sanding took about three hours and just about
from sliding around during glue up. get it right. wore me out.
My blade cut a '/8-in, wide ken, so I made the
Shake It up—For laminating jobs, I like to use cuts by moving the fence toward the blade in Installation—Earlier, when trimming the ends
urea-formaldehyde glue, which starts as a tan- 'A6-in. increments, raising the blade just enough of the curved blank, I had left the ends l'/2 in.
colored powder that must be mixed with water each time to remove the maximum amount of long. Now I used the bandsaw to cut a 1½-in.
(for more on builders' adhesives, see FHB #65, stock without cutting across the layout line. In long triangular stub tenon on both ends of the
pp. 40-45). 1 used to employ a stick or a rubber this fashion, I finished with a cross section very molding. These tenons would fit into the trian-
spatula for mixing but sometimes ended up with close to that of the factory crown. gular voids behind the factory crown molding,
lumps of powder that wouldn't dissolve. Then I Because the stock laid flat on the saw table, aligning the joints while providing solid backing
discovered a better method: Put the powder in a the cutting operation was accomplished safely for the ends of the factory molding.
plastic container, add the correct amount of wa- and easily. I also kept my fingers far away from The final step was to fit and attach the curved
ter and then shake the container in a circular the sawblade at all times. crown to the wall. Using a belt sander, I relieved
motion. Surprisingly, mixing in this way is faster the concealed edges a bit so that the molding fit
than with a stick and produces a lump-free mix- Scraping it smooth—All that remained was to snugly against the wall and the ceiling. By now
ture every time. smooth out the small, sharp steps on the molding Kreutzfeldt had decided to paint the crown
To make glue up simpler, I taped the pine strips blank. I figured this would be the easy part, but it molding, but just a small amount of belt-sanding
together edge-to-edge on my glue-up table. Then turned out to be the most difficult. produced such a tight fit against the wail and the
I spread the glue across the assembly using a First I tried sanding, but the sandpaper quickly ceiling that no caulking was necessary. I applied
paint roller. That done, I removed the masking became clogged with pine resin. I soon realized construction adhesive along the back and top of
tape, coated the masked areas with glue, tilted I'd have to scrape the pine smooth. the curved crown, then fastened the molding to
the strips upright and pressed them together. To make a scraper, I cut a 45° angle on the end the wall with screws run through the stubs.
Glue up took every clamp I had (including of a scrap piece of the factory molding and As Kreutzfeldt installed the straight runs of
C-clamps, bar clamps and pipe clamps), and it traced its profile on the blade of an old taping crown molding, I was gratified to see that just a
wasn't a pretty sight (photos facing page). 1 in- knife. Then I cut the blade to the layout line bit of sanding produced a satisfying match of
stalled the first clamp at the midpoint of the form a bench grinder and raised a cutting edge by rub- curved to straight molding. E
and worked my way toward both ends, alternat- bing the blade with a hardened-steel punch.
ing clamps above and below the form. Extra Dragging this homemade scraper along the John La Torre, Jr., is a carpenter in Tuolumne,
strips of wood placed against the outer plies of molding at a 45° angle produced satisfying no- Photos by author except where noted.
sugar pine helped distribute the clamping pres-
sure evenly. Excess was srraped off before the A homemade scraper. The author smoothed the sawcuts using a scraper made out of an old tap-
glue cured. ing knife. The edge of the scraper was shaDed with a bench grinder.
After letting the glue cure for 24 hours, I re-
moved the blank from the form. Checking the
concave side of the blank against the template, I
saw that the molding was within '/16 in. of a per-
fect fit. I decided not to shave it further just yet.

Sculpting on the table saw—Once the glue

up was completed, I ran the curved blank, top-
edge-down, through the thickness planer to re-
move slight irregularities from the bottom edge of
the blank. Then I flipped the blank over and
planed its top edge to size. Planing the curved
blank was easy—I simply steered it through the
planer to keep it perpendicular to the cutterhead.
With the sizing completed, I squared both ends
of the blank and traced the outline of the factory
crown molding on one end. Now all I had to do
was remove everything that didn't look like
crown molding.
Probably the easiest way to make crown mold-
ing is to cut it on a shaper. Many cabinet shops
nearby had a shaper, but none had a cutter
 Making Classical Columns
Interior detailing with a router and lathe

by Joseph Beals

lassically detailed columns are commer- for the router-turned shaft, the remaining parts nola, whose Comparison of Orders (written in
cially available in almost any size and config- were turned on a lathe. 1563) is often cited as the rulebook of classi-
uration, but they're expensive, particularly when cal design, an Ionic shaft should be seven mo-
it comes to custom orders. And because the A hybrid design—The columns I designed dules high (not including the necking, which
correct proportioning and form of classical de- combine a clasically correct 24-flute Ionic shaft is typically one more module high), or in this
sign isn't common knowledge, local fabrication with an Attic base, a plain necking and a Ro- case 49 in. To compensate for the relatively
may not be a practical option. In addition, even man Doric capital. This hybrid configuration short hybrid capital, and to keep the length of
a modestly sized column can easily exceed the is quite common; it captures the grace of Ion- the necking within proportional limits, in- I

capacity of most small-shop lathes. ic columnation without the need for a hand- creased the shaft height to 52 in. The design
To make the two load-bearing columns, which carved Ionic capital, a detail that can over- of the capital is quite simple and substantially
completed a room divider between our kitch- power the design, the budget, or both. correct, but is detailed to echo the trim at the
en and living room (photo below), I devised a Classical columns are typically proportioned tops of the adjacent pilasters.
method of turning and fluting column shafts in modules, with one module equal to the bot- According to tradition, the bottom third of
with a shop-built jig that uses a rail-guided tom diameter of the column shaft. The mo- the shaft should be of a constant diameter,
router as the cutting tool. Each column has six dule system is more a guideline than a set of while the top two-thirds taper inward. This
separately made components (drawing facing inviolate proportions. Evety text introduces vari- taper, or entasis, eliminates an optical illusion
page). The base plinth at the bottom and the ations, but there is very little agreement on the that makes a straight shaft appear narrowed in
abacus at the top are square sections. Except details. According to Giacomo Barozzi da Vig- the middle. Because the columns are com-
paratively short, I modified the entasis ratio to
avoid excessive taper, which might create an
awkward appearance. Vignola gives of the
module as the correct diameter for the top of
a shaft, which in this case would have equalled
about 5.83 in. I increased the diameter to elev-
en-twelfths of the module, or 6.42 in., which I
rounded off to in.

Coopering the shafts—Because the shafts

posed the greatest challenge, constructed
I

them first. never considered making solid

I

shafts because they would require a lot of

stock and be heavy, cumbersome and poten-
tially unstable (solid columns have nowhere
to expand but outward in response to increased
humidity). My instinct was to attempt coo-
pered construction, rather than the simpler
but less elegant bricklaid alternative. Commer-
cial columns are almost always coopered, but
with glue-jointed staves tapered so that, once
the shafts are turned, the thickness of the
shaft walls remains constant from top to bot-
tom. Because my columns would be relatively
short, with a minimal taper, I deemed that de-
gree of sophistication impractical.
I chose a straight-staved hexagon as a con-
genial compromise. By using 13/s-in. thick pop-
lar that I had on hand, I could ensure suffi-
cient wall thickness to accommodate a fluted,
tapered shaft without compromising strength.
I considered splining the staves, but the logis-
tics of gluing 12 edges and 6 splines and as-
the fluted columns, combined with raised-panel pedestals, pilasters and an ornamental header, sembling everything quickly and gracefully was
frame the opening between the author's kitchen and living room. The hybrid columns blend an a terrifying prospect. Having decided on a
Ionic shaft with an Attic base, a plain necking and a Roman Doric capital. plain mitered joint, I ripped all the stave stock

60 The Best of Fine Homebuilding Photo: Staff

slightly oversize, set the staves aside for a few plywood turning centers. With a few drops of oil
days, then jointed them and ripped a 60° bevel to lubricate the action, the shaft turned easily,
on each side. Now I was ready for glueup. and the holes didn't enlarge and become slop-
py before the critical flutng operation.
Rounding up the staves—Gluing each shaft A Finally, a pair of 2x guide rails were sawn
was surprisingly simple. cut the staves to
I
planed to match the entasis, or taper of
length, laid six of them across a pair of saw- and
the column shaft (photo 4, p. 63). The inside
horses, and quickly applied yellow carpenter's Capital edge was rabbeted to provide a bearing ledge
glue to the beveled edges. One by one, I stood for a plywood router base; the rabbet con-
them on the floor to assemble the hexagon— tained the router between the rails, but al-
my left hand on the tops of the staves as
placed them together, my right hand reaching
I

Echinus } lowed it full axial movement. The rails were

screwed to the turning centers high enough
for the next piece and setting it in place. above the centerline of the column shafts so
When the hexagon was complete, I tight- that the cutting bit would just clear the arrises,
ened a large stainless-steel hose clamp around but not so high that the fluting bit would need
the top to hold the parts together. Then I used to be extended excessively from the collet.
a combination of hose clamps and a heavy
Jorgensen band clamp (which applies tremen- Turning the shafts—Shalt turning was sim-
dous circumferential pressure) to pull the staves ple but tedious. The router was fitted with a
together down the rest of their length. Work- '/2-in, carbide-tipped, hinge-mortising bit, with
ing my way down the shaft, I drew the staves the initial cutting depth set to hog off as much
tight with the band clamp, secured a hose material as possible without bogging down
clamp beside it, and then moved the band the router. I quickly discovered that it wasn't
clamp to the next gluing position. I repeated practical to make the roughing cuts with the
this operation seven times. shaft driven by the lathe; the router would
I glued the second shaft the next day, and chatter and dance as each arris swept beneath
again the operation went smoothly. Because the bit, proving the machinist's axiom that an
clamp pressure is applied evenly across the interrupted cut is the most difficult of all turn-
six arrises formed by the 12 butting edges of ing operations. Instead, I did the rough cutting
the staves, the effect was to force the staves with one hand guiding the router and the oth-
into perfect alignment. Because the arrises er turning the shaft, using whatever combina-
would be removed in turning the shafts, dam- tion of motions was least tiresome at the mo-
age from the clamps was of no consequence. ment. Wasting the bulk of stock across the
As itturned out, splining the staves would arrises took some time, particularly at the small-
have complicated the preparation consider- er top end of the shaft where the cutting depth
ably and done nothing to improve the results. penetrated well into the flats between arrises.
Only when the shaft was quite round and
A shop-built router jig—My bench-top rout- very close to its final dimensions did I make a
er jig consisted of an "axle" suspended be- finish cut with the shaft driven by the lathe.
tween two turning centers, and a pair of rails This cut was accomplished by advancing the
above the axle that supported and guided the router slowly along the guide rails about five
router (photos through 6, next page). The
1
minutes to traverse the column. The result
axle was a length of ½-in, galvanized iron was close to perfect. Hand-sanding quickly re-
pipe, threaded at one end. With an outside di- moved the few traces of tool marks.
ameter of almost in., the pipe was stiff
enough to support a column shaft without Fluting the shafts—The threaded end of the
flexing. The shafts were attached to the axle axle was fitted with a standard pipe cap through
by hexagonal plywood plates that were let into which I drilled and tapped a 3/s-in. fine-thread
either end of the shaft and screwed to circular hole. That allowed me to bolt a 24-tooth circu-
plywood end-caps. These two-part wheels were lar-saw blade tight to the cap, secured by
drilled through their centers with a 7/8-in, bit a toothed washer under the boll head to
and fitted with a simple wooden clamp that ensure a non-slip fit (photo 3). A simple lever
locked the shafts to the axle (photo 1). The and stud screwed to the plywood turning cen-
turning centers consisted of a pair of plywood ter engaged the gullets on the sawblade, serv-
brackets screwed to the top of my workbench. ing as a simple but very accurate indexing
A few construction details were incorporat- head for cutting the column's 24 flutes. For sa-
ed to make the jig reliable and convenient. fety's sake, the blade should not be mounted
The head end of the axle was fitted with a 12- until just before the fluting operation.
in. dia. pulley aligned with the headstock pul- In half the photographs and drawings I stud-
leys of an old Delta lathe that sits near the end ied, Ionic flutes were of constant width, and
of the bench (photo 2). The pulley was belted the narrow fillets between them tapered toward
to the smallest headstock pulley, and the lathe the smaller diameter at the top of the shaft. In
was—set for its slowest speed. That produced the other photos, the fillets were of constant
about 200 rpm on the column shaft, which width and the fluting was tapered. There is a
turned out to be the upper limit of the router's subtle elegance to ta1jered flutes, but machin-
capacity for a light turning cut. To provide a ing them would require changing the taper of
smooth, relatively non-wearing bearing sur- the guide rails when changing over from turn-
face, I inserted pieces of copper flashing be- ing to fluting. That was more trouble than I
neath the pipe axle where it passed through the was willing to endure. A constant-width flute

Drawing: Bob t.aJ'ointe Finish Carpealry 61

produces a tapered fillet, so I needed to deter-
mine the maximum flute width by defining the
minimum acceptable fillet width at the top of
the shaft.
By referring to more illustrations, drawings
and my own instinct, I chose a strong in. as
the minimum fillet width. That produced a
flute width of about 11/16 in. and a fillet width
at the base of the shaft of a bit more than
in., which was visually ideal. The router was
fitted with a 1-in, carbide-tipped, core-box bit,
which would cut the desired flute width with-
out going too deep. The stops were tacked to
the guide rails to hold the flutes about 1/2 in.
from the tops and bottoms of the shafts. I

clamped the indexing lever into a sawblade

gullet and set the router depth for the cut.
In contrast to the tedium of turning, milling
the flutes was akin to magic. The first flute
gave a hint of classic transformation. As the
flutes began to march around the shaft, not
only was the width perfect, but the shaft itself
changed from a simple turning to a noble ar-
chitectural feature. When the first shaft was
finished, I took it down and put the second
coopered hexagon into its place. Again I suf-
fered the ordeal of turning, and then emerged
from the noise and sawdust of a cobbled-to-
gether workshop device into the stratospheric
realm of classic design. When it was over, the
two shafts had taken on a life of their own.

Capitals and bases—I had enough stock left

over from making the shaft staves to glue up
two short sections for the one-piece column
neckings. Each necking is turned to the same
diameter as the top of the shaft, with an astra-
gal at the base and a slight flare to a shoulder
at the top. The first necking was turned in the
conventional way on my lathe, It then served as
a model for the second, and was duplicated
with the aid of a caliper and a few pencil lines.
It remained to turn the 1-in, high echinus
(the broad flare at the bottom of the capital
that abuts the top of the necking) and the cy-
lindrical portion of the base. In conventional
practice these parts would be made out of sol-
id sections with the long grain horizontal. But
I was concerned with the problem of hiding
alternating long grain and end grain under a
finish coat of enamel. To avoid future regrets,
I made the turnings so the long grain would
be oriented vertically. This would produce a
narrow band of end grain around the cir-
cumference of the torus, but at least it would
be continuous.
laminated seven lengths of 13/s-in. thick
I

poplar to make a cube almost 10 in. on a side.

I crosscut the cube with a bandsaw to pro-
duce a short section for each echinus and a
large section for each base. Because the long
grain is discontinuous across the seven lami-
nations, the sections should be relatively sta-
ble. They were screwed directly to the face- Jtggrng and assembLy. ihe column shafts were turned and fluted with the use 01 a shop-built,
plate of the lathe, and turning was a straight- benchtop router jig. The core of the jig was a '/2-in, galvanized iron pipe that suspended the
forward matter of following the plan sketch. shafts between a pair of plywood turning centers. The hollow shafts were plugged at both ends
with a combination plywood cap and wooden clamp that accommodated the axle and clamped
The bottom of each column base (called the shafts to it (photo 1). For turning, the axle was fitted with a 12-in, pulley belted to an adja-
the plinth) and the top of each capital (called cent shop lathe (photo 2). At the opposite end, a 24-tooth circular-saw blade (mounted back-
the abacus) is a square section. Typically, wards for safety) served as an indexing head (photo 3).

62 The Best of Fine Homebuilding

 these parts also would be made solid with the
long grain horizontal, once again creating prob-
lems with dimensional stability. I turned to
cabinetmaking methods to avoid trouble. Each
plinth and abacus consists of a mitered frame
made of 15/8-ifl, high by P/2-in. wide poplar,
with tongued diagonal corner blocks let into a
groove ploughed around the inside perimeter
of the frames (photo 5, below). The corner
blocks reinforce the glued miter and fill the
gap that would otherwise be exposed at the
abutting turned sections. No end grain is ex-
posed, arid the narrow stock avoids the insta-
bility inherent in a solid block.
Using a shaper, I milled each abacus with a
cove that ends at a shoulder to complete the
top of the capital. The base plinths remain
square, but were made slightly taller than I
really needed to allow for trimming the col-
umns to length later on.

Assembly—The six components of each col-

umn were assembled with hidden screws and
hexagon keys. The plinth is screwed to the
bottom of the turned section of the base
through the corner blocks; the abacus
is screwed down through its corner blocks to
the turned echinus. The remaining assembly
has friction-fit hexagonal keys cut from 3/4-in.
plywood (photo 5). A key screwed to the top
of the base engages the shaft, and a key
screwed to the bottom of the echinus secures
the necking. A pair of hexagonal keys screwed
together align the necking with the top of the
shaft (photo 6); the keys are offset by 300 to
prevent them from falling into the shaft.
The keys center each component to its neigh-
bor. They also ensure that a flute is perfectly
aligned with the middle of the four square
sides of the base plinth and the abacus. This
is a traditional detail that seems trivial unless
it's overlooked. A displacement will be hard to
as a fault, but it gives the and
peculiar appearance of something not quite right.
The columns were primed, sanded, given a
first finish coat and lightly sanded again before
installation. At the site, I cut a measuring stick
to fit at each of the two column locations. Each
column was assembled on the floor and the
base plinth trimmed until the column was a
strong i/it in. taller than the measuring stick. I
used a small hydraulic jack and a post to lift the
header until each column tucked into place.
When the jack was removed, the columns took
enough load to hold everything solidly without
the need for any further fastening.
The columns were given a final finish coat
of oil-base enamel. The result gives a power-
ful, classic appearance to what before had
been an attractive but incomplete design. A
few visitors have noticed the columns at once,
but most have stood back, puzzled at the dra-
matic but hidden change. "This looks great,"
The router was fitted with a square plywood base and guided by a pair of rabbeted rails screw!d one of them said. "What did you do—have ev-
to the plywood turning centers (photo 4). For dimensional stability and to hide the end grain, eiything repainted?" D
base plinths (photo 5) are made of mitered frames reinforced with corner blocks. The
atop each capital is made the same way. To assemble each column, the author screwed the Joseph Beals is a designer and builder who
and abacus to the adjacent turnings, then joined the rest of the components with friction-fit 3/4-ta.
thick hexagonal plywood keys—double offset keys to join the neckings to the shafts (photo 6), lives in Marshfield, Massachusetts. Photos by
and single keys to join the rest of the parts (photo 5). the author except where noted.

Finish Carpentry 63
 Retrofitting a Threshold
A three-piece threshold provides
extra weather protection, especially
in exposed locations

by Gary M. Katz

used to install the ordinary type of metal

Water-return threshold. The drain pan and sill thresholds available at hardware stores. Every
cover of a water-return threshold both act as flashing, time it rained, I'd worry. I'd worry about water
capturing any water that makes its way under the door sweeping in as the door swung open, water
and directing it back outside. An aluminum screened-
trickling in around the sides of the door or water
door jamb that appèarsin accompanying photos has
been omitted from this wing for clarity. entering through the screw holes. I'd worry
about water warping a hardwood floor or stain-
ing a Persian rug.
Now I use ordinary thresholds only in protect-
ed openings. Experience has taught me that
three-piece water-return thresholds are the safest
bet. I often use thresholds made by Pemko Man-
ufacturing (P. 0. Box 3780, Ventura, Calif. 93006;
800-283-9988). A water-return threshold (draw-
ing left) consists of a threshold, a drain pan and
an interlocking sill cover. Although a water-
return threshold is a little tough to install, the
techniques I use make it simple enough, and
the extra effort is worthwhile because it saves
all of that worrying when it rains.
This example involves a door and frame al-
ready in place. While new, prehung-door units
typically come with serviceable thresholds, the
techniques discussed here could be used to add
water-return thresholds to new doors.

Start with the sill cover—Sill covers are life-

savers. They are essentially the flashing for the
Sill cover threshold and cover the rough edge of a con-
crete slab or the exposed grain of a wood floor.
They are also the perfect cure for elevation prob-
lems that can be created when, for example, a
tile floor is laid right up to an original oak
threshold and oak sill. This problem was the
case in the door opening featured here.
I start by deciding how the sill cover should
be notched around the jamb and exterior trim.
Usually the sill cover butts against the door jamb
or the brick mold (drawing left). On this job, a
screened-door jamb had been added, and the
sill cover had to remain behind that jamb so that
the screened door would shut (photo 1).

64 The Best of Fine Hoinebuilding Drawing:

 Notch the sill cover around
the door frames. First, the
broad, flat sill cover is held in
First, I cut the sill cover off square at the longest place to mark the location and
dimension needed, in this case from brick mold depth of the notch (1 and 2) that
will allow the sill cover to fit
to brick mold. To cut the sill cover, I use a small around the jambs of the exterior
circular saw equipped with a metal-cutting door and the screened door.
blade. (For more on cutting aluminum, see side-
bar, p. 67). With the sill cover cut off square, I
tip it into the opening and align it with the back
of the screened-door jamb and with the rabbet
for the main door. Using a pencil or a utility
knife (scratch marks made by a knife are easy to
see on most aluminum products), 1 scribe marks
for the notch (photo 2). I repeat the process for
the opposite side of the jamb. I'm using the jamb
in place of a tape measure and square.

Slope the sill cover to drain—After cutting

the notches, I set the sill cover in place and pre-
pare to trim the front, or vertical, edge of the
cover. On some openings this step isn't neces-
sary. But if there's a concrete porch or wooden
step just beneath the sill of the door, then the
sill cover has to be scribed in. The cover must fit
tight to the original sill, and it must have some
slope so that water will drain outside, not inside.
I tip the sill cover and check the slope with a
torpedo level; between '/s in. and in. of pitch
across the width of the sill cover is usually
enough (photo 3). Using anything handy, I shim
the sill cover in place. Then, on the inside of the
opening, I use my square to measure the dis- The sill cover slopes to the
tance between the sill cover and the floor be- outside. The sill cover must be
neath. I spread my scribes accordingly and canted so that water can drain
scribe a line across the front of the sill cover to the outside (3). Consequently,
(photo 4). Sometimes I attach a clean piece of the front edge must be lowered
by scribing it to the existing
masking tape to the sill cover to make the line threshold, which Is left in place
easier to see. I put on my goggles and earplugs (4). Masking tape makes the
and, holding the sill cover as far from my face as scribe line easier to see.
possible, I cut to the line with my circular saw.

Start with the longest dimension—I start f it-

ting the threshold the same way I fit the sill cov-
er, by measuring the widest dimension of the

Finish Carpentry 65
 door opening, the rabbet for the main door. Af-
ter making the first cut for overall length, I slide
the threshold into the opening to mark the notch
(photo 5).
Normally, the threshold aligns with the face of
the door, but for this opening I wanted to pull
the threshold inside the house 1/4 in. so that it
would cover the raw edge of the tile. I tip the
threshold, hold it against the jamb and mark the
notches (photo 6). 1 repeat these steps for the
opposite end and cut the notches.
Once the threshold is cut, I temporarily set it in
Scribe the threshold to the place on top of the sill cover. I mark the spot
door frame. Rather than tak- where the front edge of the threshold rests on
ing measurements and then the sill cover. This mark will determine the lo-
transferring them to the thresh- cation of the drain pan, which is installed
old, the author holds the thresh- between the threshold and the sill cover. It's im-
old itself against the jamb and
carefully marks the locations of portant to locate the drain pan carefully so that it
notches with either a pencil (5) catches water seeping through weep holes in
or a knife (6). the threshold but at the same time remains hid-
den from view.
I set the drain pan in position just behind the
mark on the sill cover, then scribe a line for the
notch I need to make around the jamb. The thin
drain pan is easy to cut with tin snips.
With the drain pan cut and in place, I set the
threshold on top of it and drill pilot holes for the
screws that hold the assembly to the floor. If I'm
working on a concrete slab, I run my masonry
bit through the threshold, drain pan and sill cov-
er, down into the concrete. That's the surest
method I know of getting concrete anchors in
the right spots.

Cuffing off the door—With the threshold and

the sill cover in position, I'm ready to determine
how much to cut off the bottom of the door. In
order to get a weather-tight seal, this type of
threshold requires a separate U-shaped metal
door bottom with a vinyl sweep and drip edge
(drawing p. 64); in this case I used one made
by Pemko Manufacturing. In some installations,
there is enough room for the metal door bottom
between the bottom of the door and the new
Door height is transferred threshold. In this case, though, the door is too
from the jamb to the door it- close to the threshold and has to be trimmed
self. Once the sill cover and slightly. First, I measure up from the top of the
threshold have been notched threshold ½ in. and make marks on both jamb
and set in place, the height at legs (photo 7). Then I remove the threshold and
which the bottom of the door
will be cut off can be deter- drain pan, but I leave the sill cover. The cover
mined (7). The mark made on provides a smooth surface for me to run my
the jamb takes into account the scribes along.
thickness of the metal.door bot- I spread my scribes from the sill cover up to
tom (8). The author then scribes the line I've made on the jamb (photo 8), then
the bottom of the door (9) to the shut the door and scribe a line across the bot-
new sill cover. Masking tape on
the bottom of the door makes tom of the door (photo 9). Again, masking tape
the pencil line easier to see. makes it easier to see the line.
I use different methods for cutting off doors.
On veneered doors I sometimes use a "shooting

66 The Best of Fine Homebuilding

stick" straightedge made of thin plywood (photo
10). The shooting stick allows me to cut doors
quickly without worrying about tearout. If I don't
have my shooting stick with me, I use a metal
straightedge and a knife to score a line on the
door before I cut it. Either way, it's my circular
saw that does the hard work.

Seal the threshold with plenty of silicone—

Before I install anything, I sweep out all of the
dust and dirt, especially under the sill cover. I
run a bead of silicone under the sill cover to
help secure it, though the screws that pass
through the threshold really do that job. I press A site-built tool speeds the
the sill cover into the silicone and then run an- cut. The author cuts doors with
other bead of silicone on top of the sill cover a "shooting stick," a plywood
and beneath the drain pan. I take care to keep straightedge that has a fence
the silicone away from the front edge of the against which he registers the
table of his circular saw (10).
drain pan so that it doesn't squeeze out. Then I
press the drain pan down into the silicone and
apply more silicone at the joint of the jamb and
drain pan. I also squirt silicone into the screw
holes. Finally, I set the threshold and screw it
down snug.
It's better to notch the metal door bottom First the weatherstripping,
around the weatherstripping on the door jamb, then the door bottom. Weath-
erstripping is installed on the
so I install the weatherstripping first if there isn't
side jambs first. The drip cap on
any already. The door bottom has to be cut to fit the door bottom then is scribed
the overall width of the door and then notched and notched to fit around the
to fit around the weatherstripping on the door weatherstripping (11 and 12).
jambs (photos 11, 12).

Not too tight, not too loose—After cutting the

door bottom, I slip it on again and swing the
door shut to make sure everything fits. The door
bottom shouldn't be too long and squeeze or
rub against the weatherstripping, but it should
come close. I press the door bottom down
against the threshold, but not too hard. The vinyl
sweep needs to contact the threshold, but
shouldn't be forced against it. Otherwise, the
sweep will compress over time, and the seal will
be lost.
I drive one screw at each end of the door bot-
Take care when cutting aluminum
Cutting aluminum thresholds is not my makes the 4200N, although some
tom to hold it in place. Then I check the action idea of fun, so I like to do it as quickly and distributors still have a few of these saws.
of the door to be sure that it's sealing but that it's as safely as possible. Many people use It was replaced with the model 5005 trim
not rubbing too hard. Then I drive in the rest of hacksaws. Some professional saw, which has a larger and slower-turning
the screws. weatherstrippers use portable table saws 5½-in, blade.)
All that's left then is to apply silicone to the with aluminum-cutting blades. Wearing eye protection is important
threshold and sill cover at the joint of the jamb, I use a 4%-in. Makita model 4200N trim when cuffing aluminum, not only to guard
and maybe a little caulking between the door saw (Makita U. S. A., 14930 Northam St., La against bits of flying metal but also
bottom and the door to seal out moisture. Be- Mirada, Calif. 90638; 800-462.5482) that because teeth on the combination blade
fore I leave, I check the swing of the door one cuts at a fast 11,000 rpm and a fine- can chip. In my work box I carry plastic
more time. When it's right, the door closes just toothed, combination metal blade, also by goggles wrapped in a sock to prevent
like a refrigerator, and with that whoosh of air, Maldta (model 792334-2). I've used scratches. Years ago a bungee-cord
I'm gone. carbide-tipped blades to cut aluminum, but accident took most of the vision in my left
they are expensive, especially when the eye, so I'm careful with my right one. I
Gary M. Katz is a carpenter/contractor and writer teeth begin to break off. (Makita no longer also wear earplugs.—G. M. K.
in Encino, California. Photos by the author.

