M ARKETI NG REPORT

   

MR-261
M otivation for Testing xDSL Splitters and I n-Line Filters

Issue: 1
Issue Date: May 2012

© The Broadband Forum. All rights reserved.

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

I ssue History
I ssue Approval Publication I ssue Editors Changes
Number Date Date
1 14 May 2012 3 July 2012 Andre Holley, Sparnex Instruments Original
Peter Reusens, LAN

Comments or questions about this Broadband Forum Marketing Report should be directed to
[email protected].

Editors Andre Holley Sparnex Instruments

Peter Reusens LAN

White Papers Chair Sarah Banks Cisco

Vanessa Nacamura Nokia Siemens Networks

Chief Editor Michael Hanrahan Huawei

May 2012 © The Broadband Forum. All rights reserved 2 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Table of Contents
1   I NTRODUCTI ON .............................................................................................................5  
1.1   POTENTIAL EXTENDED USE OF TR-127 .........................................................................5  
2   XDSL SPLI TTERS & I N-LI NE FI LTERS ± TECHNICAL OVERVI EW ....................6  
2.1   UBIQUITOUS TELEPHONE CABLES, SUCCESSFULLY RE-USED FOR XDSL PLUS POTS .......6  
2.2   SHORT INTRODUCTION TO THE POTS SYSTEMS .............................................................7  
2.2.1   Significance of the Ring-Trip Event ..........................................................................7  
2.2.2   POTS and in-house wiring as source of disturbance for xDSL: ................................7  
2.3   OVERVIEW OF XDSL SIGNALS ON THE COPPER LINE .......................................................8  
2.4   MULTIPLEXING AN XDSL SYSTEM ON A POTS CABLE CARRYING A POTS SIGNAL .........8  
2.4.1   6LQJOH³0DVWHU´6SOLWWHUYHUVXV0XOWLSOH,Q-Line Filters..........................................8  
2.5   DISTURBANCE OF XDSL BY POTS AND OTHER NOISES AND DISTURBANCES ................. 10  
2.6   NETWORK ISSUES CAUSED BY POOR QUALITY FILTERS AND IN-HOUSE WIRING............... 10  
3   TESTI NG DSL SPLI TTERS AND I N-LI NE FI LTERS ACCORDI NG TO TR-127 ..11  
3.1   GENERIC INTERWORKING TESTING DIAGRAM FOR XDSL OVER POTS ........................... 11  
3.2   ON-HOOK TO OFF-HOOK RING-TRIP TRANSITION: A CRITICAL TEST IN TR-127.............. 12  
3.3   THE POTS MODELS .................................................................................................... 12  
4   CONCLUSI ONS..............................................................................................................13  
5   REFERENCES AND TERM I NOLOGY........................................................................14  
5.1   REFERENCES .............................................................................................................. 14  
5.2   DEFINITIONS............................................................................................................... 15  
5.3   ABBREVIATIONS ......................................................................................................... 15  

List of Figures

)LJXUH0XOWLSOH[LQJWZRVHUYLFHVZLWK³0X[-ILOWHUV´RQDVLQJOHWHOHSKRQHOLQH ......................6  
Figure 2: Spectra of POTS and ADSL merged on a twisted pair with mux-filters ........................6  
Figure 3: $FWXDO³0XOWLSOH[LQJ´3276DQG['6/RQDVLQJOHWZLVWHGSDLU ..................................7  
)LJXUH)XQFWLRQDOGLDJUDPRIDVLQJOH['6/³PDVWHU´VSOLWWHUFRQILJXUDWLRQ ...........................9  
Figure 5: Functional Diagram of the xDSL In-Line Filter configuration ......................................9  
Figure 6: Generic interworking testing diagram for xDSL over POTS ....................................... 11  