Finish Carpentry 67
More Than One Way
to Case a Window
You can vary the look with simple combinations of flat boards
and stock moldings

by Joseph Beals III

tock trim, such as 1x4 pine or clamshell cas-

Two approaches to formal casing. With mitered casing the ing, will always be useful in routine, contempo-
same trim runs up one side, across the top and down the Butted rary construction. But after years of installing
casing entails a classical pediment sitting on top of square-cut side casings. these lifeless, standard-issue casings, frustration
drove me to cross the frontier. Using as a refer-
Mitered casing Butted casing ence the casings I'd seen in so many period New
- Cornice England houses, I built a simple pediment head
Mitered casing to surmount square-edge, square-cut side
return casings. I will not forget that first look back at the
result: An ordinary window suddenly had char-
acter, grace and purpose; and a client new to
classical trim was particularly pleased with a win-
Pediment dow worth looking at, not just through.

Mitered casings get moldings; butted cas-

ings get pediments—In this article, I'll discuss
two approaches to formal casing design (draw-
ing left): mitered casings as a stylistic option and
butted casings, which employ a pediment head,
or architrave, as a classical architectural option.
In general, mitered casings are developed by
Plain side adding layers of moldings to the perimeter of a
casing mitered flat casing. With butted casings, the ped-
iment head sits on the square-cut tops of plain
or molded side casings, and architectural detail
is developed on the pediment itself.
In a mitered casing the simplest alternative is
the use of two or more layers of molding. A bead
cut on the inside edge of a casing and a back-
band applied around the outside perimeter of a
flat casing will give a strong, three-dimensional
appearance (top right photo, facing page). A thin
molding interposed between the flat casing and
the backband adds another element of detail
Stool cap and shadowline to the profile (bottom right pho-
to, facing page).
A formal alternative to mitered casings is the
use of a pediment, or architrave, as a head cas-
ing with side casings butting into it. The pedi-
ment represents an entablature, the lower por-
tion of a classical roof, and its elements are
Horn derived from ancient Greek and Roman temples.
This style of window trim is commonplace in pe-
riod architecture, especially in Federal and
Greek Revival houses of the 19th century.

Apron All window casings start with the stool

cap—Any style of window trim must begin with
the stool cap, the piece that finishes the inside

68 The Best of Fine Ho,nebuilding Drawings Dan Thornton

 Six variations on a theme
We asked designer/builder Joseph Beats
to show us how to create various styles of
window trim without resorting to custom
moldings made with a shaper. On a test
wall in his shop, Beals mocked up six
casing designs, using mostly dimension
lumber and stock moldings from the local
lumberyard. He cheated occasionally and
used his shaper, but where he did, a
router would work just as well.

Mitered casing with

backband and a beaded inside edge
Starting with 4½-in, wide flat casings, the author mills a
simple bead on the inside edge and runs a backband
molding around the perimeter.

Intermediate
molding

The stool must stay level and perpeiulic-

ular. The author screws blocks of scrap wood
to the wall to hold the stool temporarily.

edge of the windowsill. You can buy stock stool

cap at a lumberyard, but you'll spend enough
time fitting it to the window that you might as
well make your own.
I like the stool to stand proud of the casings
about an inch. If you're thinking about a two-
piece or three-piece mitered casing, you'll have
to figure out the combined thicknesses of the
moldings before you cut the stool.
I cut the stool stock long, and I machine a bull-
nose, or half-round, on the outside edge. I scribe
and cut the two horns (horns are the ends of the
stool that overlap the drywall on each side),
check the fit, then cut the ends to allow about a
1-in, overhang past the outside edges of the side
casings. The bulinose returns can be shaped by
machine, but I prefer to remove the bulk of the
waste with a block plane and finish the job with a
sharp file or a piece of cloth-backed sandpaper.
Given the wide variety of window conditions, A third layer adds depth and detail
Interposed between the flat casihg and the backband, a
fastening the stool can be a problem. On the thinner (½ in. or less) molding adds another
Andersen window used in these photos, the stool level of detail and shadowline.
is toenailed from the top into the sill. The horns
can be face-nailed into the framing studs if nec-

Finish Carpentry 69
 Connect the marks with a straightedge. Using a straightedge to
mark the side-casing top cuts will ensure that the head casing fits perfectly.
Simply squaring cuts from the marks can result in sloppy joints.

Mark the side c with a gauge block. The autnor uses a scrap of Biscuits register side casings to stool. Common practice Is to nail up
wood, marked with the amount of offset between the window jamb and through the stool into side casings. But biscuits hold better than end-grain
trim, to mark the length of the side casings. nails and won't come out through the exposed surface of the side casing.

essaty. Glue or caulk between the stool and the mark on the casing to remind me of the angle of finished; they have been locked in place so that
sill can help hold everything together. the cut. This procedure may seem foolish, but it they can't move.
To keep the stool square to the window, I is remarkably easy to forget how things go to- Other people prefer to install mitered casings
screw 2x4 blocks to the rough framing right be- gether in the short walk to the miter saw. by working around the window: up one side,
low the stool (photo left, p. 69). Nothing is worse When I mark the two sides of a butted casing, I across the top and down the other side. I prefer
than getting a window trimmed and realizing that make tick marks on the inside edges with the my method because the two sides are done
during the course of your work, you've pushed gauge block and then connect the two marks quickly and easily, and any adjustments are
the stool out of square with the wall. The blocks with a straightedge (photo above right) and a made in the head casing. Only if the window is
hold the stool square until I add the apron at the sharp pencil. In theory, the top cuts should be badly racked would this be impractical, in which
end of the job. square to the casings, but this theory may not be event it might be better to fix the window condi-
true in practice, especially in the case of an old tion first.
Cut the side casings long, fit them to the window that you're retrimming. Before nailing
stool and mark their length—For any casing the side casings, I cut biscuit slots in the bottom Fitting a mitered head casing— I like to cut a
design, the two side casings are cut long and fit- ends and mating slots on the top of the two stool bead on the inside edge of all of my mitered win-
ted first to the stool. I cut the casing bottoms horns (photo bottom right). dow casings. A shaper, router or cutter head-
square, then hold a casing in position against the equipped table saw cuts beads equally well. The
window jamb. I judge the reveal by eye (the re- Nail the side casings home before you fit molded bead mimics the applied sash stop
veal is the exposed portion of the jamb between the heads—Some people prefer to fit the head found on a lot of old double-hung windows.
the casing and the jamb's inside edge), but if you casings while the two side casings are only For a mitered head casing, I make the 45° cut
prefer reference marks, make them with a gauge tacked in position on the theory that it's easier to on one end and then mark the other end. As the
block of some kind. make the joints perfect if both pieces are ad- photograph (left photo, facing page) shows, the
In theory, if the bottoms of the side casings are justable. Then, after everything is fitted, you nail casing is held upside down with its mitered end
cut square and if the stool is perpendicular to the pieces home in one marathon effort. I think perfectly registered on the side casing, and the
the sides of the window, the casings should fit this process invites problems because the side other end is marked with a utility knife.
tight to the stool, but theory doesn't always work. casings can shift as they are pulled against the I cut the second miter, saving the knife line. I
I close any gaps by dressing the casing bottom wall when the nails are driven home. Even set- drop the head casing in place, taking care to
with a sharp block plane. After both side casings ting a nail after it has been hammered in can keep it parallel with the window head jamb. Any
have been fitted to the stool, I tack them in posi- cause the casing to shift, and the joint you deficiencies in fit are apparent, and these flaws
tion with a few 4d finish nails. thought was perfect opens up again. Adjustments are dressed out with a sharp block plane.
To mark for the top cuts, I use a gauge block as to bad joints are awkward or impossible. I use biscuits to reinforce the mitered joints.
shown in the photograph (above left). For After I fit a side casing to the stool, mark it, cut it The biscuit slots in the side-casing miters can eas-
mitered casings, the mark I make indicates the to length and cut the biscuit slot, I nail it tight ily be made in place if they weren't done earlier,
bottom of the miter. I draw a diagonal reference and set all the nails. The side casings now are and the head-casing miters can be slotted on any

70 The Best of Fine Homebuilding

For mitered casings nail home the two Simple butted casings
Plain 4½-in, wide side casings are topped with a 4½-in, wide
sides and then fit the head. The author head casing as a pediment. The only embellishment comes
makes a 450 cut on one end of the head casing, in the fact that the pediment is made from thicker stock (5/4)
holds it upside down on the side casings and and overhangs the side casings in front and on the ends.
marks for the other cut with a utility knife.

flat surface. If allis well, I glue the slots, slide the

biscuits in and tap the head casing into place.

More layers add depth and detail to

mitered casings—As the photographs of f in-
ished casing styles show (see p. 69), mitered cas-
ings can be built up in layers to give a handsome
appearance. The tight, solid base offered by the
flat casings makes the application of additional
components easy.
There are plenty of backbands available at lum-
beiyards, or you can make your own with a table
saw and router. I install the backband in a se-
quence similar to the flat casings: fitting and nail-
ing both sides and then fitting and nailing the
head piece.
A three-part casing with an intermediate mold-
ing between the backband and flat casing adds
another level of detail to the casing. I make the
intermediate molding from 1/2-in. or thinner stock.
Making this piece from thicker stock can result in
a clunky, heavy casing. A thinner intermediate
molding has a more delicate look.
On these multilayered casings, the outside
edges may need dressing to present a single, flat
surface. I do this with a sharp bench plane (pho-
to left, p. 72), working as close as I can to the
stool horns, and I clean up the last few inches Bed molding above and a
with a sharp chisel. bead below dress up the pediment
A slightly wider head casing a nailing base for
standard bed molding at the top of the pediment. Builnosed
Pediments dress up butted casings—One of on one edge, a thinner molding in. by in.) overhangs
the beauties of adding a pediment to a butted the bottom of the head casing.
casing—aside from the aesthetic ones—is that all
ii
Finish Carpentry 71
 Run mitered returns long; trim them
with a handsaw. The author uses one hand
to hold the handsaw while he makes the cut.

Plane the euges even. It's possible that the outside edges of all of the built-up moldings of a Nail the head moldings on a flat surface.
mitered casing won't be perfectly flush after they're nailed on. A couple of passes with a bench plane To align the pediment's components, the author
makes a flat surface. The last few inches at the bottom are cleaned up with a sharp chisel. nails the pieces together on a flat surface.

of the pediments can be made on a bench and won't make much difference in the final result. I each end, as the bead does on the pediment de-
added as a complete unit to the tops of the two use a handsaw to trim each return flush with the scribed above (photo top right, p. 71). Adding a
side casings. backside of the head (photo above right). bed or cornice molding to the 5/4 stock begins to
A basic pediment is built from 5/4 stock, with To make the bead, I resaw a piece of stock echo the lines of a true entablature.
an applied bead across the bottom and a bed 7/lrin. thick by I V8 in. wide, but these dimensions The next evolution is a 6/4 head with a bed
molding across the top (bottom photo, p. 71). are not critical. I cut a bullnose along one edge molding, surmounting a square-cut 5/4 fascia or
The space between the top of the bead and the with my shaper (also an easy router job). The lower molding (photo top right, facing page).
bottom of the bed molding should be at least bead overhangs the casings by about in. at Each element overhangs the face and ends of
equal to the width of the side casings. You can each end. I shape the returns with a block plane, the element below, in the same style as the clas-
make the space wider for a bolder appearance, then finish the radius with a sharp file or a strip of sical Greek and Roman entablature.
but beware of overdoing it. Make a trial pedi- cloth-backed sandpaper. The bead is applied to Adding a cornice molding above the bed mold-
ment if there is doubt about the aesthetic effect. the head stock with glue and 4d finish nails. You ing in any pediment completes the basic ele-
To provide a base for nailing on the bed mold- can do the application free hand, but working ments for the full entablature (bottom photo, fac-
ing, I make the height of the head stock at least on a flat surface like a table saw makes it easy to ing page). For interior cornices, I use 4/4 or 5/4
1/2 in. taller than the width of the side casings. The keep the back of the bead flush with the back of stock, and I machine the curve on my shaper
length of the head stock is equal to the distance the head stock (photo bottom right). with a knife I ground for the purpose. I shape the
between the outside edges of the side casings. Before installing the pediment, I cut biscuit returns by hand, as described above for stool
I draw a pencil line along the length of the slots on each side of the bottom, aligned with and bead returns, because machine work on
head stock to indicate the bottom edge of the the centerlines of the two side casings. The slots short, end-grain sections is usually awkward.
bed molding. With a square, I mark the bottom in the side casings can be made before the cas- If you choose to include a cornice, you should
edge of each bed-molding return around the cor- ings are installed, but if you have enough ceiling increase the height of the head stock by the full
ner at each end of the head. This step is impor- height, it's just as easy to make them in place. Be height of the bed or cornice molding and install
tant because it's easy to cock a short return, and aware of where you nail the bead onto the head the cornice molding first. Remember to make
even a small misalignment will be brutally obvi- stock so as not to put nails where they will inter- the cornice long enough to incorporate the bed
ous once the pediment is installed. fere with cutting biscuit slots. molding returns. Install the bed molding tight to
The long piece of bed molding is mitered at the cornice, and take care to seat it properly. It's
each end. For convenience and safety, I cut the Pediments can range from simple to com- easy to cock the long molding in rotation, which
mitered returns long. I check the miters for a tight plex—As the photographs of casing styles show will make fitting the mitered returns frustrating.
fit, then fasten the returns with glue and a few (see facing page), the pediment can be varied
brads. To avoid nail holes, you can use masking to increase the level of architectural detail. The Aprons complete the casings—An apron is
tape to secure the returns while the glue dries. simplest pediment is a plain piece of 5/4 stock, nailed to the wall below the stool and is the low-
But this task is a slippery, three-handed job, and it cut to a length that overhangs the side casings at est component of a window casing. On a lot of

72 The Best of Fine Homebuilding

Cope the apron returns. Rather than miter
returns, the author draws an outline on the
molding profile and cuts it with a coping saw.

Different
aprons fit dif-
ferent casing
styles. The au-
thor adds the
apron after the
rest of the win-
dow is cased, ex-
perimenting
with the apron
design to find
one he thinks
works best with
a particular win- Layers of successively
dow-casing style. thicker stock look elegar It
Here, the 3%-in. side casings supp a pediment composed of
5/4 square stock and 6/4 head cas Fng, again capped with a
bed molding. Each element overt ngs the face and end of
the element below.

windows, a piece of casing stock serves as the

apron, with its ends cut square, angled or re-
turned, according to the whim of the carpenter
or designer. There is no equivalent to the apron
in classical architecture, which is why aprons in
neoclassical window trim exhibit such a variety
of profiles.
The apron has a specific aesthetic function,
and there are several profiles that I use (bottom
photo above). The apron visually returns the win-
dow to the plane of the wall below. About 3Y2 in.
of height works well; if the apron stock is wider, it
will look boxy and unbalanced, but narrower
sections can be successful. I mill the profiles on
the shaper and table saw, but a router will also
handle this job easily. The simplest pattern
echoes the backband that is used on the mitered
casings. I make this piece out of solid stock, but a
cove cut with an applied bead on the bottom is a
simple alternative.
The apron returns are coped rather than
mitered. I sketch the profile free hand, but you
might prefer to use a paper or flexible plastic tem-
plate. I remove the waste with a coping saw (top
photo above), and I dress the return with a sharp
file. This work goes quickly in contrast to
mitered, glued returns, which are awkward and
generally are not worth the effort, even on bright
finished trim. E A cornice is the touch
Adding a simple cornice piece the bed molding and
head casing completes the basic e ements for a formal
Joseph Beals Ill is a designer and builder in entablature.
Marshfield Hills, Mass. Photos by Jefferson Kolle.

Finish Carpentry 73
 Making Curved Casing
Strip-laminating arches to match straight casing profiles

by Jonathan F. Shafer

few years ago I was asked to step in Making the patterns—Our first step was to to contain the unknown radius points. We
and complete the finish work on an 11,500-sq. make patterns of all the arched windows and then used a beam compass to find the radius
ft. Tudor home that had taken 18 months to doors. How we produced the patterns varied points on the plywood by trial and error. Again,
get through drywall. Completing the trim took depending on the particular application—some the arch was then transferred to kraft paper. I
an additional 12 months and posed many chal- methods were as simple as tracing on kraft came on the project too late to have done it,
lenges, such as hanging 8-ft. high doors, build- paper (available in long rolls) against the win- but in the future I would make tracings of
ing four stairways and running thousands of dow frame, while others were as involved as each window frame prior to installation.
feet of wide casings and base. The house also mathematically computing arcs and multiple Finally, we cut out each pattern and
had many arch-top windows (photo below) radius points. checked it against the corresponding window,
and doorways of various heights and widths. One method we used on some of the more making necessary adjustments. The patterns
The curved window and door casings had to complex windows required a thin, flexible rip- also had to be extended on both ends to allow
match the existing straight casing, so I decided ping of even-grained wood long enough to fol- extra casing length for trimming later. We la-
to produce the curved casings on site with the low the arch along a window frame. This strip beled the patterns for window location and
help of a talented crew of finish carpenters. was clamped or held by helpers against the wood species.
My approach to this challenge was to strip- inside of the frame. We maintained the arch
laminate the arched casings. By alternating shape by tacking crosspieces to the bowed Building the bending forms—When the pat-
strips from two pieces of straight, even-grained strip. The more crosspieces we used, the bet- terns were ready, we built a bending form for
casing, we reproduced the casing profile. We ter the shape was held after the clamps were each one. We constructed them from 2x stock
ripped the strips from straight casing and then removed. We then transferred the shape of the cut into arcs on a bandsaw (for the design of
bent them around a form for each window and arch to kraft paper. a bending form for curved jambs, see the sidebar
door. We also laminated extension jambs for With another method, we tacked plywood on p. 77). With roundtop casings, the 2x
each window, using the same bending forms. against the window, using a piece wide enough arcs were made using a simple circle-cutting

The curved casing for these windows was fabricated on site, using two pieces of straight stock cut into narrow strips and laminated around a form.
jig fixed to the bandsaw table (drawing be- ripped with the inside edge against the fence, next stage. Once the complete casing was dry,
low). We extended the table with a piece of and so on. we scraped and sanded the casing profile, re-
'/4-in, plywood and ran a screw through it to On the first piece of casing that we did, I moving glue squeeze-out and any irregulari-
create a pivot point. The 2x stock was then made the mistake of dropping the laminating ties in the profile.
pivoted around the pivot point on a '/4-in. ply- strips into a pile, thinking I could easily sort
wood carriage. them out later. Wrong. It took me over an hour Making extension jambs—We made the ex-
To cut the more gradual arcs of the bigger to put the pieces in order before gluing them tension jambs for the windows with the same
windows, we used a 1x3 to extend the pivot up. For subsequent casing, I built a rack to bending form used for the casing—all we had
point of the circle-cutting jig across the shop hold the strips in order (top photo, next page). to do was remove the '/4-in, shims. The reveal
(photo below). The 2x arcs were screwed to a The rack was simply a pair of lx4s with saw was 1/4 in. so we made the extension jambs
plywood base or to the subfloor, depending kerfs in them, nailed to a short bench. in. thick, allowing sufficient material to secure
on how big they were. the casing. We produced strips roughly 1/8 in.
Glue up—Before we could start gluing, the thick to reduce the chance of springback and
Ripping strips—Once the bending forms were bending forms had to be adjusted to allow for wide enough to fill the space between the
completed, the strip-cutting operation was next. the jamb reveal—the difference between the window frame and the edge of the drywall,
The basic principle here is that you're taking a inside edge of the jamb and the inside edge of plus 1/4 in. extra for ripping and jointing to the
piece of straight casing with the molded pro- the casing. We tacked '/4-in, thick spacing finished width after the glue had dried.
file you want and ripping it into narrow strips shims (the width of the reveal) against the Ripping and jointing was a two-man oper-
that you can bend around a form and glue forms. We also covered the forms with waxed ation. One man fed the piece into the table
back together. But if you were to do this by paper to prevent the casing from adhering to saw or jointer, and the other helped support
ripping a single piece of casing, the resulting them during glue up. the piece as it went into and came out of each
molding would be narrower than the original We had plenty of pre-adjusted clamps on machine. Cutting and fitting the extension
because of the material lost to the saw kerf. hand to do the job—everything from bar clamps jambs to proper length was a trial-and-error
Therefore, you have to make alternate cuts on to wedges against wood blocks screwed to the process, the error always being on the long
two pieces of straight casing. floor. Our clamping cauls were bandsawn to side until the jambs fit. Next they were glued
To ensure that the laminating strips were match the outside radius of the casing. We and nailed to the window frames through pre-
cut to a uniform width, we used thin pieces of dry-fit the strips around the form so that we drilled holes.
pine as spacers resting against a preset table- could work out a clamping strategy (bottom
saw fence. This enabled us to cut the casing photo, next page). During glue up, we quickly Fitting the casings—The ease of fitting the
incrementally without changing the position and evenly brushed yellow glue on each piece. casings to the windows was directly related to
of the saw fence. Because the casing was relatively wide and the care with which the pattern had been
In our case, the saw blade, and hence the the set-up time relatively short, we glued and made. If the pattern was true to the window
laminating strip, was roughly 1/8-in. wide. The clamped the strips in three stages and let the form, the casing was relatively true to the win-
spacers were cut so that each was twice the glue dry overnight before proceeding to the dow. Because our laminating strips were about
width of the table-saw blade. We cut our spac-
ers the same length as the short auxiliary
fence on my table saw. To keep them from
slipping with the casing as it was being cut we Bandsaw circle-cutting jig
simply tacked a brad to the underside of each
spacer, which hooked over the front edge of
the saw table (drawing next page).
After the spacers were completed, we ad-
justed both pieces of casing (ripped a little off
them) so that the finished width was an even
number multiple of the spacers. Our casing
had a rabbeted back band around the outside
edge, so we were able to reduce slightly the
outside edge of the casing without changing
the profile. (The side casings were also adjust-
ed in width to make them equal to the arched
head piece.)
Next, we glued and clamped the back band
to the casing. We also filled in the plowed re-
lief on the back of the casing by gluing in thin
material and jointing it flush. This was neces-
sary so that each strip would be cut square to
the others.
We set the table-saw fence to equal the total
adjusted casing width (casing plus back band)
minus the width of one spacer. Finally, we
equipped our table saw with a riving knife (or
splitter) mounted behind the blade. This pro-
tected the thin strips from damage as they For round-top windows with
came off of the saw. Shafer cut sections of the bending forms
In order to produce alternating strips from
on the simple jig shown in the drawing.
For bigger windows, he attached a length
two pieces of straight casings, we ripped the of lx3 to the plywood carriage (above) and
first piece of casing on the saw with the out- extended it across the room to a center
side edge against the fence (drawing next point on top of a workbench.
page). Next, the second piece of casing was

Drawings: Bob LaPointe Finish Carpentry 75

 in. wide, the springback was negligible. We dow stools. Likewise, using plinth blocks responding angle was then cut on the side
were using relatively wide casings, so spring- would have simplified fitting the casing on the pieces, After the miter was judged to be tight,
ing the casing to match the window—any- bigger windows and doors. But we decided to the bottoms of the side pieces were marked
where the pattern was not true—was very diffi- miter the corners between the arched casing and cut square to rest on the window stool, or
cult, if not impossible. We had to live with and the side casings. mitered if the window was picture-framed. E
compromises in a few places. We calculated the miter by tracing the head
Just as with the extension jambs, fitting the casing and side casings right on the drywall, Jonathan F. Shafer was a carpenter in Dublin,
casing was a trial-and-error process. It was rel- then connecting the points where their inside Ohio, when he worked on this project. He has
atively simple on the windows with one-piece edges and outside edges intersected. The head since relocated to Bellingham, Washington.
casings that were butted directly to the win- casing was cut to match this line, and the cor- Photos by Kevin Ireton.

Cutting sequence
Laminating strips for arched casing are
cut from two pieces of straight casing.
With the fence locked in place, the first
piece is cut with its inside edge
toward the blade. Then the second
piece is cut with its outside edge toward
the blade. With each successive pair of
cuts, a spacer is added to the fence.

No
strip I_Blade

I
Piece 1

lstcut
Backband glued on.
Relief in molding filled in.

strip Blade
1 o ensure mat the laminating strips were cut to a uniform kept me LaDle-saw fence
at a fixed distance from the blade and used spacer strips to move the straight stock incremental-
ly closer to the blade. The rack beside the saw holds the laminating strips in their proper order
for gluing.
2ndcut —- The bending form is simply 2x stock cut into curved sections and screwed to a piece of plywood.
Here the laminated strips have been clamped up without glue to work out the clamping strategy
and eliminate some of the usual glue-up trauma.

Blade
strip 1 spacer

Piece

3rd cut

Spacer detail
Each strip is twice the
width of the blade.

Brad in end of spacer "F ence

holds it in place.

76 The Best of Fine Homebuilding

 edge-jointing later, and a few inches longer each side of the jig, Paul and Mullis worked
Reusable jig for than arc length. Paul and Mijllis first sorted away from the center, clamping to each block
curved jambs among the four plies and selected the best with just enough pressure to hold everything
for the inside finished surface. They finish- in place. With the frame now held stable in
With the trend toward New Classicism In sanded this face—a job that is much easier the jig, Paul and Mullis went back and added
architecture has come a renewed popularity on a flat bench before glue up. They laid that two C-clamps, with a caul above and below,
of arch-top windows and, to a lesser extent, ply face-down on the bench and on its back between each of the clamping blocks
doors. Making the jambs for these doors and they drizzled aliphatic glue, each man (drawing below). With all the clamps in
windows presents only minor challenges. releasing his artistic talents in a distinctive place, they tightened them firmly, and glue
The millwork is straightforward, a matter of pattern of squiggles. They spread the glue to flowed from every seam. They have found
laminating a stack of plies in a curved jig. an even film, paying special attention to that only an even clamping pressure will give
The problem is cost. One of the axioms in the coating the surface along the edges. Then, a fair arc. The key is to use lots of clamps
millworking business is that no matter how they lifted the second ply onto the first and and tighten them uniformly. The other men in
many stock jigs you have cluttering up the repeated the procedure until all four plies the shop joked about Paul and Mullis
back room, the next customer will want a were glued. They clamped the plies by the hatching another peacock (photo below), and
window of a slightly different size. Building a edges to keep them together as a unit until in fact, the radiating bar clamps did call to
custom jig for each order soon prices the clamped in the jig. mind one of those strutting birds.
work out of the market. I returned the next morning to watch the
Over the last few years, David Marsaudon Clamping—Working quickly before the glue plucking of the bird. Clamps were removed,
and the other folks at San Juan Wood Design set, Paul and Mullis lifted the group of plies scraped free of dried glue, and hung back in
have developed and refined a jig for making onto the jig and balanced it on the apex. their racks. With the bench clear, Paul and
arch-top jambs that employs reusable parts. Starting at the center, they laid a caul across Mullis laid the finished frame against the
With this jig, they can lay up fair, smooth the width of the plies, centered over a penciled layout marks. The frame had sprung
arcs of virtually any radius, and do so at a clamping block. They hooked one bar clamp to within in. of the design width.
cost that gives them a competitive edge in under the chin of the block, snugged it down —J. Azevedo, a free-lance technical
the custom market. Their jig is made with and then did the same with another clamp on writer in Friday Harbor, Wash. Photo by
some scrap plywood, strips of '/4-in. the opposite side. Working as a team, one on the author.
hardboard, a sheet of particleboard and a
shop full of clamps. I visited San Juan's shop
in Friday Harbor, Washington, and got a
close look at how the jig works. Clamping-jig detail
/ Bar damp

Building the jig—The jig, which somewhat

resembles a dinosaur skeleton, Is made of
three parts: a semicircular particleboard
panel, a set of clamping blocks and a mold
surface. The radius of the particleboard
panel determines the final size of the jamb.
I watched craftsmen Roger Paul and Jerry
Mullis make the frame. They started with a
given finished radius (the inside of the
finished window frame) of 73 in. Based on
previous trials, they estimated springback to
be 2'/s in., so the jig had to have a radius of DIOCK /
70'/z in. By subtracting the thickness of the
mold surface ('/4 in.) and the depth of the /4 Ii
clamping blocks (3 in.), they came up with a pa tlcleboara
Piles for'
jamb
panel radius of 67'/4 in. A new panel must be
cut specifically for each size window frame,
but smaller panels can be cut from larger
ones to save on materials.
Paul and Mull is propped this panel
vertically on a bench, and screwed to its
rounded top a set of clamping blocks at
about 8 in. o. c. Made from scraps of 3/4-in.
plywood laminated in pairs, these blocks are
saved for reuse in each new jig. Finally, Paul
and Mullis stapled a skin of '/4-in. hardboard,
as wide as the intended window jamb, to the
clamping blocks. They started at one end and
carefully worked their way to the other,
keeping the skin centered on the blocks and
checking the final jig for bowing. Because
any imperfections in the skin would
telegraph through to the window jamb during
glue up, Paul and Mullis checked that the
skin was perfectly smooth. As a final touch,
they waxed the skin with paraffin and
smoothed the wax with steel wool. The
wax keeps glue from sticking to the skin,
so the finished jamb will pop easily out of
the jig.

Assembling the plies—Curved jambs are

built from four plies, each in. thick, about
1/4 in. wider than finished width to allow for

Finish Carpentry 77
 Raised Paneling Made Easy
A combination of built-up moldings and layers of plywood
duplicates a traditional look

by Jim Donnelly

lenn Bostock, a cabinetmaker living in

Pipersville, Pennsylvania, was contracted to fab-
ricate a frame-and-panel wall in the entryway of a
house in New Hope, Pennsylvania. A formida-
ble task, but certainly not an insurmountable
one. There was, however, a challenge: The pan-
eled wall was to be built on the inside curve of a
circular staircase (photo facing page).
Traditionally, curved frame-and-panel walls
were made by sawing, shaping and planing solid
wood. Even with a fully equipped modem shop,
the stiles, rails and panels would require elabo-
rate jigs and extensive shaper work.
Bostock came up with a technique for dupli-
cating the appearance of traditional frame-and-
panel construction that was less expensive and
quicker than traditional methods. His technique
employs a series of overlapping layers, includ-
ing bending plywood, tempered hardboard and
kerfed pine. Each layer was nailed or glued in
place, and the whole assembly then was
trimmed out with moldings that hid the seams
and fasteners of earlier layers.
Bostock's process was simple, and the most dif-
ficult task was making accurate cuts so the mold-
ing fit together well. Granted, the construction
of the wall took a lot of time, but not nearly as
much time as a solid-wood raised-panel wall
would have required.
Although Bostock's project was a curved stair-
case wall, his techniques would work just as well
for building a flat raised-panel wall. An added
benefit for those with limited access to a shop is
that Bostock's techniques can be accomplished
almost entirely on site.

Two layers of bending plywood provide a

firm foundation—Bostock began by attaching
two layers of bending plywood directly to
the studs. Bending plywood is made of three lam-
inations; the middle layer is thinner than the two
outside layers. Eventually, the two layers of bend-
ing plywood would be covered by other layers of carpenter's glue and 1'/4-in. staples. The yellow colored in the location of the rails and stiles with
panels, stiles, rails and moldings. The bending glue turned the two pieces of the bending ply- pencil to make them easy to identify. Bostock
plywood provides a solid base upon which to at- wood into one unit. Bostock used a paint roller drew all vertical lines with a level, and he then
tach subsequent layers. to apply the glue quickly. The staples were measured horizontal distances off a level line
Bostock attached the first layer of bending ply- placed every 6 in. to assure a tight, even fit be- close to the floor (top left photo, p. 80). This hor-
wood with nails and construction adhesive. The tween the two layers. izontal line was established by marking points
construction adhesive filled any gaps between in several locations around the radius with a lev-
the plywood and the studs, and the nails were Lay out the frames and panels—Once the el and connecting the points using a thin piece of
more than adequate for drawing the plywood up bending plywood was attached, Bostock marked wood bent into the curve. Once the layout of the
to the studs. The second layer was put on with the stiles, rails and panels on the wall. Bostock vertical and horizontal lines was complete,

78 The Best of Fine Homebuilding Drawings: Vince Babak

 The traditional look is made of nontraditional methods. By building up layers r different materials, including bending ply-
wood, pine moldings and tempered hardboard, Glenn Bostock duplicated the look of a traditional frame-and-panel wall.

Bostock began the installation of the stile and foundation to which all the components are at- be mitered were 900 instead of 91° or89'/2°, he
rail base pieces. tached. All of Bostock's other layers of paneling took extra time to double-check everything (bot-
either butt against or lie on top of the stile/rail tom left photo, p. 80).
The first layer of the stiles and the rails base pieces. Although the layout of the stile and rail base
must be exact—In a traditional frame-and-panel Attaching the base pieces was simply a matter pieces was critical, the butt joints where the
wall, each stile and rail is made of one piece of of nailing or screwing them to the bending ply- pieces meet was not nearly as critical because
solid wood. Each stile and rail in Bostock's wall is wood. But Bostock stressed that this point the thumbnail pieces and the caps would cover
made of three pieces: the base; the thumbnail; the last chance he had to make sure the butt joints. Forthis curved-wall project, ¼-in.
and the cap (drawing facing page). The ¼-in, by was level, plumb and square, and because sub- thick pieces of wood flexed enough to bend eas-
2½-in, clear, sugar pine base pieces provide the sequent layers would be easier if the 0 ily around the 8-ft. radius of the wall's curve.

Finish Carpentry 79
 Miters have to be exact. Butt joints don't.
Miter joints where the bevel moldings meet will
be visible on the finished wall. Long butt joints
between the bevel moldings and the stile/rail
base pieces will be covered by the layers of the
Pencil in the locations of the frames. Bostock marked out the stiles and rails on the wall. All panel field. Screws used to attach the bevel
subsequent layers of the wall either will butt against or be attached to the sifie/rail pieces. molding also will be covered by the field.

Accurate placement of the base pieces expedites the project. Making sure that the stiles are Fasteners are hidden by gluing on a if-
plumb and the rails level will ensure that the miters of the bevel moldings layers are true 450 cuts nal layer of hardboard. One-eighth-inch tem-
and that the panels pieces have 900 corners. Along the upward sweep of the staircase, some of the an- pered hardboard is used for the panel fields.
gles had to be bisected (sidebar facing page). Rolling on contact cement to the fields and stick-
ing them to the wall covers all the staples,
screws and nails Bostock used to attach the sub-
layers of the panels.

Bostock glued and stapled the stile and rail mation on using a router table, see FHB #90, As with all the previous pieces in Bostock's pro-
pieces into place. p. 61.) A table saw also can be used to make the ject, the bevel moldings must bend. But the
pieces if a shaper or a router is unavailable. moldings are relatively thick and unyielding and
Bevel pieces mimic the edge of a raised Even though the width of the bevel moldings is had to be kerfed to ensure flexibility. Bostock
panel—There are two parts to a raised panel: the somewhat arbitrary, the thickness is crucial. The laid the moldings so that the diagonal cut was
field and the bevel. The field is the large, flat sec- molding must be exactly as thick as the two lay- face down on a power miter box. He then cut a
tion in the center of the panel, and the bevel is ers of the panel base pieces, which in this case series of kerfs into the thickest part of the back of
the diagonally cut border around the field. The totaled % in. More on the panel bases later. the moldings.
panels in Bostock's project are made of four Bostock milled his bevel molding stock to a
pieces: two panel bases; the field; and the bevel width about 2½ times as wide as needed. He ran Mitering the bevel moldings—The bevel
molding (drawing, p. 78). the boards through the shaper twice, raising both moldings had to be mitered to fit within the
Bostock made his bevel molding on a shaper edges. Then he ripped the moldings to their de- framework created by the stile/rail base pieces.
using a panel-raising cutter. A similar profile sired width. This method allowed him to run Because of the geometry involved in flexing the
could be cut on a router table. (For more infor- wider stock through the shaper more safely. beveled molding into place, some gaps were in-

80 The Best of Fine Homebuilding

evitable between the bevel moldings and the Mitering odd angles
framework made of the stile/rail pieces. These
Mitering wood at an odd angle is exacting approximate center of the angle. Repeat
gaps would be covered by subsequent work (top
work, but it's not hard if you have a bevel the process on the other leg. Again, the
right photo, facing page). The miter joints of the
gauge and a compass. The process involves distance between the metal point and the
bevel moldings, however, had to be tight.
bisecting the angle to be mitered. pencil point on the compass is
On Bostock's curved wall the run of the stair-
The first step is to use your bevel to but the setting has to be at least wide
case creates a series of triangular panels, and fit-
transfer the angle to be mitered onto a enough so that the two arcs will intersect.
ting the bevel moldings into these frames posed
scrap of wood or a piece of paper. Next, The intersection point of the two new arcs
greater challenges because their tight angles
stick your compass point in the apex of the establishes the center of the angle. Now
were not 450 miters. Rather, they were more you can reset your bevel to measure the
acute and irregular. In order to fit the pieces ac- angle and strike an arc equidistant from
the apex along both legs of the angle bisected angle from the apex to the point
curately, Bostock had to bisect all the angles in-
(drawing below). The distance at which you where the arcs intersect. Transfer the new
dividually (sidebar right).
strike the arc is arbitrary, but it is probably angle to the pieces of wood that are to
Because of the difficulty presented by their cur-
a good idea if you set your compass so that be mitered.
vature, Bostock fit and installed the horizontal
the marks you make are at least 3 in. from Keep in mind that there is plenty of room
members of bevel molding first. He used yellow
the apex. for error in this process and that it might
glue and 1Y4-in. staples to fasten the moldings,
The next step is to position your compass behoove you to cut your pieces a little long
but some pieces required screws to pull them up
at the point where the first arc intersected so that you can fine-tune the miter with a
tight. The fasteners were kept close to the out-
the angle's leg. Draw a new arc through the block plane.—J. D.
side edge of the moldings so that subsequent lay-
ers would cover them. After installing all the hor-
izontal pieces, Bostock installed the verticals.
Bisecting an angle takes a compass
Installing the panel base pieces—The panel and a bevel gauge. Mitering angles other
than those at 9Q0 can be tricky. The first
base pieces are made of, respectively, kerfed step is to transcribe with a bevel the angle
¼-in, plywood and %-in. bending plywood. Both to be bisected. Next, use a compass and
are used to fill the interior area created by the follow the steps below. After you have
bisected the angle, reset your bevel gauge
bevel moldings. Keep in mind that for a flat wall
to the new angle and transcribe the new
where the panels didn't have to bend, you would angle onto the two pieces to be mitered.
be able to use single pieces of 5/8-in. plywood to
make the panel bases. 1. Strike arcs along
Bostock wasn't too concerned about getting a b angle legs ab.
tight fit between the panel bases and the bevel
moldings because the panel field pieces would
cover the joint.

The stile/rail thumbnail molding must be

coped—The stile/rail thumbnail moldings cov- 2. Strike arcs
er the joints between the stile/rail base pieces from b to c. 3. Angle bac is the
bisection of angle bab.
and the bevel moldings. Bostock made the
stile/rail thumbnail moldings of ¼-in, pine. He
milled a quarter round on one or both edges, de-
pending on each piece's location in the project.
Bostock installed the perimeter pieces of the
thumbnail molding first. Then the horizontal
pieces, or rails, were installed. Rail ends had to
be coped into the perimeter stiles.
On Bostock's curved wall, to fit the curved rails,
the angle-and-cope cut was made on one end
while the other end ran long over the perimeter
stile. He temporarily screwed the rail into place,
which allowed for a direct marking of the rail in
relation to the stile. Once marked, he removed
the screws and routed the cope with a ¼-in, cove
bit in a router table. Then he reinstalled the rails Because nail holes or other mechanical faiten- of the wall, and its smooth surface takes paint
using glue and the same screw holes. ers would detract from finished work ancj be- readily. Like the stile/rail covers, Bostock at-
After all the rails were in place, Bostock coped cause there was no easy way to clamp the pieces tached the panel fields with contact cement. The
and fit the rest of the stiles. Because these pieces for a conventional glue joint, Bostock used don- cement can be brushed on or rolled on; Bostock
are coped into the rails, their length is deter- tact cement to install the stile/rail caps. prefers the latter method. Rolling provides for an
mined by measuring the distance between rails even coating of cement, critical to a good bond.
and adding ¼ in. for each cope. Extra wood is The panel fields are made of '/s-in. hard- When Bostock was finished with the wood-
removed during the course of coping. board—The panel fields are the final pieces that work of the wall, he applied latex caulk and glaz-
Bostock installed on his paneled wall. ing compoend to fill any gaps between layers
Stile/rail caps were attached with contact pieces cover both the fasteners he used to attach that would otherwise reveal that this project was
cement—Bostock made the last piece to cover the outside panel-base layer and the be- not conventional raised-panel construction. E
the rails and stiles—the stile/rail cap—of ¼-in, sug- tween the panel base and the bevel moldings.
ar pine that was planed in his shop. If you don't Bostock made his panel fields of Y8-in. Rem- Jim Donnelly is a cabinetmaker and writer who
have a planer, you could use '/8-in. hardboard, pered hardboard (bottom right photo, facing lives in Bucks County, Pennsylvania. Photos by
which is available at some lumberyards. page). The material bends well around the curve Glenn Bostock.

Finish Carpentry 81
 Raised- Panel Wainscot
Traditional results with table saw and router

by 1. F Smolen

nstalling a traditional raised-panel wainscot Original plaster the panel size to be uniform, so I divided each
is a good way to transform a nondescript wall
section to allow equal spacings between stiles.
room into a more formal space. It's also a The largest section has four panels, each of
handsome alternative to replastering old walls which is 12¾ in. wide; the smallest section
that have been damaged over the years by Colonial-casing
molding used as has a single 18-in, wide panel.
feet, furniture and children. A wood wainscot chair rail I milled the panel stock from 4/4 roughsawn
is more durable than plaster or gypboard, and pine boards about 14 in. wide. The wood,
it relieves the unbroken plane of the wall with originally intended for flooring, was air dried.
the delicate array of shadow lines created by I had one side planed down to a thickness of
moldings and flat surfaces, as shown in the in., which allowed for a full '/s in. of relief
photo at the top of the facing page. on the raised panel and a sturdy 'A-in, thick
1/4-in, expansion'
The term "wainscot" is often loosely ap- space tongue around the panel perimeter. Tight
plied to various paneling treatments that cov- knots were acceptable, because I planned to
er the lower part of a wall. The raised-panel paint the finished wainscot.
wainscot shown here is a traditional style
based on frame-and-panel construction. The Raising panels—I set up a cutting schedule
Raised
frames consist of vertical members, called that included all panels for the wainscot,
stiles, and horizontal members, called rails. crosscutting the planks to 23½-in, lengths,
They support panels whose beveled borders then ripping them to finished width (the dis-
and raised fields give this particular style its tance between inside edges of the stiles plus a
name. Each panel rests in its frame with its ½-in, tongue allowance on each side). Then I
grain running vertically. Panel width is limited made a template of the panel profile, which I
by the width of your stock, unless you edge- set against the sawbiade or dado head when
join two or more boards together. setting up for a cut in order to produce the
Beneath the bottom rail of the frame, a proper bevel and depth.
baseboard extends to the floor. At the top of Making a raised panel with a profile like the
the wainscot, a molding called a chair rail one shown in the drawing below left requires
covers the joint between the top rail and the three cuts on each side—one to form the bev-
upper section of the wall. Backing-band el, one to form the tongue on the panel's edge
The moldings that I used in making this molding
and one to form the shoulder on the edge of
wainscot can be bought at most lumberyards, the field.
and it's possible to make the raised panels To cut the bevel, I used a carbide-tipped
and their frames with a table saw. I used a slot combination blade because it produces a
cutter and router table to groove the inner smooth surface that needs little sanding. To
edges of my stiles and rails, but you could Raised-panel
wainscot set up the saw for the bevel cut, I first set the
handle these as well on a table saw with a construction arbor angle and fence distance to match the
'A-in. dado blade. bevel on the template. Then I clamped a guide
board to the tabletop, parallel to the fence
Panel design—The wainscot that I installed and in. away from it. This guide board
in the dining room of my late Victorian home aligned and steadied the on-edge workpiece
is traditional in design. I wanted its top to be as it was fed through the saw (photo facing
about 41 in. from the floor. This finished page, center). Without it, you'd have a trouble-
height would include an existing 7-in, base- some, hazardous time keeping the bevel
board that could be left in place, and a 4-in. straight and true. Even with the fence, the
wide chair-rail molding that would overlap blade had to cut through just over 2½ in. of
the top rail of the raised-panel frame. With the wood, so I held the stock securely and fed
stiles and rails 4 in. wide, the panels would with slow, steady pressure. The next time I
show 22½ in. of height. Their actual height need a similar setup, I'll use a 4x4 as the aux-
would be 23½ in., since ½ in. of the panel iliary fence.
edges would be let into the grooved frame all Because some boards cup slightly after they
around (drawing, above right). are surfaced, I found it best to make cross-
The width of the raised panels was deter- grain bevel cuts soon after the boards had
mined by the distances between corners and been cut to their finished sizes.
the door, window and cabinet frames in the When the bevels had been cut on all four
room. In each wainscoted section, I wanted sides of the panel, I completed the panel pro-

82 The Best of Fine Homebuilding

file using 1-in, wide planer knives (flat across
the top) mounted in a Sears molding head.
The second cut (photo bottom left) re-
moves a triangular section of waste at the
edge of the panel to create the ¼-in, thick
tongue that fits into grooves in the frame. To
allow for expansion and contraction of both
frames and panels, I trimmed in. from the
top tongue of each panel and ¾6 in. from each
side tongue (wood expands more across the
grain than along it). The inner edge of the
frame would be grooved to a depth of ½ in. to
provide expansion space at the top and sides
of each panel. Allowing for play in the fit of
each panel in its frame is necessary if the
wainscot is to survive years of fluctuating hu-
midity. Too tight a fit, and the panels are like-
ly to check or bow out of their frames.
The third cut produces a ¼-in, wide land at
the juncture of bevel and field, and a ¼-in.
high shoulder where the field begins, 2½ in.
from the panel edge. No auxiliary fence is re-
quired for this cut, but the main fence needs
to be set up exactly right. Here again I use the
template to get accurate settings for the fence
and blade. The blade should just graze the
wood surface.
As soon as the panels were cut, I prefin-

In a traditional raised-panel wainscot, a solid

wood panel sits in a grooved frame. A chair-rail
molding covers the joint between the top rail
and the plaster. Overlapping the bottom rail, a
baseboard extends to the floor, as shown in the
drawing on the facing page.

TaLe-saw setup. Top, a guide board, clamped parallel to me main fence, aligns the panel the bevel Is cut. Slow, firm feed and a carbide blade
produce a smooth cut. For safety, let the blade stop spinning before the waste piece is removed. Above, a 1-in, wide planer blade cuts the tongue along
the edge of the panel. At right, the author uses a tenon-cuffing jig, attached to the saw's miter guide, to cut the tenons on a stile.

Photos: Anne Smolen; Illustrations: Vince Babak Finish Carpentry 83

ished them to keep checking, cupping and
wood movement to a minimum while I built Modified wainscot:
the frames. I filled dings and small knots with
plastic wood; then I sanded the exposed face
a raised panel with birch-veneer
with 80-grit paper and sealed knots with shel-
lac to keep sap from bleeding through the fin-
plywood and beveled molding
ish coats of paint. Finally I gave each panel a
coat of oil-base sealer compatible with the
enamel finish I planned to use.

Stiles and rails—First I ripped 1x6 pine

boards to 4-in, width. I cut all rails about 3 in.
by Michael Volechenisky
longer than their finished length so that after small ad in the local paper got Making moldings—From top to bottom, my
assembly I could scribe them for an exact fit me interested in building a raised-panel wainscot contains a chair-rail molding
to the walls on either side. wainscot. It offered for sale "paneled scribed to fit the plastered wall; two smaller
My design called for a 'A-in, wide by ½-in. wainscoting from a 150-year-old home" moldings (a cove and a beaded stop) that fit
deep mortise-and-tenon joint between stile (wainscot is actually the correct term here). over the frame and raised panels; and a
and rail, so I grooved the inner edges of all This would be just the thing for the dining baseboard scribe-fitted to the floor, with its
the stiles and rails on my router table, using a room in my equally old house, which I was top edge covered by a modified scotia
'A-in, slot cutter. Since the slot has to be cut in the middle of redoing. But the age of the molding (drawing, facing page). I made
down the exact center of the stock, I tested wainscot was unfortunately confirmed by its these moldings with my spindle shaper, but
the setup on ¾-in, thick scrap before running condition, and in any case there wasn't could have bought similar ones at the
stiles and rails through the machine. enough of it to go around my 15-ft. by 15-ft. lumber store.
To cut the tenons on the stiles, I used a dining room. So I decided to build my own I milled the beveled molding for the
thin-rim carbide blade on my table saw and a from scratch, enlisting the advice of Luther panels that make up the wainscot on an
tenon-cutting jig on my miter gauge (photo Martin, a retired builder and woodworker old Hebert molder-planer. The Hebert,
previous page, bottom right). After a little who was sympathetic to the idea of which is no longer made, is similar to the
touchup work with the chisel, I was ready to recreating an old look with new materials. Williams & Hussey planer (Williams &
put frames and panels together.
Hussey Machine Co., Dept. 16, Milford, N. H.
Panel design—I planned to construct a 03055), and both machines are shop-size
Assembly and installation—I clamped the frame-and-panel wainscot along traditional versions of the larger, more powerful planer-
bottom rail of each panel section in the end lines, bordering it with a baseboard along molder machines used by lumber mills.
vise of my workbench and fit each section to- the floor line and a chair-rail molding along My machine has only one cutterhead,
gether dry. Small pencil lines on panels and the top. But Martin didn't want to use solid which is mounted horizontally above the
rails served as registration marks for center- wood panels because he'd seen long table and holds a pair of knives. It is driven
ing each panel in its frame. stretches of wood-paneled wainscot push by a 1-hp motor, and is powerful enough to
In assembling the wainscot, only the stile walls out of plumb as a result of normal complete a molding in a single pass,
tenons get coated with glue. The panels are wood expansion. Cracking and cupping are providing you use knot-free wood that's not
seated firmly in the frame, but not glued. This other risks of solid wood panels. So I made too dense. But to be on the safe side, I
way, they can respond to changes in humidity my raised panels from ¾-in, thick lumber- usually make my first cut to within in. of
and temperature without binding or bowing core plywood with birch face veneers, and the finished dimension and then run the
their frames. 1 used a small brush to spread with slightly modified beveled molding. The stock through a second time to get a smooth
glue on the tenons, and then assembled the molding, which is 2 in. wide, forms the surface that requires very little sanding.
frames and panels, snugging stile-to-rail joints beveled border of the panel, and the lumber- Molder-planer manufacturers sell a variety
together with pipe clamps until the glue set. core plywood is the field. A tight tongue- of molding cutters to fit their machines, but
I wanted to nail the wainscot directly to the and-groove glue joint between border and I've often made my own from precision-
studs rather than installing it over the existing field and two coats of white enamel hide the ground tool steel (it comes in many sizes
plaster, which was sound but presented quite fact that these panels weren't raised in the and thicknesses, and can be bought
an irregular surface. Leaving the original traditional fashion. wherever metalworking tools are sold). After
baseboard intact, I snapped a level line about This alternative design has several drilling two holes in each tool-steel blank so
'A in. above the installed level of the top rail advantages. First of all, the lumber-core they can be bolted to the cutterhead, I
and ripped out the plaster and lath. I scribed plywood field is far more stable than its transfer my molding outline to the blank and
the plaster along this line with a utility knife, solid wood counterpart. Expansion and start removing metal. I hacksaw as much as
then ripped it off by hand. I cut the lath with a shrinkage are negligible, as are cracking and I can, grind the shape to a 30° bevel and
chisel and pried it off the studs with a ham- checking. And the birch face veneer is better hand-file corners and coves where my bench
mer and small prybar. than solid pine or fir if you're planning to grinder can't reach. After getting both
Next, I nailed up the paneled sections, paint—as I was—because it contains no cutters as nearly identical as I can, I mount
which I had purposely built slightly wider knots or resin pockets, which could bleed them in the Hebert and run a trial piece of
than the spaces they would occupy so they through the finish. wood through. This tells me which cutter is
could be scribe-fitted to the walls. This left a Third, you can give the panel's bevel a doing most of the work by the flecks of
slight gap between baseboard and bottom rail fancier treatment than is possible with wood that adhere to its cutting edges. More
and between plaster and top rail. I used a 4-in. conventional techniques, since the border filing follows; then I hone the blades and
wide Colonial-casing molding at chair-rail isn't an integral part of the field. The install them. Since most of my molding runs
height to cover the joint between the rough molding I used, for example, has a quirk are for 500 ft. or less, it doesn't seem
plaster and the top rail of the wainscot, and a bead at the inner edge of the bevel—an necessary to harden my cutters.
15/s-in, wide backing-band molding at the embellishment that suggests far more All the moldings, stiles and rails for my
baseboard and bottom-rail junction. E intricate work than was actually involved. wainscot were cut from basswood stock I'd
been saving. Basswood works easily, and
Ted Smolen practices law and does amateur Michael Volechenisky lives in Sayre, Pa., and I've found that you can usually smooth it by
woodworking in Dan vers, Mass. Pompano Beach, Fla. Photo by the author. hand with a cabinet scraper, with little or no

84 The Best of Fine Homebuilding

sanding. It also takes a fine coat of paint, Wainscot construction
because it's knot-free and resin-free. Section view
No matter what type of wood you use on a
molder, you'll get better results if you make Chair-rail molding
sure that each board you run through the
machine has its grain oriented correctly Cove moding
relative to the cutterhead. To prevent small
chips and tears in the molding, feed your
boards so that their grain slants down
toward the exit side of the machine.
Beaded stop molding
I constructed the panels first. Each com-
pleted panel actually consists of five
pieces—the birch-veneered lumber-core 2-in wide upper rail
field and four bordering molding sections.
As shown in the drawing, the tongue along Modified casing molding
forms border of
the inner edge of the molding is designed to raised panel.
fit in a 'A-in, wide groove cut in the edges of
each panel. I grooved the panel edges on my
table saw, using dado blades. At the corners
of the panel, adjacent molding sections are
mitered. Once all the parts for a panel were
cut and test-fitted, I glued them up.
After the panels were finished, I built their 3/4-in, thick lumber-core
frames. lop and bottom rails for each 15-ft. plywood with birch
side of the room were cut from 16-ft. long face veneer
basswood boards. Using a hollow-chisel
mortiser chucked in my drill press, I Cut
mortises in the rails to receive stile tenons. 2k-in. by i'è-in. reveal creates,
Then I grooved the inner edges of the frame shadow line around field
on the table saw to receive the 'A-in, by of raised panel.
'A-in, tongue around each panel.
I test-fit the frame-and-panel sections for
each side of the room, then glued and
clamped them together. Before installing
each section, I gave the back of the frames
and panels two finish coats—the same 3-in, wide lower rail
number that the front of the wainscot would
receive, Thus both sides of the wood can
respond equally to temperature and
humidity. Perhaps this wasn't necessary, but
there hasn't been a single paint crack in the
wainscot after eight years on the wall.
Though I could have installed the wainscot
directly over the dining room's old plaster
walls, they were in such bad shape that I
stripped the room down to its studs and
nailed up rock lath. Then new plaster was
applied down to a temporary ground I nailed
just below the height of the chair-rail
molding that would top off the wainscot.
The frame-and-panel assembly was the
first part of the wainscot to get nailed up. I
made sure each wall section was level and
used 8d finishing nails, positioning them
close to the rail edges so they would be
hidden by the covering layer of molding.
Once the four frame-and-panel sections
were up, I added moldings to the rails. The
baseboard and the scotia-style molding
covering its top edge were mitered at the
corners, as were the chair rail and its two
adjacent moldings. A light sanding, followed
by a primer coat and two coats of semi-gloss
Kemglow enamel, finished the job. 0

The gluing setup used to construct the panel

consists of four bar clamps and a Formica-
faced base slightly shorter and wider than
the panel. At panel corners, the molding joint
Is a glued miter.

illustration: Vince Babak Finish Carpentry 85

 Drywall Detailing
An alternative to wood trim around doors, windows and skylights

by Dennis Darrah

rimming out doors, windows and skylights lights, archways, and even doors. Although Co. (4430 Gibsonia Rd., Rte. 910, Gibsonia,
can be a rewarding endeavor. It can also be the labor involved is comparable to that for in- Pa. 15044; 412-443-7080). I have not used the
costly, if done with the proper care and mate- stalling wood trim, the materials are much blueboard and veneer-coat system of plaster,
rials that this labor-intensive procedure re- less expensive. Don't get me wrong—I love so here I'll concentrate on the standard dry-
quires. Unfortunately, as a house nears com- beautiful trim work. But also appreciate a
I wall accessories.
pletion and costs are running over estimate, project brought in on budget. And even where Corner bead is the obvious choice for form-
this can be one area of a job that experiences money isn't the overriding factor, drywall de- ing wall returns into windows, without the use
severe cutbacks. The most frequently used al- tailing may still be the most effective and of casings. It's available in galvanized steel
ternative to first-rate trim is inexpensive pine, pleasing solution to many trim problems. This and white vinyl. Corner bead can be attached
applied as plainly as possible. Many owner- is particularly true in renovation, where old by driving a drywall nail every 4 in. to 6 in.
builders, and even some contractors, work must abut new. along each side of its length. The length of the
this route with the argument that one can go nail will depend on the thickness of the drywall;
back through the house when time and money Mud and metal—The basic materials for de- the nail should penetrate the framing in. to
allow, and redo the trim in proper fashion. Of tailing wall openings are the various beads, in. One advantage of metal corner bead is
course, we all know that the chances of this moldings and tapes that are available through that you can attach it with a crimping tool,
happening are slim. a good drywall supplier (drawing, p. 88). If which crimps the edges of the bead so that it
A more graceful alternative to the pine solu- there isn't a good drywall supplier in your grips the drywall. This saves quite a bit of
tion is drywall detailing around windows, sky- area, one good mail-order source is Bon Tool time, especially on large jobs, and also helps

Detailing a window. Before installing corner bead, the author trims the edge of the drywall with a surform plane (left photo). The bead is attached
with a metal crimping tool (right photo). A strip of drywall mesh tape spans the gap between the two pieces of corner bead.