May 2012 © The Broadband Forum. All rights reserved 3 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Executive Summary
Technical Report TR-127 of the Broadband Forum is an xDSL system test. It checks primarily if
xDSL splitters used at the Central Office (CO) and at the Customer Premise Equipment (CPE)
side are capable of protecting the xDSL signals against voltages and other disturbances present
on a telephone line. To achieve this, TR-127 contains a series of tests, which model all signals
of the Plain Old Telephone Service (POTS) in a realistic way, which includes all POTS
disturbances and transients that may affect xDSL.
Splitter devices should be tested dynamically in an xDSL system:
ADSL and VDSL (xDSL in short) are carrying High Speed Internet, IPTV and VoIP. Therefore,
the Bit Error Ratio (BER) of xDSL can significantly impact the quality of the delivered services
and consequently the user experience. Splitters or filters passing TR-127 can guarantee the
necessary quality to pass triple-play service with a low BER.
A splitter is used at the CO side and either a splitter or in-line filters (a.k.a. ³PLFUR-ILOWHUV´ DUH
used at the CPE side for the deployment of xDSL. They should minimize interference into the
broadband xDSL from signals of the traditional POTS service and other sources from the
network side / or upstream. Particularly poor quality splitters and filters are known to be sources
of errors and signal losses induced in xDSL. These splitter devices must have sufficient quality
to achieve the required xDSL performance. Therefore it is recommend that splitters and filters
be tested dynamically, which can only be achieved by the complete xDSL system test in TR-127.
Before TR-127 was compiled, splitters and filters were only tested according to the static test of
ETSI or ATIS. While still necessary these tests are known to be insufficient to guarantee the
correct operation under dynamic conditions.
It is concluded that splitters successfully passing TR-127 can guarantee a virtually error free
performance of a system of xDSL transceivers including the splitters. The usefulness of TR-127
lies in its testing xDSL systems and splitters dynamically, together with POTS voltages, signals
and other disturbances. This issue is not addressed by other specifications.
TR-127 Issue 2 also achieves dynamic tests for ISDN splitters. MR-261 will concentrate on the
POTS splitters only.
The importance of dynamic testing is emphasized. Only dynamic tests as opposed to static ones,
can guarantee the necessary QoS for triple-play xDSL services. MR-261 concludes with a brief
description of TR-127 test suite, and provides a set of conclusions and best practice
recommendations for splitter design and deployment.

May 2012 © The Broadband Forum. All rights reserved 4 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

1 I ntroduction
ADSL and VDSL (xDSL in short) are attractive systems because they allow the merging of the
xDSL signal with the Plain Old Telephone Service (POTS) on a single twisted pair line, with
little mutual disturbance.
When xDSL and the POTS service coexist on a single telephone line, the signals must be merged
by a set of multiplexing filters, i.e. the splitters or in-line filters as shown in Figure 1. If these
splitters or filters are of poor quality design, the Bit Error Ratio (BER) of the xDSL will suffer,
and the quality of the triple-play services over the xDSL will deteriorate.

Before the creation of TR-127 by the Broadband Forum, splitters were tested according to static
ETSI and ATIS standards. These tests impose static requirement on xDSL splitters and in-line
filters. However, these ETSI and ATIS standards protect primarily the quality of the POTS
signals. Therefore, although necessary, these static tests are not sufficient, because they do not
test the actual dynamic rejection of the POTS signals and transients, which can affect the BER of
the xDSL.
Dynamic tests of xDSL systems were developed by the Broadband Forum since its earliest
existence. Both functional tests and performance tests of both ADSL and VDSL were created.
However, other existing TRs of the Broadband Forum do not fully test xDSL in the presence of
the splitters with complete POTS signals. Only Broadband Forum TR-127 tests the complete
combination of splitters, in-line filters and xDSL modems with the POTS DC, ringing and
transients.
Why is the testing of xDSL splitters and in-line filters according to TR-127 needed?
- The existing POTS signals were never intended to be xDSL friendly: a 50 V DC battery, a
strong 90 V AC ringing, with large currents and a number of spiky disturbances.
- A POTS splitter or in-line filter can only reduce the POTS transients on the telephone line.
Therefore, xDSL is using digital protection to correct errors, but only TR-127 will test if the
error correction capability is able to keep the xDSL quality sufficiently high (i.e. virtually
error free), when subjected to the POTS signals such as ringing, spiky transients, etc.
- Also non-linear devices or distortion in components inside the splitters can affect the xDSL
transmission, which is only testable according to TR-127.
TR-127 is of critical value in testing the interaction between xDSL and splitters in the presence
of actual POTS signals. This is particularly important in the case of in-line filters.

1.1 Potential Extended Use of TR-127

7KHFRQFHSWRIUHSODFLQJRQH³EDVHOLQH´splitter by another one as the DUT is used in TR-127

and can be extended to other uses. For instance, it is possible to quantize the xDSL capacity
drop due to the non-linearity of surge protections, bridged taps or wire unbalance. The
quantization starts with one of the baseline tests, and instead of replacing one baseline splitter by
a splitter DUT, the extra element to be tested is inserted to the test set-up.
As an example, the effect of an in-house network on a VDSL system with in-line filters can be
quantified. As a first step, the system calibration and the test with baseline filters is executed.

May 2012 © The Broadband Forum. All rights reserved 5 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

As a subsequent step, the extra bridge taps that model the in-house network are added. In this
way, the resulting capacity drops can then be seen.
Note that the full POTS DC and AC ringing voltages are present at the input of the xDSL
modems. This is an important advantage and benefit of the TR-127.