86 The Best of Fine Homebuilding

to center the bead. If the bead is rolled to one stand up as well to daily abuse. Despite this
side, it makes the other side difficult to cover product's limitations, there are times when it
with compound. Though I use a crimping tool is the only option.
to set the bead, I always add four to six nails Where the drywall meets the window sash,
for extra security. Any structural movement, you have two alternatives: J-bead or L-bead.
particularly in a new house, is liable to show The basic difference is that J-bead shows as
around windows and doors, and the first casu- the final detail, while L-bead is mudded over
alty will always be the drywall work. But there (drawing next page). J-bead must also be in-
are times when nails can't be used, particular- stalled when the drywall is installed. Which
ly in old houses where the interior walls have one I specify depends on the type of track in
been covered with a layer of foamboard insu- which the sash is mounted. J-bead works well
lation beneath new drywall. The foam and for movable sash—window movement might
drywall create built-up corners through which cause the mud needed for L-bead to crack
nails won't easily reach. In such cases, a crimp- over time. L-bead works better for windows
ing tool is all one can use in applying stan- with complicated track where J-bead can look
dard metal corner bead. Crimping tools retail too busy. However, J-bead actually enhances
for $60 to $70. simple retrofit track, creating a handsome
There is an alternative to the rigid metal or thickening effect at the juncture of the return
vinyl corner bead: a paper tape lined with two and the sash. Although you should feel free to
strips of thin galvanized steel, which forms a use one or the other based on your own aes-
rigid corner when folded lengthwise (photo thetic, there is one caveat: if you expect high
next page). It's sold under the names Flex-Cor- condensation levels on the window or sky-
ner and Sure-Corner, and it comes in widths of light in question, stick with J-bead. Your
in. to 4 in. This product is especially use- mud work over the L-bead is liable to be
ful on oblique or acute angles, the kind that damaged by the moisture collecting on and
you're likely to find in skylight openings. It is around the window.
designed, like standard paper tape, to be em- I try to have the painters spray the J-bead
bedded in a layer of mud and then mudded before it's applied. Spray paint adheres better
over. It is folded at the center margin and ap- to metal than does a brushed-on coat. An
plied with the metal facing the wall. In my es- ideal alternative, at least for white walls, is to
timation, it is not quite as serviceable as rigid use white vinyl J-bead, but this is usually hard
corner bead; it doesn't always adhere well, it A stool cap and some paint are the only trimr to find in small quantities. I also recommend
occasionally has to be reworked and it doesn't work materials needed around this window. that moisture-resistant drywall be used for all

Three coats of joint compound are applied with a 6-in, knife (left photo). The opening is with 120-grit, open-weave silicon paper (right
photo). No sanding is done until after the final coat has been applied.

Finish Carpentry 87
returns around windows and skylights that have headaches, I would go strictly commercial; conditions. And be forewarned: no amount of
potential condensation or moisture problems. there are usually no ceilings to contend with, paint can hide a lousy taping job. Success de-
Windows are not usually that tricky or time- and most commercial spaces are lit with dif- pends not so much on how fast you can
consuming to finish (bottom photos, previous fuse or fluorescent lighting, which is very for- move, but rather on how fast you can work
page). On most returns one can usually catch giving of taping irregularities. I'm not down- well—it's not speed alone, but speed com-
both the L-bead and the corner bead in one playing the quality of commercial work, it just bined with quality that counts. When I finish-
pass of a 6-in, knife. If J-bead is used, all you offers a different set of challenges than a cus- sand a custom job, I hold a spotlight in one
have to finish are the outside corners, though tom home. On a commercial job, for instance, hand and a sander in the other. Around sky-
on large houses this can add up to quite a bit— a good taper really has the chance to show off lights, I try to sand on cloudy days. The light
keep that in mind when pricing out the job. his speed. is more diffuse then, and the spotlight helps
I finish skylights much like finish win- In a home, on the other hand, your taping
I
me to pick out minor irregularities that would
dows. First I apply the corner bead to the has to look good under wildly varying lighting be masked by strong sunlight. Or I sand at
square corners and the flex tape to the ob-
lique angles. I use the mesh tape on the in-
side corners—in my experience it holds the
compound better than paper tape. I also use A bead sampler. There are a wide variety of drywall beads on the market. Although you'll find
mesh tape to span the joints between two slight variations between manufacturers, the basic products are essentially the same. Below are
pieces of corner bead or a piece of corner a few of the more useful ones for trimming out doors and windows.
bead and a piece of flex tape. This is neces- 1. J-bead. Gives a finished edge without joint compound. Use it when the drywall edge must be
sary because the corner bead and flex tape isolated from a window sash or a door jamb — where there's a potential for condensation, or
can't overlap. Their combined thicknesses where door or window movement could crack the finished compound. Ins tall it before applying
the drywall. Galvanized steel J-bead can be painted before installation; white vinyl J-bead usually
would result in a lump that no amount of tape doesn't have to be.
would cover. For the first coat, I tape the in- 2. L-bead. Leaves a crisp edge where the drywall meets the sash or the jamb. The exposed leg
side corners, let them dry, then tape the out- is finished with joint compound. Available in metal or vinyl.
3. U-bead. Gives a clean edge along with isolation of the drywall. Face nail and finish the same
side corners. That way I'm not constantly fight- as L-bead. Usually installed before the drywall. Available in metal or vinyl.
ing overlapping knife marks. I use three coats 4. Standard corner bead. Comes with a deep knurling for reinforcement of outside corners.
of compound. Less compound is used in the Available in galvanized steel or white vinyl.
5. Rounded corners. Compound is applied to the legs; the nose is left bare and painted.
second and third coats, and because overlap- Available in white vinyl or paper-faced galvanized metal.
ping marks are less of a problem, those can
be done all at once. I don't usually sand until
after the final coat is done. I sand with 120-
grit, open-weave silicon paper; it cuts better
than standard paper, and it doesn't load up
with compound as fast, either.
Be aware, however, that skylights take time
to do right. If that isn't bad enough, your tap-
ing work is under the scrutiny of the most se-
vere lighting possible—direct sunlight. To look
good, it has to be done well. A bad taping job
is the worst of all possible finishes.

Levels of perfection—I do both custom

homes and commercial work. If I were to
choose only the biggest buck for the least

An alternative to rigid corner bead: paper tape

laminated to two strips of thin galvanized steeL
It forms a rigid corner when folded length-
wise, Although not as sturdy as rigid bead, it's
the only choice when taping oblique angles.

6. Archway bead. For trimming arched or rounded window and door openings. The white vinyl
vanety has a rounded edge; metal bead has a squared edge.

88 The Best of Fine Hoinebuilding

night, and check it out by the light of the next tant to mask out the timber before ber-frame ceiling is to install the from
day. I find it hard to sand in the glare of direct the drywall. This not only speeds up the tap- above. Simply prepaint the ceiling drywall and
sunlight and have it come out right. ing and painting, but protects the timbers lay it on top of the timber joists, gluing and
Drywall can be dry-sanded, wet-sanded or throughout the whole process. Mud is applied screwing it as you go. Lay the subfloor and
sponged. There are arguments in defense of to the wall abutting the timber, then pai4 finish floor on top of the drywall, then screw
each method. I've never wet-sanded. I used to tape embedded flat in the mud, its edge care- the drywall up into the subfloor. With typical
sponge, but I never got good results from it. fully aligned with the timber. The paper tape 2-ft. centered timber joists, there is no ceiling
The sponge always left visible ridges and will cover any gaps between the timber and taping to do whatsoever. To facilitate accurate
smears that didn't seem to happen with dry- the drywall. If the gaps are excessive, they can screwing into the joists, pre-snap chalklines
sanding. I now dry-sand and swear by it. The be pre-filled with a patching plaster or even a on the back of each piece of drywall. For
tools involved are hand sanders, pole-mount- spray foam. beams with rough surfaces, you can protect
ed or hand-held. The sanders are rubber- A trick for speeding up the finish of a tim- the drywall by installing a strip of closed-cell
backed and take precut sandpaper (100, 120, sill sealer between the joists and the drywall.
or 150 grit) or Fabricut sandpaper, which is an I did one renovation job on an old timber
open-weave silicon carbide sanding cloth that frame with exposed rough beams on 4-ft. cen-
does not readily clog with dust. This comes in Detailing a door casing ters. Using the above system, I saved myself
100 and 120 grit. Power tools have no place in some 280 ft. of flat taping next to the edges of
sanding seams. For one thing, you risk ripping Plan view I

the beams. A word of caution, however—be

the drywall paper with a power sander. Be- careful not to step through the drywall before
Covb the subfloor is in place.
sides, if the seams are that heavily overloaded moclir
with mud, go back and do a skim coat or get
out of the taping business. The bottom line —You're limited only by your
I have only had one opportunity to trim imagination in the use of drywall enhance
out interior doors with drywall. The rough ----V interior finishes. For example, I once built
opening is wrapped with drywall, the corner drywall-formed light fixtures in a cathedral ceil-
bead applied, then the door jamb and the ing (photo below right), along with a false col-
mud. A piece of cove molding helps ensure umn to provide a wire-chase and to hide a
that no gaps will develop over time between structural steel rod.
the jamb and the drywall. Although the job Once you get good at doing drywall details
really could be done without the cove mold- you'll be in greater demand. You may find
ing, I feel the door opening looks a lot better contractors and architects consullirkg you be-
ii 'ii q fore completing the design of a house. If you
when it's there.
work it right, you can get paid for your design
suggestions, as well as for their execution. LII
I

Flat-taping—Being a timber framer in addi-

tion to a drywall taper, I've done a lot of flat-
taping next to exposed timbers (photo below Dennis Darrah specializes in d,'ywall finishing
left); flat-taping is only necessary where stress- and timber-frame constmction. He lives in Mont-
skin panels aren't used. I feel that it's impor- pelier, Vermont.

A timber-frame alternative. Install the floor Drywall-formed light fixtures are a graceful alternative to a truncated cathedral ceiling. Besides
joists, then lay the drywall from above and permitting overhead lights, they add detail to the apex. Note also the drywall-formed
avoid taping against the beams. wire chase beside the chimney.

Drawings: Bob Goodfellow

Finish Carpentry 89
 Building a Fireplace Mantel
Fluted pilasters and decorative trim make for a mantel
that appears complicated but really isn't

by Gary Katz

first glance, fireplace mantels seem im- methods might be safer, they are slower and less
pressively intricate and outrageously expensive, accurate than a table-mounted router.
but frequently they aren't. I know because a I use a ½-in, round-nose (core box) bit to cut
good client asked me to build a copy of a mantel the flutes, set to a depth of ¼ in. To make the
from his previous home. After seeing the origi- process easy, I place multiple strips of medium-
nal mantel and identifying its parts, I was able to density fiberboard (MDF) between the rip fence
build the new mantel simply and economically and the workpiece (top photo, facing page) and
using solid stock, plywood and manufactured remove one strip after each pass, making it un-
moldings (photo right). Even though I've been a necessary to adjust the rip fence for each flute.
trim carpenter for many years, anyone with a ba- For the fluting on these 6-in, wide pilaster's, I
sic knowledge of woodworking can build a fire- make the MDF strips 1-in, wide: ½ in. for the flute
place mantel with the techniques I used here. plus ½ in. for the space between flutes, which
Mantel built in three sections. A 2x4 prop gives me even spacing across the face. The MDF
Strips of MDF simplify flute spacing—For this (top photo) holds the mantel top level while the strips stay put because they butt into a stop block
mantel, the first things to build are the pieces author nails a pilaster to the panel wall. The that's screwed to the table extension. The stop
flanking the firebox, or the pilasters. I make each three components—two pilasters and a mantel block is positioned so that the workpiece bumps
pilaster with three pieces—two sides and one top—were built off-site; the texture on the oak against it, resulting in flutes that end 1¼ in. from
face—mitered toge:her along their length. I rip comes from scraping out the soft grain. the bottom of each face.
the faces and sides from solid oak and cut them On the MDF strips, I draw a line 1¼ in. beyond
slightly longer than their finished length. Once the feed side of the router bit; this line indicates
the pilasters are built, I'll trim them to size on a ra- the beginning of the fluting. I carefully lower the
dial-arm saw. through the fence six times to make parallel dec- workpiece past the line for the initial plunge and
Although I rip the sides of the pilasters to width orative grooves, or flutes. slowly pull it back until the end of the stock lines
with one edge square and the other beveled 45°, Although fluting looks ornate, it isn't difficult up with the 1¼-in, mark across the MDF strips.
I rip the faces slightly wide and with square to accomplish with a table-mounted router. My Then I feed the board through until it butts into
edges. Later, I'll bevel the edges at 45° for Makita portable 8-in, table saw is designed with a the stop block, remove the workpiece and an
mitered joints, but right now I'd rather push a router mount underneath, so the bit sticks up MDF strip and repeat the process five more times.
square edge along the table saw's rip fence. The through an opening in the table. The fluting also After fluting the faces, I rip the edges at 45°.
long point of a bevel can jam under the fence, can be cut by running a router along a straight- Then I spread glue on the miters of the side
and in this case I'll be pushing the face stock edge or by using a router guide. While these pieces and the face, hold the joints tight with my

90 The Best of Fine Homebuilding

Side view shows combination
of plywood and solid wood

molding

2x ledger supports
mantel top.

soffit of oak
Frieze
plywood Cut multiple flutes wiulout moving me fence. The pilaster faces were fluted
on a router table by registering them against 1.in. MDF strips instead of directly
against the rip fence. Alter e ch pass, a strip is removed; the rip fence is not moved.

The complicated look

of the mantel comes
from applying
molding on top of
other pieces and
using blocks that step
out in dimension.
These techniques Joining the pilaster to the
create the mantel's
dynamic profile. Any capital. After building up the
wood that is not capital with two 1-in, oak spacer
exposed is plywood; blocks, the pilaster's plywood
exposed wood is end is screwed to the capital. All
solid stock. of the blocks are flush.

Pilaster face

Plywood

Base

Shoe molding

Marble flooring
Use plywood where It won't show. to me pilaster, a uackboard elimi-
U 1 nates the need to scribe fit fireplace's marble veneer to the pilaster profile. For
economy's sake, the backboard consists of two pieces of solid oak on the exposed
ends and plywood in the hidden center, all glued and biscuited together.

Finish Carpentry 91
Drawings: Christopher Clapp
 material to; otherwise the installer would have
to cut that marble to fit the profile of the pilasters.
With the pilasters attached to the backboard, the
installer has an easier job.
Because most of the backboard is hidden be-
hind the pilaster (there's a 3A-in. reveal beyond
the capital and base), there's no reason to make
the whole backboard of solid stock. I use two
pieces of 1x4 oak for the area that won't be cov-
ered by the pilasters and a plywood filler in the
center. I join these three pieces with biscuits to
make a 12-in, wide backboard. With the pilaster
covering most of the backboard, none of the
joints will be visible; the joints just have to be
Frieze box has plywood upper half and solid-wood lower half. This part of the mantel top flush to avoid gaps between the pilaster and the
will be capped with a shelf and covered with trim, and only the solid-oak lower half will be exposed. backboard. I make the backboards longer than
The plywood and the solid wood are biscuited together, and blocking supports the joints. the pilasters. Once the pieces are assembled, I'll
trim the backboards flush with a circular saw.

Assembling the pilasters—Now it's time to put

all the pieces together. I glue up the first spacer
block and center it on a capital or base so that
the reveals are an even 3/4 in. The pieces must be
flush at the back, or else there will be gaps be-
tween them and the backboard. I nail the first
block in place and position the second smaller
block in the same manner. Then I attach a capi-
tal and a base to each pilaster, running screws
through the backer blocks to draw the first spac-
er block tight (center photo, p. 91). I also nail the
backboard to the pilaster so that there's a 3/4-in.
reveal at the capitals and bases.

The frieze is just a big box—I build the top of

the mantel starting with the frieze box that will sit
above the capitals and support the mantel shelf.
In this case I copied an existing frieze box,
though usually I rely on a scale drawing or a
mock-up of the mantel to determine the frieze
dimensions. This frieze box overhangs the capi-
tals by 3/4 in. both in the front and on the sides.
Fitting the colonial stop. The joint between Fastening the crown. Once the snelf is fas- Although the frieze box is high, I use lo-
the frieze box's solid-oak lower half and its ply- tened to the frieze box, the crown molding is ap- in. wide solid oak only for the exposed portion of
wood upper half is covered by a 1x4 oak apron plied. The front piece goes on first, and the the box, the part below a 1x4 oak band, or apron.
to which is affixed the colonial stop. mitered ends are glued and nailed in place. Over this apron I apply colonial doorstop and a
wide crown molding, both stock items. The up-
per portion of the frieze box, therefore, is cov-
ered with trim, so I make the upper portion of
fingers and nail the miters together. Plywood or The spacer blocks are ripped from 1x2 oak and the box from %-in. plywood and biscuit it to the
MDF backer blocks nailed inside the pilasters at are mitered on a chopsaw. I make them with four solid oak (top photo, left).
each end support the joints (left drawing, p.91); sides, like picture frames, with all miters glued Building the frieze box is just like building the
the backer blocks also serve as a backing both and nailed together. I miter the corners because capitals and bases in that the front and sides are
for the bases and for the capitals, which are the this mantel will be stained rather than painted, mitered, glued and nailed, except I install blocks
decorative blocks that will adorn the bottoms and I don't want any end grain to show. across the biscuit joints to strengthen the con-
and the tops of the pilasters. The capitals and bases are built just like the pi- nections. I also nail in a couple of midspan
It's important to close up the miters as tightly as lasters, with a face and two sides mitered togeth- blocks that provide backing for the bottom of
possible while the glue is still wet. I use a scrap of er. I use my table saw and rip fence to cut the the box, or the soff it, and the backing for the lid,
hardwood or my hammer handle to roll the 450 bevels, glue and nail these joints together,
both of which sit within the frieze box. The soffit
sharp edge down just slightly, which flattens the roll the edges and sand them. I install a lid at one will be exposed, so here I use oak-veneer ply-
mitered edge and closes the joint. Fine sandpa- end using any material handy, usually 3/4-in. ply- wood. The lid of the box can be made of any-
per finishes this process, leaving the mitered wood. The lids are backing for attaching the thing. Its only purpose is to provide backing to se-
edges softly eased. spacer blocks, and they aren't exposed. cure the mantel shelf.
When installing the backing blocks, I take care
Making capitals and bases—The pilasters have Backboard eliminates scribing—On this job, a to hold them 11/2 in. inside the back of the box so
a capital and a base; each capital and base has marble-veneer fireplace surround will be applied that I can slip the box over the I V2-in. mounting
two 1-in, spacer blocks (right drawing, p. 91). The after I install the mantel. To make the mason's ledger and have it fit snug to the wall.
bases, capitals and spacer blocks are solid oak, job easier, I mount the pilasters on a backboard On the assembled frieze box, I draw pencil
and each one is larger than the one that it's (bottom photo, p. 91). This board provides a lines showing the positions of the crown mold-
attached to, creating a profile of 3/4-in, steps. straight line for the marble installer to butt the ing, colonial stop and 1x4 apron. I place a short

92 The Best of Fine Homebuilding

piece of crown on the frieze box with the top of
the crown butted against a framing square as if it
were the underside of the mantel shelf and scribe
a line along the bottom of the crown. From that
mark I locate and draw lines for the colonial stop
and for the 1x4 apron. I cut and install the apron
and the stop, mitering and gluing the outside cor-
ners (bottom left photo, facing page). The crown
stays off until the shelf is installed.

Putting on the shelf—I make the mantel shelf

from a single piece of 6/4 solid oak. The owner
wanted the end grain to show, so I didn't miter
the ends. On this job, the shelf overhangs the top
of the crown molding 3 in. That overhang is con-
sistent around the front of the mantel, too, so the
width of the shelf is equal to the frieze box, plus
the coverage of the crown, plus a 3-in, overhang.
After ripping, crosscutting and sanding the shelf,
I use a 3/16-in. roundover bit in my router to ease
the edges.
To secure the mantel shelf to the frieze box, I
place the shelf upside down on my workbench
and position the frieze box on the shelf, centered
and flush at the back. Then I screw through the
frieze lid into the mantel shelf.

Installing the crown molding—With the shelf

in place, I now can install the crown molding
(bottom right photo, facing page), which goes
under the shelf and flush with the back of the
frieze box. The main concern is getting the
crown's outside miters tight.
I place the front piece of crown upside down
on my compound-miter saw as if the saw table
were the mantel shelf and the saw fence the
frieze, and cut one end at a 45° miter. I tack this
piece in place on the colonial stop, then cut the
return piece. I'm no mathematician, but I am fair
at trial and error, and that's the method I use to
find the perfect angle on a compound-miter saw
to cut crown molding. I cut scraps until they fit Masking tape makes the scribe line visible. Although the pilaster's backboard eliminates
just right to determine the setting for the saw. the need to scribe the pilaster to the wall, the capital must be scribed to fit beneath the frieze soffit.
Wish it were easier. The pilaster is shimmed up until it touches the soffit; then it's scribed and cut. Masking tape makes the
Once the saw is set, I cut the right side of the scribe line easier to see on the finished capital.
front piece first, then I hold it up to the colonial
stop, where the crown eventually will be nailed. I
check the fit on the right end with a short scrap,
also mitered, then mark the bottom back edge of allowing me a little room to adjust the final loca- oak plugs to cover the screw holes, then sand
the left end. This is the ideal spot to guide in the tion of the mantel top. Then I fasten the ledger to them smooth for the finisher to touch up.
blade of my Hitachi sliding compound-miter saw. the studs with 3½-in, screws and panel adhesive.
Once the miters are tight, I cut the backs of the On a masonry fireplace I would use plastic an- Installing the pilasters—Because the pilasters
return pieces so that they are flush with the back chors and panel adhesive. are mounted on backboards, they only need to
of the frieze box. Then I glue the mitered ends The frieze box slides over the ledger, and a cou- be scribed to the frieze soffit (photo above), not
and securely nail all three pieces both to the stop ple of 2x4s temporarily support the box while the to the wall. If there were no backboard, I would
and to the shelf. shelf, the soffit and the sides are scribed to the install a cleat on the wall, much like the mantel-
wall. The mantel top is prefinished, so I top ledger, and after scribing the pilaster to the
Mantel top hangs on ledger—The mantel top masking tape along the edges and scribe the wall, nail the pilaster to the cleat. In thIs case I
is mounted on the wall at the combined heights lines on the tape, which makes them easy to see. don't need a cleat. I spread panel adhesive on
of the pilasters, the frieze box and the shelf, plus I cut the top with a worm-drive circular saw. the backboard and nail it right to the wall. I hold
1Y2 in. of clearance for the marble flooring. I Because I cut all the scribe lines at a slight bevel, the pilasters I ½ in. above the floor, leaving plenty
scribe this height on the wall and measure down the bottom of the cut won't jut out and cause of room for the marble flooring to slide under
from the mark the thickness of the mantel shelf gaps to be visible between the mantel top and the base. The resulting ¼-in, gap leaves an ex-
and the frieze lid combined, which is 2¼ in. With the wall. cuse to return to tack on some prefinished shoe
a straightedge I scribe a line at this elevation Once the mantel top fits perfectly tight to the molding and to take a picture (top photo, p. 90)
across the wall. This line locates the top of the wall, I countersink holes and screw through the of the finished mantel for my portfolio.
ledger that supports the mantel top. I rip the mantel shelf into the ledger. I use an inexpensive
ledger from a piece of 2x8 and crosscut it 1/2 in. plug cutter by Vermont American (P. 0. Box 340, Gary Katz is a carpenter and a writer in Reseda,
shy of the inside dimension of the mantel shelf, Lincolnton, N. C. 28093; 704-735-7464) and make California. Photos by the author.

Finish Carpentry 93
 _______

Building a Federal-Style Mantel

Neoclassical detailing in poplar and ceramic tile

by Stephen Sewall

or the early European settlers in America, about Dutch tile and wanted to incorporate a The basic frame—I built the mantel out of
domestic life centered around the hearth. As reproduction of it in a new house they were poplar. White pine might have been a more
the source of heat and the means of cooking building. Working from a photograph in the traditional choice, but it's more difficult to
food, it was a natural gathering place for the magazine, a friend of mine, David Stenstrom, carve and dents more easily. The basic frame,
family. The earliest fireplaces simply had oak made a measured drawing of the mantel, and or field, was 5/4 stock that I dressed to in. (I
lintels supporting the masonry, but gradually that gave me a starting point. added scribe strips later). There are two verti-
the wooden framework around the firebox be- My clients wanted to reproduce the entire cal pieces (side frames), each 81/4 in. wide by
came a decorative element. By the Federal pe- fireplace, including a tile surround, a slate in. long, which form the base for the ta-
riod (1790-1825), fireplaces were still the sole hearth and a cast-iron fireback. Therefore, the pered pilasters, and a horizontal piece 14¼ in.
source of heat in the home, but weren't neces- size of the firebox and the exact dimensions wide by 69¼ in. long (drawing below). I had
sarily used for cooking. Formal fireplace man- of the mantel were determined by the tiles we to glue up two boards to make the horizontal
tels, with their elaborate carvings and moldings, selected (Delft tiles, manufactured in Holland piece, and when I did I oriented the growth
became symbols of wealth and prosperity. The by Royal Makkum and imported by Country rings the same way. When boards are glued
term mantel refers to the entire decorative Floors, 15 E. 16th St., New York, N. Y. 10003- up with the growth rings oriented in opposite
framework around the fireplace opening, not 3 104). The opening of the mantel is 12 tiles directions, the resulting piece has a tendency
just to the shelf above it. wide and 8½ tiles high. Once these dimen- to be wavy if it does cup.
I recently built a reproduction of a Federal- sions were established, the slate hearth was The horizontal piece had to be as smooth
style mantel (photo facing page) from a house ordered through the Bangor Slate Co. in Ban- as possible because I was going to rout flutes
in Croton-on-Hudson, New York. My clients gor, Me. The hearth is 16 in. wide, 72Y2 in. in it, and a wavy surface would mean that the
had spotted the mantel in a magazine article long and 2½ in. thick. flutes would be uneven. After face- and edge-

Mantel A Federal-style mantel

shelf Horizontal
Mantel section (5¼ in.
—--------------.
frame
in.) (14¼ in. by
69¼ in.)
Cornice -
Side
13/e-in. by frame
13/8-in. cove Frieze
Carved
block

Arch itrave
half-round
with flat
sections

Pilaster

Tapers
from
33/81fl
to l¼in.(j
7/8-in.
side frame
'1/2-ifl
scribe strip 8¼in.
Pilaster screwed and glued to frame from back.
Plinth

94 The Best of Fine Homebuilding

jointing the pieces, I glued the boards togeth- I marked the holes between the flutes with the width of the board (top photo next page).
er, using biscuit joinery to align the pieces. a compass—drawing the entire circumfer- The width of the hole cut in the jig is the same
Then I made one more pass to joint the en- ence rather than simply marking the center as the base plate of the router I used—a 3-hp
tire width before planing and sanding to the point—because I was cutting them on a drill Makita plunge router fitted with a %-in. core
final 7/8-in, thickness. press with a Forstner bit, which has no cen- box bit. The length of the hole creates the
I joined the horizontal piece to the sides of ter spur. These bits cut flat, stop for the machined portion of the flute. I

the frame with a miter at the inside corner holes. The molding separating the architrave marked center line, top and bottom on the jig
where the joint would show, then switched to from the frieze was also penciled in, and I and aligned these marks with the center lines
a butt joint where it would be hidden behind drew in the location of the carving block in for the flutes. Because the force of the rout-
the pilaster. I cut slots for biscuits along this the center of the mantel, along with the cor- ing was resisted by the cleats underneath, I

joint as well, but did not glue the frame to- nice molding. I laid out the 5/s-in. flutes last, needed only one clamp to hold the jig in
gether yet because it's easier to do the rout- keeping the area between them—known as place. I made two passes with the router: one
ing and carving first, before gluing up. the land—as close to in. as spacing would to remove most of the stock and a second
permit. I marked these flutes by their centers light pass to give a clean, smooth cut.
Routing and carving flutes—The anatomy because they would be cut with a router. The ends of the flutes had to be carved to
of a mantel relates directly to the elements of The jig for cutting the 5/8-in, flutes was made make them look better. I chopped the bottom
a classical entablature and is typically com- of '/2-in. Baltic birch plywood with cleats end of each flute into a convex shape with a
posed of two pilasters supporting the archi- screwed underneath to position the jig along 5/8-in. gouge. With the same gouge, I tapered
trave, frieze and cornice, the top of which
serves as the mantel shelf. I cut 5/8-in, vertical
After seeing a magazine photograph of an 18th-century mantel, the owners of this house com-
flutes down the middle of the horizontal sec- missioned the replica shown below, including everything from the carvings to the DeIft tiles,
tion of the frame, which corresponds to the slate hearth and parged firebox.
frieze in the entablature. Just below these
larger flutes, I carved sets of smaller, 5A6-in.
flutes, alternating with 7/8-in. dia., ¼-in deep
holes. These smaller flutes and holes make
up the architrave. I laid out all the details in
pencil, drawing both sides of the smaller
flutes (rather than just center lines) because
they were to be carved entirely by hand.
These flutes were so narrow and short that I
could carve them faster than I could rout
them. The bottoms of these smaller flutes run
right off the edge of the board; the tile mold-
ing I applied later creates the stop.

Frame joint
Butted portion
hidden behind
carved block.

==

The curves were laid out with a

pattern, then cut on a bandsaw. Side
cuts were made first, but weren't cut
all the way through until after the
front Cut was made.
 the top of the flute to a slight point. As modi-
fied, the flutes look to be pointing upward
(bottom photo).

The pilasters—Made up of four sections, the

pilasters are 58% in. long overall. The top
piece (1 in. by33/8 in. by 14¼ in.) is carved in
a sunburst motif—a vertical ellipse with an
oval applied to the center (more about this
later). The flutes radiate out and stop in a
convex chopped end. I also carved another
set of 5/16-in. flutes across the bottom of the
carved blocks.
The second section of the pilaster is I in.
thick by 35¾ in. long, and it tapers from 3¾
in. down to 1¼ in. at the foot. I cut the tapers
with a tapering jig on the table saw—the
same way you would cut a table leg. I started
by drawing the tapered cut lines on the
stock. For a jig, I found a scrap of plywood
slightly longer and wider than the stock and
ripped it on the table saw once, leaving the
rip fence set. I lined up the cut line of the
stock with the ripped edge of the plywood jig
and screwed small notched blocks into the
jig at either end of the stock to hold it in
place. Then I ran the jig through the table
saw again and had my first tapered cut. I cut
one side of both pilasters, then repositioned
the blocks on the jig for the last two cuts.
With a shop-built jig and a %-in. core box bit in a plunge router, the author cut the vertical flutes The foot of the pilaster was cut from a
that create the frieze under the mantel shelf (photo above). Where the joint would show between block in. thick by in. high by 2% in.
the horizontal and vertical members of the frame, the pieces are mitered (photo below). The wide. I made a paper pattern to lay on the
butted portion will be hidden behind the pilaster. Notice how the ends of the flutes are carved so block and marked the three sweeping curves.
that they appear to point upward. (Photo below by author).
Then, on the bandsaw, I cut most of the way
through two of the lines (pilaster drawing
previous page). 1 stopped short, though, so
as not to lose the third line. I cut the third
line all the way through, then finished off the
other cuts by eye. The plinth block at the
base of the pilaster projects ¼ in. out from
the foot on three sides. The block is in.
thick by in. wide by 43/4 in. high.

Assembling for humidity—When all the

carving of the frame was done, I glued it up.
Later, I sanded it carefully, then attached the
four sections of each pilaster with glue and
screws run in from the backside. At the top,
where the carved sections of the pilasters cov-
ered the joint between the vertical and horizon-
tal portions of the frame, I attached only the
carved sections to the vertical boards. Their
.4' grain ran in the same direction, and the boards
would therefore expand and contract in the
same way. If I had attached them to the hori-
zontal portion of the frame (the frieze) as well,
the boards might have split because the grain
would be running perpendicular.
By not attaching the small carving, the
frieze is allowed to move freely with the sea-
sonal changes of humidity. The mantel has
been in place through two winters now, and

UI 4
the miter on the inside of the frame has
opened up slightly, indicating that the frieze
board has shrunk in width.
The sunburst carving in the center of the en-
tablature was done on a board ½ in. thick by
in. high by 16 in. long. I mounted this

Drawings. Bob LaPointe

temporarily onto a piece of ¾-in. piywood and to catch the light. I mounted the large carving The strips were 1 in. wide to allow good nail-
screwed a small block on the back of the ply- to the frieze with glue and screws because the ing to the wall. Before the installation, I took
wood so I could hold the whole setup in a vise grain was running in the same direction. time to sand the carvings just enough to smooth
for carving. I clamped a Universal pattern-mak- any rough edges and didn't tiy to remove the
ers woodworking vise onto my regular bench Applied moldings—The mantel shelf is a 1- tool marks. I broke all of the sharp edges of
vise to give me the extra height to work com- in. thick piece of poplar 80¾ in. long (the the wood so that the paint would adhere well.
fortably and prevent back strain (photo below). same length as the frame). It is 5¼ in. wide, It was much easier to do all of the finish sand-
To lay out the carving, I calculated the allowing a ½-in, scribing space to account for ing with the mantel flat on a work bench.
lengths of the major and minor axes (14 in. an uneven wall. The front edge of the shelf The installation was easier that I had expect-
and 8 in., respectively) appropriate to the size jogs out 1 in. over the pilasters. I created these ed because the slate hearth was installed level.
of the block. I made up a trammel using a jogs by sawing out the center portion of the I only had Ia scribe the mantel to make it
scrap of wood, drilling a hole in one end to mantel on the table saw. I finished off the cut plumb to the wall, which was out some. After I
hold a pencil. Two nails driven through the by hand and smoothed it up with a chisel. nailed through the frame at the scribe strips, I
trammel guide it in making an ellipse. The first A 5/8-in. ogee-and-bead molding was glued slid cedar shingles behind the frieze at framing
nail is half the length of the minor axis from to the front edge of the shelf, and the shelf points and nailed off the mantel. There was a
the pencil, and the second nail is half the was set aside to be installed last. I ran a ¼- bow in the wall, so I did have to scribe the
length of the major axis from the pencil. I drew in. half-round molding around the tops of the mantel shelf before nailing it to the cornice.
intersecting lines on the block along the two pilasters to cover the joint between the ta- When the tile was installed—a blue and
axes and placed a steel square over one of the pered section and the carved block. This also white Delft pattern called "Little Flowers—we
quadrants created by the intersecting lines. Us- forms a stop to the flutes on the carved were careful to leave an even space between
ing the trammel, I drew one quarter of an el- blocks. I applied a 3/s-ifl, half-round molding the tile and the frame. This would allow for a
lipse, keeping the nails against the square (draw- to the junction of the tapered leg and the tile molding that would come to the edge of
ing below). I repeated the procedure with each foot, then added a '/2-in, half-round at the the tile, yet not overlap it too much. In this
quadrant and had ellipses drawn on the block. base of each foot, resting on the plinth block. case the space varied from ¾ in. to in. I
I penciled in the center oval by eye. Then, A 5/8-in. wide cove-and-bead molding that made up a 7/8-in, half-round and scribed it to
with dividers laid out equal segments on
I projects in. is just under the main carving the tile so that it projected out from the face of
the ellipse. I drew lines from these outside block separating the architrave from the frieze. the mantel by ¼ in. and nailed it to the edge of
points to the center to represent the land be- It continues around the pilasters and dies into the frame.
tween the rays. I used three sizes of gouge, the frame. The same molding is used in con- After the tile was installed, the sides of the
each with the same sweep, to carve the rays. junction with a large 1%-in, by 13/a-in. cove to firebox were parged. This left a clear transition
I had to work carefully to avoid tearing out form the cornice under the mantel shelf. between firebox and tile and also set off the
grain. I turned an ellipsoid on the lathe, cut it All of these applied moldings were cus- cast-iron fireback. The mantel was then paint-
in half and sanded down the section to about tom-made in the shop. They were glued in ed with a traditional grey paint from Pratt &
in. thick. The oval was glued to the center place where the grain direction matched that Lambert (1234 Saline, North Kansas City, Mo.
of the carved block. The smaller carvings at of the piece on which they were mounted. 64116) called Gossamer II. The mantel re-
the top of the pilasters were laid out the same Otherwise, just nailed the moldings, but I
I quired one primer coat and two finish coats.
way as the large carving in the center of the did glue all the miters. Poplar doesn't seem to cover as well as
entablature. The primaly difference in the carv- I glued ½-in, scribe strips on the back of the
ings is that the flutes on the large carving end- two outside edges of the frame (no nails be- Stephen Sewall of Sewall Associates, Inc., is a
ed with a reverse bevel, creating a flat surface cause I might have hit them when scribing). custom homebuilder in Portland, Maine.

Pencil
Trammel
Majo axis
¶
First nail

Minor
Second
axs nail
I
I

Blade Tongue

Drawing and carving an ellipse. 1) The first nail was set half the length of the minor axis from the
pencil, the second nail half the length of the major axis from the pencil. 2) With a framing square
over one of the quadrants, the trammel was moved so that the nails rode along the blade and
tongue of the square. 3) The square was moved to each of the other three quadrants and the
procedure repeated. 4) Dividers were used to step off the segments around the outside of the
ellipse. The last step (photo right) was to carve the sunburst rays.
 Baronial Inglenook
Laminated mahogany arches and antique English details surround a fireplace

by Scott Wynn
 he pulsating electric-blue glow of the tele- width of the casing as it moves from the control on these critical laminae should they
vision has long since replaced the flicker of to the apex of the arch. I think this gives the need shifting or an extra bit of pressure.
firelight in most houses. Modern, efficient arch a bit of tension, making it seem less statici I used MDF for the form because it's cheaper
heating systems have just as surely rendered Somewhat more difficult to determine was than plywood and stronger than parhcleboard.
obsolete the fireplace and its messy stack of the style and size of the moldings I wanted to After cutting out the basic shape, I used a rasp
kindling and logs. Nevertheless, the comfort- apply to the casing. A plain casing seemed to take out any inaccuracies, checking con-
ing image of a hearth, a fireplace and the appropriate, but it needed a feeling of depth stantly with a square to make sure the edges
magnetic dance of the flames will probably al- and richness, of being "deeply carved." I set- remained perpendicular to the face of the form.
ways remain central to our concept of home. tled on a half-round and a cove (drawing A, When I was satisfied with the line, I took a
Creating a place for appreciating the fire and next page). To maintain the tension, these second piece of MDF and cut it to within in.
its tradition was one of my primary tasks in moldings had to taper to the apex of the arch of the line, then clamped it to the first and
this remodeling project. along with the casing. trimmed it to the exact profile with a flush-
My client has a house that think is best
I
trimming bit in a router. Accuracy is critical
described as Tudoresque. It doesn't have the Scaling the arch—I began to construct these here because any variation between the two
heavy-timbered, adzed-beam woodwork typi- arches by blowing up the scale drawing of sides of the form will cause the laminae to
cally associated with the Tudor era, and that's curve to full size. I did this by plotting points shift out of alignment as they are glued up. I
fine with him. His tastes tend more toward or- above the springline of the arch at every ripped the scrap into strips to use as spacers
derly and refined trimwork—the kind that more frequently at the smaller radius curve—i-- between the two sides of the form.
wouldn't be out of place in a ship-captain's onto brown paper (drawing B, next page). i I Next I glued and nailed two layers of '/8-in.
private cabin or in a baronial library. The fire- connected the points, sketching a curve and bending poplar to the business edge of the
place inglenook that I eventually designed and correcting it as needed. Once I was satisfied form provide a smooth and consistent sur-
built not only became the focal point of the with the curve, I drew a second line in. oth- face to back the clamped laminae. I took care
living room (photo facing page) but also set side of it. This line compensates for the five to set the nails so that no dimples would mar
the tone for the rest of the remodel, a project layers of '/8-in, thick larninae of the arched the clamping surface. Then I drilled 1'/2-ifl.
that occupied my crew and me for nearly jamb, and the two 1/8-in. thick pieces of ply- dia. holes near the form edge at roughly 4 in.
three years. wood that would be applied to the bending to 8 in. apart to allow room for the clamps
form for a clamping surface (drawing C, p. (drawing C, p. 101).
Already gutted—The job began in late spring Using carbon paper, I transferred this line 'a Finally I could begin assembling the arches;
of 1986, when my client bought a shell of a sheet of %-in. MDF. Note that the form is I started by resawing mahogany for the jambs.
house in the hills of Berkeley, California. The cave—not convex. That's because lik I I chose a piece of 1x8 for the faces that had a
former owner/builder had planned on remod- have the finished side where I can see it striped figure, and set aside the first two
eling the house, but after eight years of fitful ing glue-up, which gives me a little bit r pieces that came off the saw so I could later
effort, he bogged down and put the house up bookmatch them. The faces are backed up by
for sale. one more layer of mahogany, and three of
Evolving arches. The author used the poplar bending plywood. I oriented the face
We decided to stick with the existing floor over the front door (photo below) as a
plan. The existing living-room space, howev- of departure for a unifying design elei grain of two of the layers of bending plywood
er, was too narrow. Widening the room over Inside, the inglenook (photo facing pag so they were perpendicular to the length of
its entire length was impractical because the framed by a softer, curved version of the the arch; this would counteract springback
made of laminated mahogany. The COfl( from the mahogany. I've found that if you ori-
second and third stories were bearing on the fireplace surround repeats the same arch
living-room wall. We decided that a large fire- surround blocks were cast on site in ent the face grain of all the layers of bending
place alcove would add space as well as visually mine-coated particleboard forms held te poplar perpendicular to the curve, the radius
expand the room and give it a focal point. er with screws for easy disassembly. will actually decrease as the glue dries.
I also cut enough of the bending plywood to
Finding the arch—The exterior of the house use as clamping cauls for the curved portion
was decorated in false half timbers (photo of the arch. For the cauls, the grain runs paral-
right). This seemed to make our choice of Tu- lel to the curve to help spread the clamping
dor a rather obvious one, but I needed a motif pressure. On the straight parts of the arch, I
to tie it all together. I took my cue from the used a piece of scrap MDF as a caul. All the
entry to the house, which was framed by a cauls were about a in. wider than the jamb
stylized, angular Tudor arch. We settled on a to ensure full bearing.
softer, more elegant variation of it called a
"four-center arch." Its name implies that the Glue-up—I used Titebond, an aliphatic resin
arcs originate at four points, but I couldn't glue (Franklin International, Corporate Center,
find a formula for laying it out; I relied on my 2020 Bruck St., Columbus, Ohio 43207; 614-443-
eye instead. 0241), to bond the laminae together, and spread
I used the shape of the arch liberally around it around with a 4-in, putty knife. I find this
the house. But by far its most notable applica- method to be faster than using a paint roller,
tions were as laminated jambs and tapered though it takes practice to get the glue thick-
casings around the fireplace and the entry-hall ness consistent. Either way, it's a messy job.
openings (photo next page). To make sure I could later release the jamb
from the form, I lined it with wax paper.
Refining the casings—I made repeated line I collated the glue-covered laminae on a flat
drawings on graph paper to determine the workspace, clamped one end to keep them
right width of the casings in relation to the from shifting around, and then took them to
size of the opening and finally settled on one the form. Jamb in place, I started at the small-
8 in. wide. I knew, however, that making the est radius and worked outward, setting the
width of the casing continuous around the clamps as I went.
opening would result in an awkward and dis- Once the glue set (I allowed at least 12
proportionate look. Instead, I narrowed the hours) I removed the iamb, scraped off the ex-

Finish Carpentry 99
cess glue along the edges (this speeds curing) small shop; therefore, ran them through the
I full-size template on brown paper (drawing D,
and set the jamb aside while I glued up the table saw, making sure to keep the back of the facing page). I joined them with biscuit wafers
next one. Before I started to joint any of the jamb tight to the table as it passed the blade. A at the miter, and then glued the casings to the
jambs, I let them sit for at least 24 hours after I long fence to serve as a reference plane is prac- jambs. Once the glue was dry I used a saber-
removed them from the form. tically a necessity for this operation. saw to cut away the larger portion of the
After taking a light pass on one side, I used waste. I followed this by using a router with
Straight edges—To clean up the edges of a hand plane, when necessary, to take out im- a flush trimming bit to match the casing to
each jamb I removed the remaining lumps of perfections. Once I was satisfied with the edge, the exact curve of the arch. Then I cut the
glue with an old iron plane, sighting along the I ripped the opposite edge, leaving enough cove in the edge of the casing with a piloted
edges to get them as straight as possible. Nor- along the first joined side to be able to rip a cove bit.
mally, I put a laminated jamb on the jointer, clean edge and still get my finished width. Next, I used a straight bit with an adjustable
and with a helper, maneuver it over the knives, pilot to cut a slight rabbet at the edge of the
taking off just enough to get a straight line. Casing layout—The casings for each arch cove. This makes a narrow reveal that mimics
That's the typical sequence, but in this case, section join at the tight radius in the corner. I the setback of a casing around a door. I did
the long pieces wouldn't fit on the jointer in our plotted their angle and orientation by doing a this on all the arch pieces, and then, using my

Arched jambs. Laminated mahogany jambs trimmed with tapered casings frame the entries to
the dining room, living room and the inglenook in the distance. The arch also turns up in smaller
scale as a border atop the fret-sawn panel applied to the newel post in the center of the photo. A. Section through Drywall
trim at arches
2½

in.

3/4-in, radius

Laminated jamb

Casing

B. Laying out the arch

Plotted points

Edge of paper as spring

line (where curve begins(

Line of jamb face

paper template, I laid out the line of the out- To do the radius portions of the half-round scraped and sanded the little blocks to match.
side edge of the arch on the casing and cut it moldings I glued four blocks of mahogany to- The moldings, as they go from the radiused
with a sabersaw. gether, with the grain oriented diagonally, and corner to the apex of the arch, are not only
attached them to a piece of plywood (drawing tapered but curved. I roughed out the mold-
Coves, half-rounds, donuts—I fashioned the E, below). I mounted this assembly on the ings on the bandsaw, then closed in on the
moldings with a router, using standard cove faceplate of my lathe, turned the blocks intQ final profiles with router bits. I finished them
and roundover bits. The half-round was done "donuts," and then ran the router over them with molding planes. On the cove portion, I

with passes on opposite sides of a mahogany with the half-round bit. Then I bandsawed actually took a piece of the standard cove mold-
2x3. Even though the bit has a pilot, I used a them off the plywood backing and glued them ing and bandsawed a bit off both sides to get
fence to guide the router. That's because on in place, trimming moldings on the straight the taper. Then I finished the cove with a mold-
the second pass, the pilot rides on the pre- section of casing as required to match. ing plane, flexed it into place and glued it.
viously cut radius, preventing it from cutting a The radiused cove sections were too small
true half-round. I glued and biscuit-joined these to make easily with a router, so I bandsawed a Jambs on site—What I took to the site looked
moldings to the straight portion of the jamb, piece to fit and installed it unshaped. After the like a truckload of huge hardwood boomer-
running them slightly long. adjoining moldings were installed, carved,
I angs. I had sections for two arches, each one

E. Making corner blocks

3/
1. Glue up Direction 2 Turn
blocks of grain donut

C. Bending form

Spacers from waste•

Line of cut

Line of face of form equals

line of cut p/us two layers
¼-in, bending plywood.

in. equals three layers

¼-in, bending plywood plus
two layers ¼-in, mahogany.

2. Slide top backband piece to donut, scribe to fit. Finish cut by

placing handsaw in joint, evening kerf. Repeat until fitted.
D. Casing layout 1. Trim end of donut approximately
perpendicular to line of backband.

Bandsaw cove
block to fit; shape
to match
adjoining coves.

Cove

Line of jamb face

Finish Carpentry 101

divided into finished halves with moldings mount- and 17th centuries. One excellent reference I I decided that a single large mantel would
ed on both sides. I made the sections so that I used is A Histoiy of the English House by Na- be too rustic. Instead, I combined a more so-
had about 3 inches of play per side between thaniel Lloyd (Omega Press Ltd., I West St., phisticated curved mantel that reiterated the
the back of the jambs and the rough framing. Ware, Hertfordshire, England). From this refer- crown molding at the ceiling. The overniantel
To fit the sections, I first checked how level ence I learned about the individual decorative between them would be divided into horizon-
the floor was across the opening to find the elements of the fireplace and found useful ex- tal sections: a base, a top frieze and a band
low side. Then I slid the jamb assembly for amples that I could mix and match to make between them containing three decorative pan-
that side over the wall, plumbed it, and scribed an elegant fireplace surround for this job. els (drawing below). These would be divided
it to the floor. I marked a cutline at the apex, To get my proportions, I began with the fire- by pilasters similar to the ones that support
removed the first assembly, slid the second box opening. We wanted this to be as large as the mantel.
assembly over its wall and repeated the proce- we could get it, limited only by the ratio of the
dure. Itrimmed the arch tops to the pre- opening to the cross section of the flue. We Refining the zones—Having arrived at the
scribed marks, and then slid them back in decided to use a manufactured metal flue, as proportions of the different elements, I began
place. To get the fit just right at the top, I but- the cost of a seismically reinforced masonry to articulate them. In some places, I used
ted the two halves together and ran a Japa- flue would be prohibitive. Our maximum flue moldings to differentiate or highlight areas.
nese draw saw, made for cutting hardwoods, area was 154 sq. in. (14-in, flue). Working on For example, a watertable molding with a
between the butted ends to create an equal the theory that says the ratio of firebox-open- cove accents the step between the base of the
gap. Then I pushed them together for a tight ing-to-flue area should be 10 to 1, we could overmantel and its central panels, and an as-
fit. Last, I nailed the assemblies to the wall. have a firebox opening of approximately 38 in. tragal heightens the border between the frieze
They are secured to the framing by toenailed by 40 in. board and the carved panels.
finish nails at the base and finish nails run The fireplace opening carries the motif of The overmantel pilasters (photo facing
through the casings into the studs. the arch. By code, any wood has to be at least page) are Jacobean in their design: a flat-
12 inches away from the opening. I laid out faced column tapering in two dimensions to
Surrounding the fireplace—The richness these dimensions on a piece of paper, along the plinth. The plinth has a double-tapering
of the wood and the complexity of the trim in with the height and width of the alcove. Then I cove top, and contrary to tradition, the capital
the rest of the house demanded that the fire- divided the surround into individual zones tapers to a waist. The applied geometric pat-
place detailing be of great intensity, if it was (drawing below). Pilasters flank the opening, terns, such as the diamond plaques, are typi-
to be the focal point of the house. To inform supporting a mantel. Above that is an over- cally Jacobean as well. I chose the sections of
my designs, I researched English residential mantel, which is capped at the ceiling by a the moldings and their proportions based on
architecture, with an emphasis on the 16th crown molding. photos of work from that era. I also used

Zones of fireplace surround Ceiling Individual elements

2x6 blocks affixed

to studs provide
nailing for over-
mantel

Fret-sawn panel
over 1/4-in, plywood
backing

-- Plywood nailer
16 in. o. C.

Plywood backer

Plywood spacer

Plywood backer
screwed to framing

:rete surround

Plan detail of pilaster

some dentils, which seem to be common to overmantel pilasters, were still about even with Next I used chisels and gouges to distinguish
nearly all eras. the major pilasters under the mantel. The to the levels of different elements. Once these
The most difficult part, however, I had yet to crown forms kind of a soffit over the whol were established refined the volumes and
I

tackle. Namely, how much projection does ev- thing (drawing facing page). I
details of the figures, taking care to not let
erything have? It was much harder for me to some areas get too far ahead of others.
visualize the relative depths of the parts—how Making the parts—I took care to avoid I don't use sandpaper unless absolutely nec-
much they would need, how much would be crossgrain construction—instead, large panels essary, so to finish the figures I made burnish-
too much—than perhaps any other aspect of float inside their frames. also used as few
I
ing cuts and used tiny curved files called "rif-
nails as possible, as they are always visible. flers." The trick to burnishing is to make cuts
the design.
I began by determining priorities based on The mantel is backed by a piece of plywood that are nearly parallel to the surface of the
visual importance. The mantel would be the for a nailing surface. The overmantel is work. Opposing cuts must meet exactly so
strongest line, the crown at the ceiling would frame-and-panel assembly joined with there is no torn wood where a chip has been
step back from that but still make a strong cap Its pilasters are screwed to it from the back. "pried" out. The rifflers make it easy to clean
at the ceiling. The pilasters needed to be un- I cut the pilasters from solid stock and in- the crevices.
mistakably unified with the mantel, while the cised their converging flutes with a router and
overmantel had to fall somewhere between a corebox bit. I did this by carefully piottingi Installing the mantel—The alcove's panel-
the location of the flutes on the finished face ing had already been installed, so I took the
the mantel and the wall.
To assign dimensions, I started with the pi- before cutting the side tapers. Then I ripped a mantel to the job site in four pieces (mantel,
lasters. I didn't want a glue line to show, so I taper parallel to the center flute and used it as overmantel, and two pilasters). Each piece
made them from 16/4 stock. They were ap- a guide for my router to cut the flute. I had been prefinished in the shop, with three
plied over backing pieces that are U-shaped in ripped tapers parallel to the next set of coats of sanding sealer and four coats of lac-
plan (detail drawing, facing page). Their com- and again used them to guide my router. I did quer. I also brought enough crown molding to
bined thickness, plus the curve of the mantel's this successively until all the flutes were cut, finish the alcove and enough horizontal band-
crown molding, determined the projection of then ripped the final shape of the pilaster. ing to extend the lines of the mantel and the
the mantel. Simple enough. The overmantel Because I needed less than 15 ft. of most base of the overmantel into the corners. I
was less so. I felt its deepest part should still molding shapes, I made them in the shop us- coped one end of all the moldings in the
be slightly proud of the wall, so I began step- ing router bits, or roughly shaped them with shop—the large crown molding is particularly
ping forward from there with the different ele- the table saw and then worked them down difficult to cope, so doing it in the shop on the
ments, in 3/4-in. and 1Y2-in. increments. I ad- with planes. Any flat spots left by this I bandsaw was much easier than field-fitting.
justed the steps so that the foremost part, the removed by using either flexible scrapers or The day before we instahed the mantel, I se-
scrapers ground to the exact profile. The latter cured horizontal 2x6 blocks to the wall behind
I ground from throwaway Japanese I the overmantel. I positioned them exactly so
find they remove stock very quickly, and fre- that I could nail through the sides of the over-
quently there's no need for follow-up mantel, out of sight, into the end grain of the
2x6s. There was nothing to nail to across the
Carved panels and capitals—For decora- center of the mantel—so it had to be nailed
tive motifs for the overmantel panels I referred near its ends.
to two sources: Historic Ornament: a Pictorial I affixed the mantel's wide shelf to the bot-
Archive, by C. B. Griesbach, (Dover tom of the overmantel assembly. That allowed
tions, 31 E. 2nd St., Mineola, N. Y. 11501, me to fit the many pieces of narrow cove
1975. $10.95, softcover, 280 pp.; molding that wrap around the bases of the
and Manual of Traditional Woodcarving, by small pilasters while still in the shop. Also,
Paul N. Hasluck, (Dover Publications, 1977. leaving the wide shelf off the mantel assembly
$10.95, softcover, 568 pp.). 1 found some gave me access to the mantel's plywood back-
propriate floral decorations from the era and ing from above. Once I had it positioned atop
redesigned them to fit our needs. the pilasters, I affixed the mantel to the studs
Ifound a lot of other details in these with nails through the plywood (drawing fac-
books for miscellaneous decorations that ing page). To attach the two mantel sections, 1
worked into this job. I
toenailed the crown molding to the bottom of
I used a technique called "fretted carving"
the wide shelf.
to make the panels. This is simply a Truth be told, the only significant hitch that
design applied to a panel before the final we encountered during installation was a re-
tails are carved. While the finished sult of my not accurately measuring the space
look relieved (photo left), this method is a lot for the overmantel. The embarrassing thing is,
faster. I thought we had enough room for a ½-in, re-
I had originally envisioned the pilasters
veal all around the top, but because of vari-
topped by crouching medieval ations in tne framing and taping (which I

goyles of some sort. These evolved, thought I had accounted for) the overmantel
into figures of repose and meditation more was about Y8 in. too tall. Because of its assem-
classical and refined (photo, p. 98) than the bled weight, this was no small obstacle. Luck-
figures 1 first imagined. These were my conL ily it was hitting the ceiling in the back, out of
cession to the present: I didn't really set out to sight, so by removing wood there we were
make a reproduction of a style but rather some- able to get it in, though not without lifting it
thing more of an impression of a style. up, driving it on and prying it off three
Before I picked up the chisels, I made clay
mockups of the figures to help me Scott Wynn is an architect/contractor who also
them. Then I sketched their outlines on the designs and builds furniture in San Francisco.
faces and sides of the 16/4 pilaster stock and All drawings are by the author. Photos by
cut away the major waste with a bow sawF Charles Miller.

Fi& arpentry 103

A Built-In
Hardwood
Hutch
When working with solid
wood, joinery techniques
must accommodate
seasonal movement

by Stephen Winchester

IThe opening. New

studs frame me sides, nut Ash matches. A new ash hutch built into an old closet looks like the chestnut woodwork of the
the back wall of this former closet was straight- original room. The hutch was finished with two coats of Minwax Polyshades—half maple and half
ened with lx3s and shims. walnut—followed by a slightly thinned top coat.

love an old house. Working on one makes et in each room, one on both sides of the wall, studs. So I straightened the back wall with
me appreciate the skill of the carpenters who and when the wall came down, the closet area shims and 1x3 strapping (left photo, above). At
came before me. It's amazing to see the level of was a natural location for the built-in hutch. Built- the bottom I tacked a 1x3 across the old wall
craftsmanship the old-timers attained using only ins ought to look good and last a long time, so and into the old studs. Placing a straightedge
hand tools—especially in their trimwork. I re- this hutch was built of solid hardwood and de- on the 1x3, I tucked some shims behind the
cently renovated an early 1800s farmhouse in signed to accommodate wood movement (draw- low spots to bring them out to the straightedge.
New Hampshire that had some beautiful chest- ings facing page). But before I started building, I Next I tacked a 1x3 to the top, again shimming it
nut trim. I got the chance to match this wood- straightened and leveled the closet area. straight. Then I tacked on more horizontal lx3s
work when I added a family room with a built-in in. o. c. Moving from left to right, I held the
hutch (right photo, above). Roughing in the hutch—New studs on the left straightedge vertically, against the top and bot-
I made the new family room by removing a and right made the sidewalls plumb and straight, tom strapping, and shimmed the intermediate
wall between two small rooms. There was a cbs- but there wasn't room on the back wall for new strapping out to the straightedge. The wall was

104 The Best of Fine Homebuilding

Left photo, this page: Stephen Winchester
 47
Detail of glue joint
Shaping the edges of boards
with a glue-joint cutter (or router
bit) helps align boards that are being
glued up and pro vides a larger gluing
surface than square edges do.

Putting the carcase together

The cabinet floor went in first, nailed along
the front only to allow for wood movement
at the back. Then the spline-and-groove
wainscot below rhe countertop was blind-
nailed in place, with the back wall nailed up
first. The countertop rests on the wainscot,
Wainscot divider but a gap was left at the back to allow for
is biscuited in place. movement. Then the upper wainscot was
installed; two full lengths of wainscot fit into
notches in the front of the countertop.

straight when all the pieces of strapping were almost every American chestnut tree. Today, you glue-joint cutter in the shaper. (Jointing is the
even with each other. can get salvaged chestnut from old buildings or process of straightening a board's edge or face
The hutch rests on a 2x4 base; I installed it level get it resawn from beams or sheathing, but it's and is typically done with a jointing plane or with
by shimming the low end and nailing it to the expensive. I chose white ash instead, which has an electronic jointer. Joining is the process of
new 2x4 walls on each side. With the new level about the same grain pattern and texture as the connecting two boards) The glue-joint cutter
base, I didn't have to scribe the cabinet sides and chestnut woodwork on this job. But ash is hard, makes edges that look something like shallow
back to the floor, which had a big hump in it. so it's more difficult to work than chestnut. finger joints (detail drawing above). These edges
align the boards and provide a larger gluing sur-
Chestnut substitute—Chestnut was once used Gluing up wide boards—The cabinet floor and face than simple square edges do. Glue-joint bits
for almost eveiything in a house, from sheathing the counter were glued up out of several boards, are also available for use in router tables.
to door and window frames to trim. But during as were the wide shelves for the bottom cabinet. First I lined up the boards so that their grain
the first part of this centuly, a blight wiped out To joint and join the boards in one step, I used a matched, and I marked them so that they

Drawings Bob Goodfellow Finish Carpentry 105

 Cap

Head

Trimming the cabinet

The cabinet face frame was
assembled, then glued and
screwed in place. The head casing
was mortised to fit over the legs
to keep the joints from opening
up. A fixed, full-span shelf above
the countertop supports a divider
and adjustable upper shelves.

Pocket-screw joinery. To attach the bottom rail to the stiles, a spade bit makes a pocket hole that's 1% in. short of the rail's edge (left). A pilot hole
is then drilled up through the edge to connect with the pocket hole and the boards - I and screwed ter

wouldn't get mixed up during the glue-jointing Spline-and-groove wainscot—One of the orig- cutter on the shaper, but a ¼-in, slotting cutter in
operation. I used numbers—Is on the first two ad- inal small rooms had beaded wainscot all the a hand-held router or a dado-blade assembly in
joining edges, 2s on the next two and so on. way around, so I decided to use beaded wain- the table saw would work, too. I centered the
I don't have a wide planer, so I had to flatten scot inside the hutch. To make the wainscot, I ½-in, deep groove on the edge of the board. The
the glued-up boards with a belt sander. With a ripped ash boards on the table saw into random '5A€-in. splines were ripped from ¼-in, plywood. I
60-grit belt, I sanded across the grain first, then widths, from 5¼ in. to 3¼ in. didn't use biscuit joinery because, when wain-
with the grain. Then I used a 100-grit belt and fin- To join the pieces, I used a spline-and-groove scot shrinks, gaps appear between the biscuits. A
ished with a 120-grit belt. The countertop, the joint rather than a tongue-and-groove joint (detail full spline looks like a solid tongue.
most visible of these wide boards, was finished drawing, p. 105). First I jointed the edges of each Using a beading bit, I beaded one edge of each
using 180-grit paper on a random-orbit sander. board. To make the groove, I used a ¼-in, straight board to match the original chestnut wainscot.

106 The Best of Fine Homebuilding

Installing the wainscot—I installed the floor of Matching molding mark 1½ in. from the edge of the rail. Then I
the cabinet first, flush to the front of the 2x4 base. drilled a pilot hole in the edge of the rail at an an-
I nailed the floor at the front only and left a 3/8-in. gle up through the pocket hole (middle photo,
space at the back to allow for wood expansion. facing page). Finally, I squeezed a generous
I put up the wainscot for the bottom half of the amount of glue between the stiles and the rails,
hutch by blind-nailing through the splines and clamped them together and ran the screws in
into the walls as I would any T&G material. I (right photo, facing page).
didn't glue the splines because each piece of After the glue was dry, I sanded the joints flush
wainscot should expand and contract indepen- and installed the face frame. I glued the bottom
dently. This wainscot rests directly on the cabinet rail to the front edge of the cabinet floor and
floor; if the floor butted into the wainscot, a seam screwed the top rail to the underside of the coun-
would open. I avoided visible seams in the cor- tertop (drawing facing page).
ners by putting up the back wall first and then
butting the sidewalls into it. And I allowed for ex- Mortised head casing—The trim, or casing,
pansion by installing the first board on the back AL....8 saw and a shaper were used to around the hutch was installed next. The 3/4-in.
wall % in. from the corner. make ash molding (right) that matches thick side pieces, or legs, went on first; I ran
I also made a wainscot divider for the bottom the original chestnut trim (left). them ¼ in. long at the top. The 1-in, thick top
cabinet. It was biscuited to both the floor and piece, or head, was mortised to fit over the legs
the underside of the counter. I used just a dab of (detail drawing facing page). You could think
glue in each biscuit slot to prevent any unneces- of this as being a mortise-and-tenon joint, with
sary glue squeeze-out. the legs being the tenons. I set the head on top
1. The author started
The counter sits on the wainscot. Before I in- with a length of 11/16-ifl. of the legs and with a sharp pencil traced the
stalled the counter, I notched its two front edges, by 1 Vs-in. ash stock. outline of each leg onto the bottom edge of
which would allow an entire length of wainscot the head. Then I scored the marks with a sharp
at the front of each sidewall. Along the sides, the knife. Scoring makes for a cleaner mortise. I mor-
counter is nailed into the wainscot so that it stays tised these sections of the head a good ¼ in.
put, but to allow for expansion and contraction, deep with a hinge-mortising bit in my small
the back edge of the counter isn't nailed. router. Finally, I used a chisel to square the cor-
Now I was ready to put the wainscot in the ners of the mortise. This joint practically guaran-
2. The first pass was tees lasting beauty: If the header shrinks, the joint
top of the hutch. I set the wainscot on the coun- with a Ys-in. bead cutter
ter and blind-nailed it through the splines into in the shaper. still looks tight.
the walls. Putting the wainscot up in two sec- I wanted the under the front of the
tions, bottom and top, eliminated the wood- counter and at the top of the head casing to
shrinkage gaps that would have resulted from match the original molding at the top of the
running the wainscot from floor to ceiling and doors and the windows (photo above). This
butting the counter into the wainscot. With the molding wasn't something I could have picked
front edges of the counter notched, I installed up at the lumberyard, and I couldn't find any cut-
the front pieces of wainscot on each sidewall. 3. Below the bead, waste ters the right shape, so I combined two different
Because the unit is recessed into the opening, I was ripped on the table shaper cutters to make the molding (drawing
saw in. deeper than left). The result was a perfect match.
wanted a full length of wainscot from floor to the flat above the bead.
header with no seam.
Making doors—I made frame-and-panel doors
Pocket-screw joinery—ln my shop, stiles and for the cabinet at the bottom of the hutch.
rails for the face frame were cut to width but not The door stiles are 2 in., the top rail is 2½ in.,
to length. Stiles and rails are the vertical and hor- and the bottom rail is 3½ in. After cutting the
izontal frame pieces, respectively. pieces to size, I used my shaper to mold the in-
I assembled the face frames on site. I cut the 4. Then the shaper was side edges of the frame, cut the panel groove and
used to cut a 1/4-in, flute. make the cope-and-stick jont between the stiles
stiles and the rails to length and clamped them to
the cabinet to check the fit. After some slight trim- and the rails.
ming on a compound-miter saw perfected the I assembled the frames dry to check the door
face-frame joints, I laid the stiles and the rails on size and to get the panel size. I allowed 1/8 in. on
the bench and screwed them together. each side of the panel for expansion. The ash
The top rail was narrow enough to allow the panels on these doors were raised (beveled
stiles to be joined to it with screws driven straight around the edges) on the shaper, so I glued up
through the edge. But the bottom rail of the cab- 5. Another pass on the boards with square joints to make the wide
the table saw panels. If I had used the glue-joint cutter, the glue-
inet was wide, and the intermediate stiles butted removed in. of
into it, so here I screwed the rail to the stiles waste below the flute. joint profile would have been visible when I
through pocket holes. A pocket hole is a cut shaped the raised edges.
made on the face of a board that doesn't reach To be sure everything fit, I dry fit the panel with-
the board's edge. in the frame before gluing up. Then I glued the
There are several jigs on the market to make doors and clamped them. I used a small amount
pocket holes—from simple guides for a hand-held of glue on the joints because the squeeze-out
drill to dedicated pocket-hole machines. I don't could glue the panel in place, and the panel
have any of them, so to make the pocket holes to should be free to expand and contract. E
6. A ½-in, flute cutter on
assemble this face frame, I used a spade bit, start- the shaper cut a cove to
ing the hole with the drill held vertically and tip- complete the profile. Stephen Winchester is a carpenter and wood-
ping the drill back as I fed the drill bit in (left worker in Gilrnanton, N. H. Photos by Rich Ziegner
photo, facing page). The pocket hole ended at a Cove except where noted.

Finish Carpentry 107

Photo this page: Stephen Winchester I I
 ______

Simple Closet Wardrobe

A biscuit joiner and a layout jig make construction quick and easy

by Jim Tolpin

loset storage systems are in vogue these

38 in.
days. Homeowners are buying prefabricated
in. drawer units and paying closet specialists to in-
Hat shelf stall expensive epoxy-coated wire bins and
shelves. Yet a handsome closet wardrobe can be
constructed in just about a day by anyone who
learns how to use a biscuit joiner and a shop-
made layout jig.
A simple layout technique makes the process
foolproof. The same approach is also effective
Length to suit closet for many other kinds of case pieces like book-
H shelves and cabinets. Plus, all of the work can
be done on the job site with just a few tools and a
73 in. Shelf standards workbench no more elaborate than a flat surface
on a pair of sawhorses.
This surprisingly compact wardrobe unit (pho-
to facing page) features a bank of slide-out
Adjustable shelves sweater drawers, adjustable shelves and a hang-
ing locker with a hat shelf. A typical bedroom
closet is 2 ft. deep and 5 ft. or 6 ft. wide, so I
make my wardrobes 16 in. deep and devote half
A site-built the width of the closet to the drawers and the
wardrobe. shelves. The size of the accompanying hat shelf
Adjustable shelves,
drawers and a hat and closet pole can be adjusted to fill the rest of
N shelf make the the space.
N wardrobe versatile. The project can be built from either 3/4-in.
The base and the hat
shelf are made boards or 3/4-in, sheet stock like hardwood ply-
separately, with the wood or melamine (particleboard covered with
3 in. hat shelf sized to the a thin plastic laminate). Some suppliers stock
width of the closet.
The mitered base melamine that is precut to 16-in, widths with a
frame should be thin strip of the same laminate applied along one
nailed and glued edge. This gives the piece a finished look.
Drawer bins on full together with glue
extension slides blocks added for
rigidity. Maldng the base frame—The wardrobe is es-
sentially a tall box with a vertical divider down
the center. One side is devoted to drawers and a
fixed upper shelf, and the other to adjustable
shelves. The case sits on a 3-in, high base (draw-
ings left), with a I 1/2-in, toe space in the front and
a I Y2-in. overhang on each side. The base-frame
height and the rear inset can be adjusted to clear
any existing baseboard.
Start with the base, cutting the frame compo-
3 in.
nents to length and width for whatever size
wardrobe you've decided to build. I rip sheet
stock for the base, but you could also use lx4s or
even 2x4s.
Mitered corners Miter joints, which I use on the front corners
Glue of the base frame, may be made with an electric
blocks
miter saw or a handsaw and a miter box. After
the base pieces are cut to size, hold the front and
Base frame
back pieces together and mark on their top
edges the location of the interior spacer at the
middle of the base.

108 The Best of Fine Homebuilding

 Finish nails and glue are adequate for assem-
bling the cabinet base. You also may install glue
blocks on the inside corners for additional
strength. Biscuit joinery isn't necessary for the
base because it is not subjected to much stress
once the cabinet is installed, and using the joiner
may actually take longer. But if you do use a bis-
cuit joiner, be careful. The base pieces aren't
wide enough to engage the anti-kickback pins
on the faceplate of the joiner, so the tool may
slip while cutting slots (for more on biscuit join-
ery, see FHB #70, pp. 50-53).

Cutting the components—Cut the case sides,

the top and bottom pieces and the interior parti-
tion to length and width. Be sure the crosscuts
are perfectly square, or the case may rack when
assembled. I don't put a back on the case, so I
use a cleat to help stiffen the case and to give
me a way of attaching it to the wall of the closet.
Cut a notch in the upper back corner of the cen-
ter partition for a 3/4-in, by 5½-in, cleat.
Shelf standards on the right-hand side of the
case may be surface-mounted later. Or you can
use a router and an edge guide at this point to
cut shallow dadoes and let the shelf standards
into the side of the case. But take care to stop
the dadoes before they reach the top edge; oth-
erwise, the ends of the dadoes will be visible af-
ter the case is assembled.
Lay out the location of the center partition on
the top and bottom components. Also mark the
location of the fixed shelf on the edge of the left
side of the case and on the partition. Remember,
however, the partition and the side pieces are
unequal in height, which must be taken into ac-
count when marking the position of the fixed
shelf. Also mark where top and bottom pieces
will intersect the sides.
Before beginning the slotting process, dry as-
semble the case on a
flat surface. You may Marking for
need to apply light assembly
pressure with pipe
clamps to stabilize
the structure. When all
case components are
aligned to their marks,
check to be sure ev-
erything has been cut
to the proper length
and that the case is
square. Lightly mark
the case pieces so that
you can assemble
them quickly and cor-
rectly later (drawing
right). I chalk an "X"
on the faces of the components to help locate
the slotting jig and to orient pieces to the bench
top during the end-slotting process.

A simple and effective Jig—The case is held

together by 32 biscuits, which means that 64 slots
had to be cut, and they had to line up with each
other. To cut the slots quickly and accurately, I
built a jig (drawing, p. 111) that locates the slots Closet wardrobe made easy. A
and helps orient and hold the biscuit joiner to space-saving closet wardrobe fast and
the work. Positioned like a try square, the jig can centive for finish carpenters.
be used from either the front edge or the back
edge of the stock. Making the drawers.
Drawers are made with the biscuit joiner and the jig shown on
The entire jig can be made from 3/4-jfl, hard- the facing page, with ¼-in, plywood bottoms and applied drawer
wood plywood. It consists of an arm that is sand- faces of solid/umber. Make drawers 1-in narrower than the
wiched between two crosspieces at a 900 angle. opening to leave room for the drawer slides.
The crosspieces make up the base of the jig. The fl
length of the arm extending beyond the edge 15 in. stock
of the base is the same as the width of the mate-
rial you are joining—in this case 16 in. Once
you have cut out the pieces for the jig, align them
to a framing square and assemble them with
glue and screws. But jigs can be made to any
width and marked where you want to cut the
slots. For the case of the wardrobe, I marked four
spaced slots for each joint.
Once the jig has been made, cut the slots in
the side pieces of the case. These slots will re-
ceive the biscuits that are inserted into the ends
of the top and bottom pieces. Position the jig
back from the end of one of the case sides by
the thickness of the stock and clamp the jig in
place (top photo, facing page).
The centerline of the joiner's base should be
aligned with one of the marks on the arm that in-
dicates the center of a slot. Because the joiner is
being used near the end ut the workpiece, a
scrap of case stock placed under the tool's face
will add support and assure that the slot is per-
pendicular to the face of the work. This also is
the time to slot the inside faces of the top and
bottom pieces to receive biscuits that are insert- of the glue, making adjustments impossible. If Making the drawers—The drawer components
ed in the ends of the partition. Position the jig to you are working alone, consider assembling this are cut, marked and joined just like the case
the X side of the mark you made to locate the project in two stages. was assembled (drawing above). Most draw-
partition on the top and bottom pieces and make The first stage is joining the center partition er hardware requires a '/2-in, space between
the slots. Slot the fixed shelf the same way. to the top and bottom pieces. Apply glue to the the side of the drawer and the inside face of the
Biscuit joinery gives you some latitude in the slots (I use yellow glue), preferably with a special case. Therefore, if the cabinet opening is 16 in.,
placement of the slots. In side-to-side alignment, applicator tip on your glue bottle to keep the the drawers should be 15 in. wide. But check the
discrepancies of up to '/8 in. will not affect the process mess-free. Then insert the biscuits and specifications of the hardware you buy to see
outcome of the joint. But accurate vertical align- pull this I-shaped assembly together with pipe how much clearance the slides require. The
ment is critical. Work away from the chip dis- clamps. Be sure the components stay perfectly drawer fronts are cut nearly as wide as the open-
charge of your biscuit joiner; otherwise, you will square to one another. Don't move this assem- ing, which keeps the slide hardware out of sight
have to clear the work surface each time you bly until the glue is completely dry (at least a when the drawers are closed. I use full-extension
reposition the tool for the next slot. couple of hours). slides because they allow easy access to the en-
The second stage is joining the rest of the com- tire depth of the drawer.
Cutting end slots—Use the same jig to position ponents to the first-stage assembly. Assemble the Cut the drawer parts to size (wait until later to
the joiner for slotting into the ends of the top, the case on a flat surface that can support the entire cut the faces). Then dry assemble the drawers
bottom, the partition and the fixed shelf. For this structure. Orient the marked front edges upward, and mark them for location and orientation. You
operation, align the jig flush with the end of the making sure all the components are in their prop- can use the same positioning jig to cut the slots
piece to be slotted and clamp it securely in place er positions. Then apply the glue, insert the bis- for the drawers, but you will have to make new
(bottom photo, facing page). Orient the side of cuits and clamp the components together. Be slot-reference marks because the drawer pieces
the piece marked with an X during dry assembly sure the front edge of the boards are flush and are not as wide as the case. Cover the arm of the
face down on the workbench and make sure take diagonal measurements to the outside cor- jig with masking tape, then mark the locations
the piece lies flat. The presence of chips or other ners of the case to check for square. for three biscuit slots on the tape. Use the jig to
foreign objects between the work and the bench, Instead of clamps, 1%-in, drywall screws may make slots in the face and the ends of the pieces.
or warps in the bench itself, can easily misalign be used to hold the case together while the glue Cut a ¼-in, by ¼-in, groove along the bottom
the end slots. sets. Screw holes made through the sides of the inside face of the parts to receive the ¼-in. ply-
To slot the end of the stock, place the joiner case should be counterbored and then plugged. wood bottoms. Then glue and insert the biscuits,
flat on the work surface and orient the tool's cen- Or screws can be capped with plastic covers. run some glue inside the groove and assemble
terline to one of the marks on the jig's cross arm. These plastic covers are available through most the sides around the bottom panel. Drawer bot-
The base of the jig provides surface area for one mail-order hardware suppliers. toms that fit snugly will help keep the drawer
of the joiner's anti-kickback pins, helping to hold When the glue has dried, remove the clamps. square when it is clamped. Apply band or pipe
the joiner in place. Again, work away from the Flip the case over on its face edge, then glue and clamps, wipe off excess glue and check each
expulsion of saw chips. screw the cleat to the notch in the partition. The drawer for square with diagonal measurements
ends of the cleat will butt against the inside faces before setting it aside to dry.
Assemble the case—When assembling a bis- of the case, where they are glued and nailed us-
cuit-joined case of this size and complexity, you ing finish nails. For additional strength, install Installing the hardware—Next, put the case
must move quickly through the procedure. If the several screws through the top into the cleat, on one side and lay out and install the shelf stan-
process goes on for more than about 15 minutes, counterboring and plugging the holes or capping dards and the drawer slides. Flip the case over
some of the biscuits may swell from absorption the screws. and install the hardware on the opposite side.

110 The Best of Fine Homebuilding

Secure the drawer hardware to the case sides us-
ing only the slotted holes; this will allow final ad-
justment later. A tick stick, a length of scrap lum-
ber marked to show the location of each drawer
slide, will help you lay out both sides of the draw-
% in.
er bay quickly and evenly.
Now it is time to install slides on the drawers.
Cut the drawer faces and cut the semicircular
hand grips in the top edges. Screw the faces to
the drawers from the inside. If you use plywood
for the drawer fronts, you will have to band the
edges with a thin strip of solid material or veneer. 51/2
Test the drawers by sliding them into the case on
their slides. Adjust the gaps between the drawer
faces by moving the slide hardware on the
case sides up or down. When you are satisfied
that the spaces between the drawers are even,
secure the slides by adding screws to the round
holes. Then cut the adjustable shelves to size and
tiy them on the standards.

Installing the wardrobe—Set the base frame

assembly, still unattached to the case, into posi- Biscuit-joiner jig.
tion on the floor of the closet. Carefully level the This shop-made jig makes reliable biscuit joinery a speedy process. It can be used
from either the right or the left and includes reference marks that line up with the
frame from side to side and from front to back. centerline on the biscuit joiner.
Shim where necessary between the bottom of
the frame and the floor. Screw the leveled frame Cutting side slots. With
to the floor by running 2-in, drywall screws from the jig lined up at the mark
inside the frame down at an angle. Locate the on the edge of the workpiece
screws so that they will pass through the shims, and clamped in place, the
biscuit joiner cuts slots accu-
but be careful that screws don't break through rately. Note the scrap piece
the face of the base. Score and snap off the shims of case stock beneath the
(or saw them off with a veneer saw) where they tool for added support. Here,
protrude from the frame. the author uses a narrower
jig than the one shown above
Remove the drawers and the adjustable shelves to make an l1'/4-in. deep cab-
from the case to lighten the load, then set the inet. Jigs can be made to suit
unit on its base frame. Be sure the overhang of any cabinet width.
the case is equal side to side and parallel to the
face of the base frame. Then join the base and Cutting end slots. To cut
the case together. Finish nails will hold the case the biscuit slots to the ends
of the top and bottom pieces,
permanently in place. the jig is positioned flush
Fasten the wardrobe to the wall by running with the end of the work-
2V2-in. drywall screws through the back cleat into piece. After the piece is
wall studs. But make sure the face of the ward- clamped, make sure it Is per-
robe is square before securing it to the wall. Be- fectly flat so that the slots
will be aligned correctly.
cause the case doesn't have a full back, it could Then line up the joiner's cen-
rack during installation if you're not careful. If terline with marks on the jig
the wall is leaning back out of plumb, slide shims and cut the slots.
between the cleat and the wall surface first.
The hat shelf, with closet pole below, is built
to suit the closet width and is best tended to
after the rest of the wardrobe is in place. Cut
the two shelf cleats, then drill holes in them
to accept the ends of the 15/8-in, closet pole. In
locating the holes on the cleats, make sure a
hanger will clear the shelf and the back wall. Or
you can use surface-mounted rosettes for the
pole, which are commonly available from build-
ing-supply stores.
Simply screw one cleat to the wall and another
to the side of the cabinet with the pole in be-
tween (cut the pole about in. short to make its
installation easier). Then cut the shelf to fit the
opening, scribe the back and the far edge to
the wall and drop it into place.

Jim To/pin is a woodworker and writer living in

Port Townsend, Wash. Photos and drawings by
the author.

Finish Carpentry 111

 Bed Alcove
Convert wasted attic space into a bed that has drawers, bookshelves and a vanity

By Tony Simmonds

hen the middle one of my three daughters would face north and therefore would not be Site-built cabinet—The underframe of the bed
grew too old for the loft bed I built for her, the subject to heat-gain problems; second, it would is a large, deep drawer cabinet. You could have it
youngest, Genevieve, was happy to inherit it. The illuminate the shoji from behind; and third, there built by a custom shop while you get on with
loft is in a small bedroom on the second floor of was the emotional pressure from my client— framing, wiring and drywalling. Custom cabinets
our house in Vancouver, B. C., Canada. Like some drivel about the stars and the treetops and are expensive, though, and after nearly 10 years
many second floors of old houses, this one is re- falling asleep to the sound of rain on the glass. in the business of building them, I appreciate the
ally a half stoly, with sloped ceilings where the virtues of their old-fashioned predecessor, the
rafters cut across the intersection of wall and Tight layout—Juggling existing conditions is the model A, site-built version. It's economical in
roof. The bedroom has only about 80 sq. ft., so its challenge of remodeling. None can be consid- terms of material and expense, and you can usu-
bed had to be on a raised platform to leave ered in isolation. For example, I had to decide ally get a closer fit to the available space.
space for a dresser and a desk below. whether or not to keep the existing 7-in, high The partitions supporting my daughter's bed
Soon after she moved into the loft, however, baseboard. I could have moved it, but I wanted are made from %-in. plywood sheathing left over
Genevieve started bumping her head on the ceil- to leave it in place, partly for continuity and part- from a framing job (the rewards of parsimony).
ing over the bed. When she eventually moved ly to avoid as much refinishing as possible. Each partition is made from three layers of
the mattress to the floor, I knew it was time for Starting the drawers above the baseboard also sheathing (drawing below). The center layer
the old bed to go and for a new one to take its meant that the baseboard heater already on the runs the full height of the partition, but the outer
place. The bed alcove shown in the photo on adjoining wall wouldn't have to be moved to pro- ones are cut in two, with the drawer guide sand-
the facing page was the result. vide clearance for the end drawer. wiched between the top and bottom pieces. The
Four drawers fit into the space between the guide is simply a piece of smooth, fairly hard
Will it fit?—The kneewalls that defined the sides baseboard and the mattress platform. The draw- wood, 3/4 in. thick and wide enough so that it pro-
of the room had originally been a little over 6 ft. ers are 7 in. deep (6'/2 in. inside), which is ample jects in. into the drawer space.
high, leaving a great deal of wasted space be- for all but the bulkiest items. This brings the mat- Unless circumstances demand the use of me-
hind them. I proposed to recover this space by tress platform to a height of about 18 in. With a chanical drawer slides, I prefer to hang drawers
moving the kneewall over 4 ft. to accommodate 4-in, thick mattress on top of it, the bed still ends on wooden guides. I have provoked derision
a 3-ft. wide mattress and a bedside shelf beyond up at a comfortable sitting height. from cabinetmakers because I use wooden
that. Given the 12-in-12 pitch of the roof, this In plan, the mattress takes up almost exactly guides in kitchens, but when it comes to bed-
would bring the ceiling down below 3 ft. at the three-quarters of the 9-ft. long space. The leftover rooms I am almost inflexible. Even large draw-
new kneewall. Would this be claustrophobic? To corner accommodates a makeup table with mir- ers like these will run smoothly year after year
answer the question, I mocked up the space with ror above and more drawers below. I imagined if they are properly fitted and if the guides are
packing crates and plywood to make sure there that the shojis would draw a discreet curtain over securely mounted. And for me there is a subtle
would be room to sit up in bed. A high ceiling is the wreckage of eyeliners, lipsticks, mousse and but important difference between the sound
not a necessity over a bed—within reason, the re- everything else that was supposed to go in the and the feel of wood on wood vs. even the finest
verse is true: A lower ceiling increases the sense drawers but never would. ball bearings.
of shelter and enhances the cavelike quality hu- I knew that this vanity area, and especially the I attach the guides with screws rather than with
mans have always favored. Furthermore, a bed mirror, would need to be lit, but beyond making glue and nails so that they can be removed,
in an alcove that can be closed off from the rest sure there was a wire up there somewhere, I planed and even replaced without difficulty
of the room has qualities of privacy and quiet didn't work out the details during the prelimi- should the need arise. A groove in the partition to
that are difficult to achieve in any other way. To nary planning. I was in my fast-track frame of house them is not necessary, but it's a way of en-
get that extra layer of privacy, Genevieve and I mind at this stage of the project. suring that they all end up straight and exactly
decided that her bed alcove should have four where you want them.
sliding shoji screens. For this job, the pairs of guides on the three
The 9-ft. length of the space would provide Partition middle partitions had to be screwed to one an-
room for a dresser and a vanity of some sort, as
detail other, right through the core plywood. I drilled
well as the bed. Drawers underneath the plat- and counterbored all the screws and clamped
form would triple the existing storage space. the partition to my workbench to make sure ev-
Light and ventilation would come from an oper- erything stayed tight while I drove the screws.
able skylight over the bed. Then, with the partition still on the bench and
I had some misgivings about the location of after inspecting every screw head carefully for
this skylight in spite of the obvious benefits it depth below the surface, I set the power plane
would confer in terms of light and space. Having for the lightest possible cut and made three pass-
never slept directly under one myself, I didn't es over each guide: first over the back third only,
know whether a skylight so close to a bed would then over the back two-thirds and, finally, over
make sleep difficult. But in the end I was se- the whole length of the guide. Tapering the
duced by three arguments. First, the skylight guides so that they are a fraction farther apart in

112 The Best of Fine Homebuilding Drawings: Bob La F'ointe

Tight fit. Into this 9-ft. long space, the author squeezed a single bed, a by a lamp benmu die mirror. the uoseuoard reveals me line of the origi-
row of 30-in, deep drawers, a bookshelf and a vanity. A recessed fluores- nal wall. Above drawer fronts cut from a single lxlO are screwed from
cent fixture illuminates the mirror from above while the vanity table is lit behind to the drawers. Photo by Charles Miller.

the back allows the drawer to let go, rather than the front edge of the partition and secured it with The bed slats also act as ties to link all the par-
tighten up, as it slides home. a second screw near the bottom of the base- titions together (bottom left photo, p. 114). 1 used
board. With the front edges located and the par- dry 1x6 shelving pine for the slats, but almost
Partition alignment—Installing the partitions titions standing straight, the next job was to align anything that will span the distance between sup-
is the tiickiest part of a site-built cabinet job like them to create parallel, square openings. ports will do. I left an inch between the slats to
this one. I said earlier that you could save on ma- I built the new kneewalls 483A in. back the keep the mattress well aired. I learned this the
terials by building the cabinet in place, but you inside face of the baseboard. This hard way when an early bed I built on a solid
can't save on time. After all, anyone with a table a couple of 1x4 straps horizontally across plywood platform developed mildew on the un-
saw cain build a square cabinet in the shop, but the studs to provide anchoring surfaces for the derside of the mattress cover.
building one accurately in a closet or in an un- 48-in, partitions (drawing next page).
finished space under the rafters takes patience To align the partitions, I used hard board chit to Fifing the drawers—Before putting anything
and thoroughness. The key to success is to es- the full opening width (top left photo, p. 114). As on top of the platform, I built and fitted the draw-
tablish a datum line, then lay out everything from long as the hardboard is cut square, and the par- ers. The drawers have 1/8-in. clearance between
this line, leaving the wedges of leftover space titions are secured so that the hardboar4l fits their sides and the partitions. The 3/8-in. projec-
around the perimeters to be shimmed, trimmed, snugly between them, the resulting open will tion of the drawer guide thus creates a ¼-in, in-
fudged and covered up as necessary. also be square. I used screws to fasten the ply- terlock with the sides. All the drawers are 30 in.
In Genevieve's room, the existing baseboard wood flanges that held my partitions in deep, but I let the sides extend 6 in. beyond the
provided a datum line in both horizontal and just in case adjustment should be necessary. back of the drawer. The extensions support a
vertical planes. First, I divided the baseboard's When all the partitions were in place, 1 cut drawer right up to the point where its back
length so that the four drawer fronts would lie di- pieces of 1x2 to the exact dimension between comes into view.
rectly below the shoji screens. I ran one screw each pair of drawer guides. Centered on the If time and budget allow, I use a router jig to
into each supporting partition, about 1 in. below drawer fronts, the lx2s are gauges that how dovetail the front of a drawer to its sides, but the
the top edge of the baseboard. Then I plumbed deep the grooves need to be in the drawer back just has tongues cut on each end that are

Finish Carpentry 113

Bed-alcove anatomy
Pine shelving and plywood
corn pose the framework of this
bed.

Aligning partitions. Load-bearing parlitions

made of three layers of 3/8-in, plywood sepa-
rate the drawer bays under the bed and sup-
port the mattress platform. The drawer guides
are sandwiched between the outer layers of
plywood. The photo at left shows the hard-
board panels that helped to align the parti-
dons. Once the panels were in place, the parti-
lions were screwed first to the baseboard and
then to strapping along the stud wall. The lx2s
clamped to the leading edges of the panels are
gauges that will be used to determine the
depth of the grooves in the drawer sides.
Linked by slat. The partitions are lied to one
another across their tops by 1x6 pine slats (bot-
tom left photo). Spaces between the slats pro-
vide ventilation for the mattress. At the right
side, the carcase for the vanity drawers sits di-
rectly atop the bottom drawer partitions.
Bookcase wall. Shelves deep enough for pa-
perbacks are affixed to a 3/4-in, birch plywood
panel between the bed and the vanity (photo
below). The squares at the end of each shelf
frame rice-paper windows that are backlit by
bulbs behind the vanity drawers.

114 The Best of Fine Homehuilding

 tablish just how big it should be—
glued and nailed into dadoes in
the sides (I take care not to put any
the trade-offs being the height of
nails where the groove for the the fixture, the size of the shade
and its proximity to errant elbows. I
guides will be plowed out). The
drawer bottom rides freely in a thought I had a satisfactory bal-
groove cut in the front and the sides ance, so I went ahead and made
the lamp. But Genevieve put her el-
and is nailed into the bottom edge
of the back, which is only as wide bow through it the first night she
as the inside height of the drawer. slept in the bed. I forgave her and
Fastening the bottom here helps to accepted the lesson. The second
reading light ended up above the
keep the drawer square.
Fitting the drawers should present head of the bed (photo left).
few problems if they are built square
and true and if time and care have What aboul the shojis?—The
been invested in positioning the par- shoji screens have yet to be made,
and it now seems unlikely they ever
titions. Don't try for too tight a fit, es-
pecially in the width of the groove. will be. Although she was initially
keen to have them, Genevieve now
My guides were 3¼-in. material, and I
believes they would get in the way,
plowed out a 3A6-in. dado in the
and I agree with her. We analyzed
drawer side. They're not sloppy.
the patterns of opening and closing
On the other hand, you should be
that might be required during a typ-
more stingy about the depth of the
grooves. Remember, the guides ical day and night. It became clear
have been planed to allow increas- that in spite of the desirability of
drawing a curtain over the unmade
ing clearance as the drawer slides
home. Too much slop here can bed by day and the unfinished
Headboard. A reading light inspired by the luare-and-tri- homework by night, this teenager
cause the drawer to bang about left, wooden
angle motif lights up the headboard side of the bed. On would rather live and sleep in one
from side to side and actually hang tracks for shoji screens frame the alcove. room—at least for the time being—
up on the diagonal. You can always
than be bothered sliding screens to-
plow a groove out a little deeper. A
router with a fence or a guide attached is the narrow strip of mirror along the bottom edg€ of and-fro all the time. A feeling of confinement was
the large mirror, angled upward to bounce also a factor. Having tried out the bed myself one
ideal tool for this because you can easily make
very small adjustments. If things go wrong, you light where it can fill in shadows. night when she was sleeping at a friend's house, I
can glue a length of wood veneer tape into the too felt I might want more distance between my-
dado, but it's nicer not to have to do that. Bookcase wall—As for the partition beti self and any enclosing screen.
I dress the groove with paraffin wax, but only dressing table and bed, my fast-track convi I admit that this was something of a blow to my
when I'm sure the drawer doesn't bind. Patience that it could not be frame and drywall he vision of the room. What about the function of
in working toward a fit has its reward here. The better. My daughter wanted more bookste the skylight as a for the shoji? What
and the foot of the bed was a logical place about the square-and-triangle motif I was going to
moment that a wood drawer on wood guides just
slides into its opening and fetches up against its them (photo facing page, lower right). I incorporate into the shoji lattice? Ah, well, at
stop, expelling a little puff of air from the cabinet, the back of the bookcase out of 3/4-in. bircn least I hadn't made them already. And the
is a moment that provides much satisfaction. wood, which could be finished naturally on grooves in the bottom track appear to work per-
book side and painted white on the dress fectly as 9-ft. long pencil trays.
Beyond the footboard—With the drawers and table side to look like a wall. The rejected shojis and the difficulties I had
the platform in, I had to decide what to do about To light the makeup table, I mounted a stan- with the makeup light and the height of the mir-
the divider between the bed and the vanity. Here dard incandescent ceiling fixture in the space ror were all results of my decision to keep the
was where the self-imposed constraint of using behind the mirror. On a playful impulse, I wired bed alcove within the area beyond the existing
the existing baseboard as the perimeter of the al- another of these lower on the sloped ceiling in kneewall. If I had moved this line 6 in. to 12 in.
cove began to bite. Because its height was deter- the space behind the vanity drawer (the case for into the room, I could have raised the upper sho-
mined by the slope of the ceiling, the mirror over these has no back, so the fixture is easily acces- ji track a few inches, creating plenty of space to
the dressing table had to be as far forward as pos- sible). Then, after carefully laying out the lqca- mount the mirror light, the reading light and the
sible. But to bring it right up against the inside tion of the bookshelf dividers and following a shoji screens. The amount by which this would
edge of the upper shoji track would eliminate square-and-triangle motif suggested by the don- have reduced the size of the room would have
the space required for a light above the mirror. junction of the ceiling and the shelves, I jigsawed been insignificant in relation to the space gained
And even that would put the top of the mirror at the holes in the birch ply and glued rice paper by building in the bed and the dressing table—a
barely 6 ft. Temporarily derailed on the fast track, over them. This created little backlit rice-paper case of choosing the wrong existing condition
I tried to find other ways to light the mirror and windows in the bookshelves. The dividers cèver to work Irom.
kept coming back to the necessity of recessing a the edges of the paper. The only slight snag in At least the client is satisfied. The project was
fluorescent fixture into the ceiling. this assembly is that the plywood thickness caus- completed during one of the long dry spells that
The fixture I used is a standard T-112 fluorescent es a shadow line, which can be seen where the Vancouver is famous for. Finally, one morning
fixture equipped with an liltralume lamp (Philips backlighting travels at an angle through the win- when the spider webs were glittering and the
Lighting Co., 200 Franklin Square Dr., Somerset, dow. If I'd thought of it in time, I could have easi- earth smelled refreshed and autumnal, Gene-
N. J. 08875; 908-563-3000). The lamp emits more ly eliminated the shadows by beveling these vieve appeared downstairs with a beatific smile
lumens per watt than a standard cool-white lamp edges with a router. on her face. "It rained on my skylight last night,"
and has a higher Color Rendering Index, both The bedside reading light was more of a prob- shesaid. E
important factors in getting an accurate reading lem, Initially, I placed my standard ceiling
on colors, like those at a makeup table. under the skylight as far down the slope of the Tony Simmonds is a designer and builder in
Casting an even light across the face of the per- ceiling as I could. I made a cardboard moqk-up Vancouver, B. C'., Canada. Photos by the author
son standing at the mirror is important. So I put a of the rice-paper shade that I had in mind to es- except where noted.

Finish Carpentry 115

 Bookshelves can transform a room Into
a library. shelves provide plenty of eas-
fly accessible room for books and collectibles.
Glass-panel and raised-panel doors provide
more secure shelf space, and related supplies
by Bruce Greenlaw can be stored In lockable drawers.

t's hard to define the quintessential bookshelf. The one above my writing Between these two extremes lie a wide variety of shelving options. Basi-
desk, for example—a plastic-laminated particleboard shelf supported by cally, though, bookshelves are either housed in bookcases or supported by
three inexpensive metal wall brackets—was quick to build and perfectly wall brackets (unless they're propped on milk crates or cinder blocks),
suits my needs. The fixed shelf puts my reference books within arm's and they're either fixed or adjustable. To my mind, the best shelving sys-
reach of my chair, and foam-padded steel bookends that I got at Wal-Mart tems complement their surroundings and don't droop when they're
keep them from falling off. Thos. Moser Cabinetmakers' furniture-grade loaded with books.
cherty bookcases (Thos. Moser Cabinetmakers, P. 0. Box 1237, Auburn, This article contains food for thought on designing bookshelves, plus
Maine 04211; 800-862-1973) (photo above), on the other hand, are de- an appraisal of shelving materials and hardware (sidebar p. 119). Of
signed to be heirlooms. course, this information can be applied to virtually any type of shelving.

116 The Best of Fine Homebuilding Photo this page: courtesy of Thou. Moser Cabinetmakers
 SIZES, STANDARDS AND BRACKETS

I
I Shelf pins are in-
conspicuous. Shelf
pins are less visible
than shelf standards.
Top to bottom: wire
clip, which hides inside
a groove at the end of
a shelf; spring-loaded
locking clip that pre-
vents shelves from slid-
ing or tipping; plastic
clip; cushioned metal
L-support; ornamental
Bookshelves can be quickly and easily installed on wall brack- solid-brass pin; zinc-
eta. Fixed brackets are screwed directly to the wall; adjustable brackets plated steel paddle; "li-
mount on slotted standards and can be moved up or down. brary" pin with sleeve.

Measure the books, and size the shelves—Standard paperback books Melal standards are
are about 4 in. wide (from the spine to the outside edge), but my binders,
strsng but conspic-
uous. Although metal
portfolios and biggest reference books are about 11 in. wide. Unless shelf standards come
shelves will be used for storing old LP records (which are 121/4 in. wide) or with various disguising
giant art books, you'll rarely need to make bookshelves that are more finishes, they tend to
than 11 in. deep. Some shops make 8-in, to 10-in, deep shelves and let be more visible than
hole-mounted shelf
wide books overhang. supports. Most stan-
If a series of shelves will be fixed in a bookcase or on nonadjustable dards can be recessed
wall brackets, measure the heights of the books that will be stored on (at right) to make a fin-
them and space the shelves accordingly. Books range in height from about ished appearance and
in. tall for standard paperbacks to well over 12 in. tall, but most are in to minimize gaps at
shelf ends, or they can
the 8 in. to 12 in. range. Don't forget to add 3/4 in. to I in. of clearance be surface-mounted.
above the books to allow fingers to grip them.

Shelf standards and hole-mounted clips make bookcases ad-

justable—Unless they're required for structural integrity, fixed shelves are
probably unnecessary in bookcases. Adjustable shelves take much of the
worry out of planning, offer more long-term flexibility and often result in
more shelves fitting into a given space. The joinery for a bookcase without
fixed shelves is simplified because you're basically just building a big box.
Also, the supporting hardware for adjustable shelves is relatively inex-
pensive and easy to install.
Most shops don't use mn-of-the-mill metal shelf standards to support ex-
posed shelving. But metal standards are stronger than most alternatives,
and I think they look okay if their color complements the surrounding spaced about 1 in. o 2'/2 in. o. c., and they're positioned so that shelves
bookcase. The standards are screwed or nailed in pairs to bookcase sides. overhang the pins ajbout I in. to 2 in. front and back.
They can be mounted on the surface, but most also can be recessed into Drilling shelf-pin holes can be tedious work, but a drilling template
dadoes for a more refined look (bottom photo). Barbed plastic standards makes the job easier. A simple one can be made from a piece of tem-
that are simply pressed into dadoes (no nails or screws are required) also pered pegboard, has ¼-in. dia. holes spaced I in. o. c. (top photo,
are available. But I've heard that these standards can be difficult to align, p. 118). Using a drillibit with a drill stop mounted on it speeds the work and
resulting in wobbly shelves. prevents the bit froijn boring through the workpiece.
Shelf pins are generally regarded as a step up from shelf standards (pho- Pegboard jigs don1t last long, though. Commercial jigs are a better choice
to top right). These pins are strong enough for most applications, and for production wo4. The acrylic shelf jig sold by The Woodworkers' Store
they're less conspicuous than shelf standards. Shelf pins typically have (bottom photo, p. has oversize holes that guide a self-centering '/4-in.
'¼-in. dia. shanks that fit into holes of the same diameter. The holes are Vix bit (or a 5mm Vix bit for metric work). The jig only 19 in. long, but it

Finish Carpentry 117

Templates simplify boring of shelf-pin
holes. Drilling jigs make it easy to bore shelf-
pin holes in parallel rows in bookcase sides.
This disposable shop-made jig is tempered-
hardboard pegboard, which has '/4-In. dia.
holes spaced 1 in. o. c.

Commercial hole-drilling Jigs are du-

rable. The acrylic jig shown here (sold by The
Woodworkers' Store; see sidebar, facing page)
has oversize holes that guide a self-centering
Vix bit. A shelf pin is used as an indexing pin
to extend the range of the jig.

can be indexed with a shelf pin to bore any number of holes accurately. into holes bored into the sides of bookcases, and they hide inside sawkerfs
The jig costs $16, but the Vix bit adds another $35. Another shelf-drilling jig cut into the ends of shelves.
is made by Veritas tools and is available through mail-order outlets (Veritas
Tools, 12 E. River St , Ogderisburg, N.Y. 13669; 800-667-2986). Wall-mounted shelf brackets can be fixed or adjustable—Truth is,
Woodworking and hardware suppliers sell a variety of shelf pins and most of the bookshelves I've put up sit on plain metal wall brackets. The
supports (photo top right, p. 117). Trussed plastic shelf pins are the least brackets install quickly with hollow-wall anchors or by screwing them di-
expensive and the most obtrusive. Metal L-supports are stronger and more rectly to studs. Deluxe shelf brackets that install just as easily also are avail-
subtle than plastic pins. Padded L-supports cushion glass shelves and help able. Two examples are the brass brackets sold by Renovator's Supply
protect fragile finishes. Locking pins made of plastic or metal help prevent and the oak gingerbread brackets sold by The Woodworkers' Store (photo
shelves from sliding and tipping. They allow bookcases to be shipped top left, p. 117).
with shelves installed, and they help anchor shelves in earthquake country. For adjustable wall support, most hardware stores sell single-slotted met-
To my eye, the best-looking shelf supports on the market are spoon- al standards that screw to walls and carry flimsy metal brackets that hook
shaped pins made of nickel or brass. Several shops I know of bore oversize into the slots. But heavier-duty, twin-slotted systems (which have two rows
holes and tap special metal sleeves into the holes to support these pins. of slots and hooks instead of one) are also available. The system sold by
The sleeves help prevent the holes from deforming, and they lend an air of The Woodworkers' Store includes special screws that prevent shelves from
refinement to unused holes. Dave Sanders & Company, Woodworker's sliding off the brackets and bookends that clip to the standards.
Supply and others sell matching spoons and sleeves. Shelf pins can also be Dave Sanders & Company has the best selection of wall-mounted stan-
hand-carved, cut from dowels or even fashioned out of brass brazing rods. dards and brackets I know of. One type is mounted to studs before dry-
For invisible shelf support, wire clips are the best choice. They're inserted walling so that only a slim slot is visible afterward.

118 The Best of Fine Homebuilding

 DESIGNING SHELVES FOR LOOKS AND FUNCTION

Well-designed shelves don't sag—Most cabinet shops use simple rules

of thumb to determine shelf spans. Generally (when 3/4-in, thick stock is
used to support heavy reference books), particleboard and medium-
density fiberboard (MDF) shelves span up to about 20 in.; softwood and
plywood shelves span up to about 34 in.; and hardwood shelves span
about 36 in. Shelves that carry paperbacks will span significantly farther.
For more predictable control over shelf sag, some shops use the shelf-
deflection table published in Architectural Woodwork Quality Standards
(available for $50, or $5 for membeis, from the Architectural Woodwork In-
stitute, 13924 Braddock Road, Suite 100, Centreville, Va. 22020; 703-222-
1100). The table lists the uniform loads that cause various unfixed 8-in.
and 12-in, wide shelving materials to deflect 1/4 in. when spanning 30 in.,
Aprons stiffen the
shelves. Putting a so!-
36 in., 42 in. and 48 in. id-wood apron on a
AWl's chart points out a number of ways to beef up shelves to increase shelf increases its load-
spans or to support unusually heavy loads. For instance, gluing a '/8-in. carrying capacity. The
thick edgeband to a 3/4-in. thick particleboard shelf increases the shelf's sagging top shelf is
¼-in. particleboard.
load-bearing capacity by about 15%. Veneering the particleboard on two The virtually straight
faces and one edge with 5/ioo-in. thick plastic laminate increases its capac- bottom shelf is 3/4-in.
ity by about 200%. Gluing a 1x2 solid-wood apron to the front edge boosts particleboard with a
it by a whopping 300% to 400% (photo right). Greg Heuer, AWl's director of 1x2 apron glued to the
member services, notes that applying an apron to the back of a shelf, even front edge.
pointing up (so that it hides behind books), gives the same results as a front
apron. Although most people would consider a shelf deflection of 1/4 in. to case and fastening the back to the shelves, or by fastening intermediate
be excessive, halving the values listed in the chart would produce a less support posts to the front edges. Adjustable bookcase shelves can be re-
noticeable deflection of 1/8 in. inforced at the back with shelf standards and brackets or with hole-mount-
Several other strategies can be used for increasing shelf spans or load- ed shelf pins.
bearing capacities. The most obvious is to use thicker shelves. Shelves If you doubt the ability of a proposed shelf to support a gwen load, prop
made of 2x lumber, or two layers of 3¼-in. plywood, for example, will span a sample shelf on a pair of blocks, load it with the weight it will carry and
at least 48 in. Applying a half-round molding to the front edge of layered check the sag.
plywood creates a solid builnose shelf.
For a significant span, consider building a torsion-box shelf, which works Recessing and trim make shelving look built in—Louis Mackall,
somewhat like a box beam or hollow-core door. These shelves consist of a owner of Breakfast Woodworks Inc. in Guilford, Connecticut, tells me that
grid of wood or plywood strips glued between two plywood skins. The the best way to integrate shelving into a room is to recess it at least partway
biggest torsion-box shelf that I've heard of is 4 in. thick and 27 ft. long. into a wall (top photo, p. 120), even if this procedure requires furring out
Fixed shelves in bookcases can be reinforced by putting a back on the the wall. According to Mackall, recessed shelves not only look better than

Sources for bookshelf hardware

Here's a short list of some useful companies duty metal wall brackets that support up to Sells assorted shelf pins, including solid-brass
to know about if you're in the market for 1,000 lb. per pair. pins and brass sleeves that reinforce
bookshelf hardware.—B. G. pinholes.
Rangine Corporation
Dave Sanders & Company Inc. P. 0. Box 128, Millis, Mass. 02054 The Woodworkers' Store
107 Bowery, New York, N. Y. 10002 (800) 826-6006 4365 Willow Drive, Medina, Minn. 53470
(212) 334-9898 Makes storage systems, including the Rakks (800) 279-4441
Sells an impressive assortment of shelf Shelving System: extruded-aluminum, wall- Sells a range of knockdown fasteners,
standards, brackets and shelf pins, including mounted shelf standards with locking, standards, brackets and pins; the best
Magic Wire concealed shelf supports and infinitely adjustable aluminum brackets. assortment of edgeband I know of; and tools
shelf pins with matching sleeves that and accessories.
reinforce pinholes. Renovator's Supply
P. 0. Box 2515, Dept. 9898, Conway, N. H. Woodworker's Supply Inc.
Knape & Vogt Manufacturing Company 03818-2515 1108 N. Glenn Road, Casper, Wyo. 82601
2700 Oak Industrial Drive NE, Grand Rapids, (800) 659-0203 (800) 645-9292
Mich. 49505-6083 Sells ornamental brass shelf brackets. Sells assorted standards, brackets, shelf pins,
(800) 253-1561 and tools and supplies.
Makes cabinet and shelving hardware, Woodcraft Supply
including twin-slotted, wall-mounted P. 0. Box 1686, Parkersburg, W. Va. 26102
standards and brackets, shelf pins and heavy- (800) 225-1153

Finish Carpentry 119

 Recessed shelving doesn't infrude. Breakfast Woodworks in Gull-
ford, Connecticut, recesses shelving to blend it with the architecture
and to create the illusion of space.

projecting shelves, but they also appear to add space. Projecting shelves
appear to subtract space. Continuing a room's base and crown moldings
(if there are any) around a bookcase also helps to unite the bookcase vi-
sually with the room.
San Francisco woodworker Scott Wynn considers horizontal details in a
room when sizing built-ins. If window head casings are a prominent fea-
ture, for instance, he'll make the built-ins the same height as the head cas-
ings. Wynn also makes bookcases supported by deeper base cabinets.
The cabinet tops serve as oversize shelves that support art books. Wynn
cautions that bookcase partitions must be placed directly over cabinet
partitions or the shelves will sag.
Except for furniture-grade pieces such as Thos. Moser's, most bookcases
are boxes with routed edges or applied trim. The boxes can have butt
joints held together with nails or screws in concealed locations, or with bis-
cuits or plugged screws in exposed locations (for more on biscuit joinery,
see FHB #70, pp. 50-53). Special knockdown fasteners can also be used, al-
lowing units to be dismantled and reassembled. Tall bookcases should
have '/4-in. hardboard or plywood rabbeted into and tacked to the back for
stability, a lx nailer at the top for attachment to walls, or both.
Bookshelves don't have to nest in bookcases to look good. Philadelphia
woodworker Jack Larimore has designed and built economical wall-
Bracket-mounted bookshelves can look built In. Philadelphia mounted shelving that looks like pricey built-in furniture (photo left). The
woodworker Jack I.arimore applies fancy aprons to basic wall-mounted shelves are supported by metal shelf standards, but they're dressed up with
bookshelves to produce ornate library storage that is available for a decorative aprons. Ornamental metal and wood brackets can also be used
modest price. to enhance the appearance of wall-mounted shelving (left photo, p. 117).

120 The Best of Fine Homebuilding Top photo: Louis Mackall. Bottom photo: Mitch Mandel/Rodate
Stock tmages.
 CHOOSING THE RIGHT MATERIALS

Solid wood is beautiful but unstable—Thos. Moser builds most of its

bookcases out of 3/4-in. thick American black cherry. At the other extreme,
I've used low-cost #3 knotty pine for a number of shelving jobs because it's
relatively strong and inexpensive. Unlike some shelving materials, solid
wood doesn't require edgeband, and edges can be detailed easily with a
router or a hand plane.
Unfortunately, solid wood is unstable. It shrinks as it dries, and it ex-
pands and contracts across the grain in response to changes in humidity. It
also tends to warp and cup more than sheet stock, causing adjustable
shelves to seesaw.
One nifty idea I've seen is to use bullnose hardwood stair-tread stock
for shelves. It usually comes in red or white oak, it's often glued up out of
narrow strips for dimensional stability, and it's a sturdy I in. to 1'/16 in. Cover exposei wood edges with edging. Options include (top)
thick. Millworks and many lumberyards sell it, and when you factor in 1-molding; pine en bead; plastic 1-molding; peel-and-stick hardwood
presanded finish and ready-made edge treatment, the price is reasonable. edgeband (oute: 1); and iron-on polyester edgeband (inner roll).

Hardwood plywood is an industry favorite—Most shops I spoke with

shun solid wood in favor of sheet stock such as hardwood plywood, par-
ticleboard or MDF (bottom photo). Sheet stock is more stable than solid
wood, and it normally can be turned into shelving parts faster because it
doesn't have to be flattened and straightened as solid wood often does.
Hardwood plywood is available with a veneer core, an MDF core, a par-
ticleboard core or a lumber core. Veneer-core plywood is strong, has ex-
cellent dimensional stability, holds screws well and is widely available
(especially in birch, red oak and lauan). But it can have voids in the core
that could occasionally cause a hole-mounted shelf pin to sag or to fall out.
MDF-core and particleboard-core plywood don't have voids, but they
also don't hold screws as well as veneer-core plywoods do, although using
particleboard screws (which have a deep, coarse thread) helps. These Sheet stock is stable and umforin choice for shelving. Options
panels are also heavy, weighing almost 100 lb. per 3/4-in. thick sheet. This range from hat 'ood-veneer plywood to medium-density fiberboard
weight not only makes them awkward to machine, but it also makes them panels with a di tble melamine surface.
a questionable choice for portable bookcases.
Nevertheless, some shops use MDF-core plywood exclusively for their
bookshelves. They like its solid core and its cost, which is about 5% to des, resulting in p nel surfaces that are as flat as glass and edges that ma-
20% less than veneer-core plywood. Georgia-Pacific's Fiber-Ply core ply- chine beautifully vithout chip-out (machining produces clouds of fine
wood and Columbia Forest Products' Classic Core plywood are com- dust, though, so w aring a high-quality dust mask a must). The Home De-
pelling hybrids. They look like conventional veneer-core plywood, but pot in my area sell 3/4-in. MDF for about $20 per sheet, which is a bargain.
they have homogenous layers of fiberboard instead of solid-wood plies di- One problem with MDF and particleboard panels is that unsealed edges
rectly beneath the face veneers. drink paint, in an uneven finish. You can seal MDF edges with two
Lumber-core plywood is strong, it holds screws tenaciously, and it is sup- coats of polyurethane finish, but quick-drying sanding sealers
posed to be voidless. But today's lumber-core plywood, much of which is and white PVA diluted 20% with water will also work.
imported, can have shrinkage voids in the core. Good-quality lumber-core Paint-grade called MDO (medium-density overlay), can also be
plywood is expensive and can be hard to find. used for painted shelving. This exterior-grade veneer-core plywood is coat-
Regardless of the type of plywood used, exposed edges need to be cov- ed with a paper that makes a superb, smooth substrate for
ered with moldings or edgeband (top photo). Moldings can be anything paint. MDO can a so be taped like drywall, so it makes a good material for
from pine screen bead to solid-wood or plastic 1-moldings that fit into built-in shelving.
grooves in panel edges. Edgeband is peel-and-stick, iron-on or glue-on ve- MDF and particleboard can also come veneered with melamine or plas-
neer that comes on a roll. Solid-wood, polyester and even metallic-foil tic laminate. Melamine is a thermally fused, resin-impregnated sheeting
edgebands are available. that resists stains, heat and chemicals, and it's available in a va-
riety of sizes, patterns and colors. The Home Depot sells melamine panels
Particleboard and MDF work well for paint-grade work—Particle- that are ripped standard bookcase widths and that are predrilled with
board and MDF are hard to beat for paint-grade shelving systems. They're '/4-in, holes that standard shelf pins. Melamine can be tricky to
relatively inexpensive and stable, and they don't require edgeband. Most work with it chips easily. But it's a versatile material, and it has a
lumberyards sell 3/4-in. industrial-particleboard shelving with filled bull- hard prefinished (For more on working with melamine, see FHB
nose edges. Available in many widths—typically 8 in. or 12 in.—precut #99, pp. 68-73).
shelving is quick and convenient.
contributing editor olFine Homebuilding. Photos by the
Called the "Buick of particleboard" by Albuquerque woodworker Sven Bruce Greenlaw 4a
Hanson, MDF is made of highly compressed wood fibers instead of parti- author, except w ere noted.

Finish Carpentry 121

 A Home Library
Maple, purpleheart and an enthusiasm for small details

by Charles Wardell

writer once said that there's no furniture

as charming as books. Too often, though, the
furniture made for books barely rises above
the level of utility. But high-grade materials, a
thoughtful design and attention to detail can
turn what might have been little more than
storage into a home's focal point.
Such was the case when Robert Hillen of
Interior Woodworking in Cincinnati was hired
to redesign the interior of a condo. The new
owners were trading down from a larger home,
and though that meant shedding excess pos-
sessions, there was one thing they were un-
willing to part with: their 360 lineal feet of
books. In fact, they wanted room to expand it
to 420 ft. Hillen's design included a complete
redo of the interior, but its centerpiece was a
massive maple bookcase that covered 30 ft. of
the living room wall (photo right).

Durable elegance—After completing the de-

sign, 1-lillen handed the project over to the
company's head woodworker, Bradd MacCal-
lum. MacCallum, who spends the bulk of his
time in the shop building custom furniture,
has a head for numbers and a passion for ac-
curacy. Everything was built in the shop, then
assembled like a puzzle at the job site. This
demanded both a great deal of planning and
precise measurements. Many of the pieces were
made to tolerances of h/&4 in.
Although the bookcase would be a dominant
feature of the house, it couldn't be overwhelm-
ing. Purpleheart detailing was used all around
to this end, helping to make the bookcase an
piece of workmanship, rather than an
imposing mass of wood. The purpleheart's
deep red contrasts sharply with the blond ma-
ple of the shelves, while at the same time
complementing the violet hue of the carpet.
The details are where the fumituremaker's hand
really shows, enticing the observer to draw
closer rather than forcing him to step back.
But if appearance was important, the weight
of all those books meant that strength was
crucial. Luckily, the condo had been built on
a thick concrete slab, so floor strength wasn't

When Robert Hillen was asked to redesign the

Interior of a condo, the main question was
what to do with all the clients' books. Hillen
answered with an attractive wall unit that cov-
ered the 30-ft. length of the living room wall.

122 The Best of Fine Homebuilding

as much of a problem as the strength of the long, then trimmed them on site for a tight fit
bookcase itself. [-lillen specified a 5-in, thick- to let the front of each unit sit squarely on the
ness for the two center uprights, and 3 in. for floor, in spite of a varying floor level.
the rest. Solid wood was out, though. Not only A 4/4 poplar block glued into the top of each
would solid maple be unreasonably expen- upright served as a nailer for the purpleheart
sive, but its weight would make the uprights top shelf and built-up pediment. MacCallum
too heavy to transport and install. MacCallum made the top shelf oversize, scribed its back
considered using wooden frames clad with edge to compensate for variations in the wall
in plywood, but those would present similar plane, then cut the leadthg edge. This gave him
problems. Finally, he hit upon a solution. a tight joint at the wall and a consistent 5/8-in.
overhang in front. The pediment was crowned
A high-grade prefab—To make the uprights, with a purpleheart arch, added on site to achieve
MacCallum built a sandwich of Styrofoam and a precise fit (top photo on p. 124).
1/2-in. bird's-eye maple plywood (see drawing The shelves themselves are 5/4 maple joined
on p. 124), laminated together with Wilsonart edge to edge. They're trimmed with a half-
spray epoxy (Ralph Wilson Plastics Co., 600 round purpleheart accent strip set in a routed
General Bruce Dr., Temple, Tex. 76504; 817- slot at each shelf's leadling edge. At the back, a
778-271 1). The sandwich is both lighter and The purpleheart handrail doubles as a ¼-in, gap separates the shelves from the fin-
less costly than framed uprights would have support rail for the rolling ladder (photo ished drywall. This gap not only protects the
been, so MacCallum could install them with above). The ladder's two pairs of rail wall, but it leaves room for wiring, giving the
hooks allow it to be hung from the rail clients the option of using the shelves for
one helper. The foam added stiffness to the as- vertically for storage or slid along in a
sembly, enabling him to use ½-in, plywood diagonal position for use. Note the felt speakers or other electronic equipment.
for the sides instead of 3/4 in. —with a 30% sav- pads on the underside of the hooks; they
ings in cost. The system left the option of easi- help the ladder slide along the railing. A double-duty railing—One of the most in-
ly trimming the uprights if need be, because genious parts of the bookcase design is the
the completed sandwich could easily be run continuous purpleheart support rail for the
through a table saw. And Styrofoam muffles rolling library ladder (top photo, left). After tak-
sound; by contrast, a wooden frame with ply- ing a 20-ft. run across the front of the book-
wood sheathing would feel and sound hollow. case, it turns a corner, goes across the balco-
Half-round maple facings flanked by purple- ny, then travels down the U-shaped staircase.
heart accent strips cover the leading edge of The railing was made from glued-up strips (3
the uprights. The facings were made from laminations in all). Their edges were rounded
glued-up maple that was rounded with a router with a quarter-round carbide bit. Routing one
and a carbide quarter-round bit, then glued to a edge at a time, MacCallum was able to cut
flat maple base (photo on p. 124). about '/16 in. per pass. All joints are staggered
To ensure a tight joint at the wall, MacCal- at least 2 ft. apart. A sliding table rig on the
lum scribed and trimmed the backs of the shop table saw ensured precise crosscuts. Al-
uprights. Then, using a router and template, though the raiiing itself was laminated with yel-
he removed the foam at the top, bottom and low woodworker's glue, the turns were joined to
rear edges of each upright to a depth of 2 in. the straight runs with dowels and cold-pressed
The resulting channel at the rear slid over 8/4 resin glue (National Casein Co., 6W W. 80th St.,
by 2-in, poplar nailers that had been screwed Chicago, Ii. 60620; 312-846-7300). It was diffi-
through the finished drywall to horizontal block- cult to clamp the curved connection accurate-
ing between the wall studs. Finish nails se- ly; the resin glue set up fast enough to let the
cure the panels to the nailers. connection be made with hand pressure.
A router was also used to make the holes The purpleheart rail is supported by adapter
for the brass shelf pins (Hafele American Co., screws — double-ended screw-bolts with ma-
3901 Cheyenne Dr., Archdale, N. C. 27263; 800- chine and wood threads—covered with hol-
334-1873). MacCallum made an 8-ft. long birch low dowel spacers that MacCallum bored out
plywood template with %-in. dia. holes at each in the shop with a router (drawing left). The
shelf-pin location. After clamping the template wood threads were turned into the upright fac-
to an upright, a plunge router with a %-in. ings, the balcony and the stair half-walls; the
bushing and a Y4-in. straight plunge-cutting bit dowel was then installed over the machine
made short work of the holes. There were ad- threads and the railing was fitted over the
vantages to this method, the most important dowel. A countersunk hole at the front of the
being that it proved exceedingly accurate. The The ladder wheels were made in Hillen's shop railing allowed the installation of a nut over
after he tried—and failed—to find an off-the- the machine threads. This hole was later filled
router also did a cleaner job than a drill; there shelf set that worked with the design.
was no tearout at any of the 1,000 or so holes. with a slotted maple plug. A purpleheart wedge
And sharpening the bit made it a hair under- forced into the slot secured the plug in the
size, just small enough to guarantee a secure concrete floor with Tapcon screws (ITW Buil- hole. The whole assembly—shelves, uprights,
fit for the '/4-in, shelf pins. dex, General Construction Systems, 1349 W. railings —was sanded smooth with a palm sand-
Of course, aligning the finished uprights Bryn Mawr Ave., Itasca, II. 60143; 708-595- er and finished with three coats of Watco oil.
during installation was essential, if the shelves 3500). Each upright was shirnmed until the
themselves were to line up. A level line mark on the back met the line on the wall. The library ladder—A rolling ladder is a wel-
snapped on the wall and a corresponding mark The block and the shim space were later cov- come touch to any library, but 1-lillen's 16-ft.
on the back edge of each upright solved that ered with baseboard. high shelves would have been useless without
problem. The bases of the uprights rest on Holding the half-round upright facings back one. The ladder rails taper from a width of 3'/2
poplar blocks with lengthwise shoulders (draw- 6 in. from the floor left room for a purpleheart in. at the top to 5 in. at the bottom. The steps
ing on p. 124) that were first screwed to the plinth block. MacCallum cut these blocks a bit sit in dadoed slots and are held fast with #8

Finish Carpentry 123

 Purpleheart pediment shelf

2-in.
Half-round Styrofoam
maple facing

The was capped a built-up pedi-

t

ment, fabricated with foam and plywood just

as the shelf uprights were.

½-in, birds-eye maple

plywood

A foam sandwich
The plywood skins
were laminated to the
Styrofoam with a spray
epoxy. Everything else
is attached with
standard carpenter's
glue. Vertical nailers
bolt through the drywall
Purpleheart details were used all around. They to horizontal blocking
encourage the observer to look closer. Li between the studs.

¼-in. dia.
brass shelf pin
accent strips
wood screws 21/2 in. long. The countersunk
holes were plugged just like those in the rail-
ing, but this time with purpleheart plugs and
maple wedges. The steps are 4 in. deep. Like
the shelves, they are trimmed with half-round
purpleheart accents set into their leading edges.
A 6-in, shelf at the top of the ladder serves as
a seat for browsing.
5/4 by 10½-in.
The top of the ladder sports two pairs of maple shelf
maple rail hooks (top photo). Each was made
from 20 pieces of maple veneer laminated
around a curve that matched the handrail.
Their undersides are covered with felt so that
they'll easily slide along the rail. The top hooks
are the working hooks; when the ladder is in
use, they hold its wheels 3 ft. from the front of
the bookcase. And while the angle of the lad-
der is admittedly on the steep side, it was the
best that the available floor space would al-
low. The lower pair of hooks allow the ladder
to be hung vertically, keeping it out of the way
when floor space is needed for other activities.
Hillen had originally specified stock wheels
for the ladder. Unable to find any that worked
well with the design, however, he asked
MacCallum to make a pair (bottom photo on
p. 123). Each wheel consists of two pieces of
cross-laminated maple that turn on a 5/16-in.
metal axle and a pair of flanged roller bear-
ings. The wheel assemblies are set into slots
at the base of the ladder.

Charles Wardell is an assistant editor of Fine Shoulcered floor block

Homebuilding. Photos by the author.
Plinth block
Drawings: Bob LaPointe
 INDEX
A Chamberlain, Samuel and Narcissa:
Southern Interiors of Charleston, South
G
Aluminum: cutting, cautions for, 67 Glue:
Angles: board "recorder" for, 8, 9 Carolina, 37
epoxy, aerosol, source for, 123
Arches: Circular saws:
for lamination, 59, 123
for casings, building, trim, for aluminum, 67
urea-formaldehyde, mixing, 59
four-center, 98-99, 100 See also Sawblades.
Griesbach, C. B.: Historic Ornament,
Architectural Treasures ofEarlyAmerica Closets: wardrobe for, making,
cited, 103
(Early American Society): cited, Columns, classical:
37, 38 39 books on, cited, 60
Architectural Woodwork Institute: making, 60-63 H
Architectural Woodwork Quality See also Pilasters. Hardboard: for mock raised panels, 78-81
Standards, cited, 119 Compasses: scribing with, 8-10 Hasluck, Paul N.: Manual of Traditional
Concrete: arch blocks of, cast, 98-99 Woodworking cited, 103
Cornices: for windows, 72-73 Hutches: built-in, 104-107
B Cupboards: See Hutches.
Bandsaws: circle-cutting jig for, 75 Curves: See Arches. Compasses. Ellipses.
Base hooks: using, 8, 9 I
Baseboards: Inglenooks: baronial, building, 98-103
coping, 30-31 D
curved corners for, 32-33 da Vignola, Giacomo Barozzi: Comparison
cutlist form for, 36 of Orders, cited, 60
Design: Joe Frogger: using, 12-13
installing, 8, 30-31, 34-36
for arches, 99 Joinery:
measuring blocks for, 8, 30-31
books on, cited, 102, 103 butt,
plate-joined, 23
for carved projections, 103 plate-joined, 21, 22
Basswood (Tilia americana): for trim,
standards for, book of, 119 vs. scarf joints, for crown
Beads: applied, for casings, 72
See also Colors, molding, 53
Beds: built-in, for kneewall space, 112-115
Doors: coped, 30-31, 50-53
Benchstones:
cabinet, making, 107 glue-joint, 105
diamond, source for, 18
thresholds for, 3-piece water-return, miter,
Japanese water, source for, 18
64-67 on crown molding, 51-52
Bevel boards: making, 8, 9
See also Transoms, plate-joined, 20, 22
Bevel gauges: using, 8, 9
Drawers: shooting board for, making, 16
Biscuit joinery: See Joinery: plate.
with biscuit joiner, 110, 111 plate,
Bits: plug-cutter, source for, 93
guides for, wooded, 112 advantages of, 20
Bookcases:
under-bed alcove, 112-1 15 for casings, 70-71
in bed alcove, 113, 114, 115
DrillIng, templates for: for curved corner pieces, jig for, 32
book sizes for, 117
commercial, 117-118 epoxy with, 23
brackets for, 117, 118, 120
from pegboard, 117, 118 jig for, 111
custom, 116
Drywall: for mantel miter, 94-95
designing, 120
beads for, 87, for narrow stock, 23
formal, 41
corners for, rounded, supply sources for, 23
hardware for, sources for, 119
crimper with, 507 using, 20-23
metal standards for, 117
moisture-resistant, around windows, wardrobe with, 108-111
pins for, 117, 118
sanding, 88-89 pocket-screw, 106, 107
recessed, 119, 120
supply source for, spline-and-groove, 105, 106-107
spans for, 119
trim in, 86-89 tenons, table-saw jig for, 83
Jointers: glue joints with, 105

Cabinets: E
Ellipses: drawing, 97 L
built-in, 104-107
Entryways: arched, angular, 99 Ladders:
plate-joined, 22-23
library, rolling, making, 123-124
See also Drawers. Hutches.
wheels for, 123, 124
Carving: of sunburst, for mantel, 96-97 F Lamination:
Casings. Fiberboard: See Particleboard. for casings, curved, 74-77, 99-100
aprons in, 72-73 Fireplaces: glue for, 59, 99, 123
arched, 98-102 hearths for, slate, source for, 94 of molding, crown, curved, 57-59
jambs of, arched, 100-102 surrounds for, carved pilastered, Lathes:
laminated, curved, 74-77 98, 102-103 column jig for, 61-63
methods for, 68-73 See also Ingienooks. Mantels, fluting columns on, jig for, 61-62
mortised, 106, 107 Flooring: plate-joined, 22, 23 Light fixtures:
plate-joined, 20, 22, 23 Floors: posts in, scribing to, 12-13 for alcove, 113, 114, 115
stools in, 68-70 Flutes: drywall-formed, 89
window, jambs for, extension, 75 converging, making, 103 fluorescent, 115
Ceilings: boards in, scribing, 10-12 making, 90-91, 95-96 Lloyd, Nathaniel: History of the English
Forms: House, A, cited, 102
for arches, laminated, 99, 101
bending, 75, 76, 77
for laminating molding, 57

126 Index
M Planes:
Japanese,
T
Mantels: Table saws:
books on, 16 molding on, 48-49
ledger strips for, 91, 93
using, 15, 16 crown, 58-59
making, 37, 38 39 90-93 94-97
sharpening, 18-19 raised panels with, 82-83
tiles for, 94 97
Mirrors: bottom-lighted, 114, 115 soles of, truing, 15 tapers on, 96
Miter boxes: for crown molding, 50, 52 source for, 16-17 tenons on, jig for, 83, 84
types of, 14-18 Templates: for plasler molding, 43
Miters: for odd angles, 81
Plaster: moldings in, 42-45 Tile: Delft, source for, 94
Miter saws, power: crown molding with,
46-47 5455 Plywood: I

Timber framing: drywall with, installing, 89

bending, for curved panels, 78
I

Molder-planers: Transoms: elliptical, plaster molding for,

edging for, 121
beveled molding on, 84 42-45
for shelves, 121
I

grain orientation for, Trim:

Posts: scribing to, 12-13
I

knives for, making, 84 beading on, applied, 72

for whole-house job, 38 casing, arched, 98-100
Molding: R curved, plastered, 42-45
applied, for mantel, 97 Railing: and ladder support, integral, 123 in drywall, 86-89
backing blocks for, 55, 56 Routers: formal, from stock, 68-69, 70-72
coping, 81 column-turning jig for, 61-63 window, fitting, 68-71
crown, 39, 40, 41 fluting with, i
See also Baseboards. Casings.
curved, laminated, 5759 converging, 103 j
Wainscot.
installing, 50-53 jigs for, 61-62, 90-91, 95-96 Tudor style:
for mantel, installing, 93 glue joints with, 105 book on, cited, 102
two-piece, installing, 5456 shelf pin-holes with, template for, 123 fireplace in, 98-103
wide, compound-miter calculations See also Bits. I

for, 46-47
wide, source for, 46
S Wainscot:
curved, 98-100, 101
Safety: in cutting aluminum, 67 elements of, 82, 84-85
Greek-key fretwork, 39 I

Sanding: of crown molding, 59 finish for, 85

making, 37-41
with table saw and shaper, 107 Sandpaper: non-tear, source for, 59 I
installing, 11, 84,85
Sawblades: for aluminum, 67 making, 39-41, 106-107
mitering, 80-81
plaster, making, 42-45 Saws: panels for,
coping, for crown molding, 50-51 particleboard, 38, 40
on table saw, 48-49, 58-59 j

in window trim, 68-73 See also separate machines. plywood, 84

See also Beads. Scrapers: making, for crown molding, 59 raising, 82-84
Screws: Walls: curved, bending plywood for, 78
adapter, for railing, 123 Wardrobes: for making, 108-111
N for concrete, source for, 123 WIndows:
Nailers, power finish: Scribing: drywall around, moisture-resistant, 88
cordless, 25 ceiling boards, with spiling battens, stools for, scribing, 10, 12
reviewed, 24-29 10-12 See also Casings. Dormers. Transoms.
Nails: for power nailers, good vs. bad, 27 pencil for, 8 Wiring: chases for, drywall-formed, 89
techniques of, 8-13
0
Ornament: designing, books on, 103
Sharpening: of plane irons, system for,
18-19
Shelves:
aprons for, fancy, 120
P facings for, inlaid, 123, 124
Paint: Federal gray, source for, 97 ladder for, rolling, 123, 124
Panels: library, maple, 122-24
raised, pins for, brass, 123
with bending plywood, 78-81 plate-joined, 21, 22, 23
for curved wall, 78-81 scribing, 10
on table saw, 82-83 spans for, 119
scribing, 10, 11 stock for, 121
See also Wainscot, strengthening, 119
Particleboard: uprights for, foam-sandwich, 123, 124
edges of, sealing, 121 See also Bookcases.
plastic-laminated, 121 Shooting boards: for miters, 16
for shelves, 121 Siding: installing, 11
for wainscoting, 40 Skylights:
Pencils: mechanical, guaranteed, 8 over bed alcove, 113
Pilasters: drywall around, moisture-resistant, 88
Jacobean, 102-103 Slate: for hearth, source for, 94
making, 91, 92, 93 Spiting battens: using, 10-12
Staircases:
circular, frame-and-panel wall under,
78-81
skirtboards of, scribing, 10
wainscoting for, making, 41

Index 127
The articles in this book 8 Basic Scribing 64 Retrofitting a
Techniques Threshold
originally appeared in October 1992 (77:58-63) Aprii 1996 (102:70-73)
Fine Homebuilding
14 Hand Planes for 68 More Than One Way
magazine. The date of Trim Carpentry to Case a Window
first publication, issue August 1992 (76:80-85) October 1995 (98:54-59)

number and page 20 Plate Joinery on the 74 Making Curved Casing

numbers for each article Job Site April 1991 (67:82-85)
October 1991 (70:50-53)
are given at right. 78 Raised Paneling
24 Survey of Finish Nailers Made Easy
October 1995 (98:72-77) April 1995 (95:90-93)

30 Installing Baseboard 82 Raised-Panel Wainscot

August 1986 (34:40-41) February 1984 (19:50-53)

32 Curved Baseboard 86 Drywall Detailing

Corners October 1990(63:50-53)
August 1994 (90:68-69)
90 Building a
34 Running Baseboard Fireplace Mantel
Efficiently February 1995 (93:66-69)
August 1992 (76:51-53)
94 Building a Federal-
37 Molding Character Style Mantel
June 1987 (40:33-37) June 1989(54:54-57)

42 Making 98 Baronial Inglenook

Plaster Molding August 1991 (69:82-87)
February 1995(93:76-79)
104 A Built-In
46 Cutting Crown Molding Hardwood Hutch
June 1991 (68:79-80) December 1993 (85:64-67)

48 Table-Saw Molding 108 Simple Closet

April 1981 (2:48-49) Wardrobe
December 1992 (78:44-47)
50 Installing Crown
Molding 112 Bed Alcove
February 1989(51:64-67) August 1992 (76:42-45)

54 Installing Two-Piece 116 Bookshelf Basics

Crown June 1996(103:98-103)
December 1991 (71:85-87)
122 A Home Library
57 Making Curved December 1990 (65:85-87)
Crown Molding
April 1992 (74:79-81)

60 Making
Classical Columns
December 1990 (64:54-57)
4
4
 FINISH
CARPENTRY
THE BEST OF
Fine Homebuilding

Quality finish carpentry can transform

ordinary walls into finely crafted rooms.
In this collection of articles from Fine
Homebuilding magazine, you'll learn the
secrets of the master carpenters on The Best of
designing and installing baseboard, Fine Homebuilding
molding, window casing (including series:
elegant curved casing), fireplace
mantels, paneling, and built-in storage.
Inside, learn how the pros: Doors

• fit trim to uneven surfaces Finish Carpentry

choose the right tools Foundations &

for any finish carpentry job Concrete Work
• install baseboard efficiently Framing Floors, Walls,
• re-create historic molding profiles Ceilings
• case windows and doors Framing Roofs
for weathertightness and good looks
Porches, Decks &
build an elegant fireplace mantel Outbuildings
• design built-in bookshelves
Roofing
Since it began publication in 1981, Fine Stairs
Homebuilding has been the magazine of
choice for builders and homeowners Windows & Skylights
interested in working smarter, faster,
and better. 'The Best of Fine
Homebuilding" series organizes the most
useful articles from the first 100 issues
of Fine Homebuilding magazine by
subject. Whether you are working on
your own house or in the building ISBN 1—56158—183—6
trades, you will find the field-tested
information you need right here. 9000C

Taunton
BOOKS & VJDEOS
9 781 561 581832

$14.95 U.S. for fellow enthusiasts Taunton Product #