2 xDSL Splitters & I n-Line Filters ± Technical Overview

All xDSL systems are used to deliver broadband acceVVWRWKHLQWHUQHWH[WHQGHGWRGD\WR³triple-
play services´$'6/ZDVILUVWGHYHORSHGWREHDEOHWRUH-use the ubiquitous copper cable plant
all over the world for high speed internet, while keeping the POTS service intact under it. VDSL
continues to be deployed over POTS.
NOTE: The earlier use of voice-band modems to access the internet was not only slower, but
also was incapable of simultaneously telephony and internet access via a single line.

2.1 Ubiquitous Telephone Cables, successfully re-used for xDSL plus POTS

,WLVRIWHQVDLGWKDW³telephone cables have been buried for a long time, but that they are still very
PXFKDOLYH´ ,QGHHGWHOHSKRQHFDEOHVZHUHJLYHQD³VHFRQGOLIH´E\WKHLUUH-use for xDSL.
The main aspect of the re-use of the existing telephone cables for xDSL is that the POTS service
and the new xDSL systems can share a telephone cable, and as such may disturb each other.
Central Office or Cabinet Customer Premises
Low Frequency Low Frequency
Telephony Service
LPF Line Length
LPF Telephony Service
MUX MUX
High Frequency Telephone Line (Twisted Pair) High Frequency
Digital Subsc. Line
HPF HPF Digital Subsc. Line

Figure 1: 0XOWLSOH[LQJWZRVHUYLFHVZLWK³0X[-ILOWHUV´RQDVLQJOHWHOHSKRQHOLQH
The coexistence of POTS and xDSL on a single telephone line (or twisted pair) requires the
FRQFHSWRI³PXOWLSOH[LQJ´VKRZQLQFigure 10XOWLSOH[LQJLVDEEUHYLDWHGWR³PX[´$PX[-
filter is inserted in each signal: a low pass filter (LPF) in the low frequency telephony signal and
a high pass filter (HPF) in the high frequency signals of a Digital Subscriber Line (DSL).
The mux-filters merge and separate the POTS and xDSL signals that are using different
frequency bands, as shown in Figure 2. 7KH/3)LVFDOOHGD³VSOLWWHU´7KH+3)LVDOZD\VSDUW
of the xDSL transceiver.
Power Spectral Density

Splitter (= Low Pass Filter) xDSL High Pass Filter

0 dB
Insertion Loss Insertion Loss
INSERTION
LOSS

í40
dBm ADSL
ADSL
Upstream
§90 VRMS
RINGING

Downstream
BATTERY
§50 VDC

Band
POTS Audio Signal Band
(- 38 dBm/Hz) 40 dB
(- 40 dBm/Hz)
(< 2 mWatt)
Log Freq.
~
~

Hz
DC 50 100 200 400 1k 2k 4k 10 k 20 k 50 k 100 k 200 k 500 k 1M

Figure 2: Spectra of POTS and ADSL merged on a twisted pair with mux-filters

May 2012 © The Broadband Forum. All rights reserved 6 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Figure 3 VKRZVWKH['6/PXOWLSOH[LQJIXQFWLRQLQPRUHGHWDLO7KH/3)RU³VSOLWWHU´LVDGGHG
externally to the POTS system. The HPF is contained in each xDSL transceiver by design.
Two splitters added in
series with the POTS
transmission Customer Premises
Line Length
POTS Central
Office
LPF LPF CPE POTS
Telephone Line (Twisted Pair) Terminal
IN-HOUSE
MUX MUX Equipment
WIRING
ADSL or
VDSL
HPF ADSL or
HPF VDSL
Central Office or Cabinet

Figure 3: $FWXDO³0XOWLSOH[LQJ´3276DQG['6/RQD single twisted pair

Often a single splitter is replaced by multiple parallel in-line filters, as explained in Section 2.4.1.

2.2 Short I ntroduction to the POTS Systems

7KH³3ODLQ2OG7HOHSKRQH6HUYLFH´RU3276VHUYLFHLVXVLQJWKH lower end of the available

³spectrum´RQWKHFRSSHUSDLUOHDYLQJWKHUHVWRIWKHVSHFWUXPIRURWKHUSXUSRVHVVXFKDV['6/
In general the POTS signals contain the following power components and signaling elements:
- A 2 mW POTS audio signal that is transmitted bi-GLUHFWLRQDOO\RYHUD³WZLVWHGSDLU´OLQH
- 7KHDXGLRLVFRPELQHGZLWKD'&³EDWWHU\´YROWDJH § 50 V) feeding the phones.
- A DC current signals the off-hook condition to make a call or accept an incoming call.
- A cadenced AC ringing signal (up to 105 VRMS) is superimposed on the DC battery.
At the left side of Figure 2 is the spectrum of the POTS signals: Battery DC, ringing AC and
POTS audio. It also shows a possible shape of the rejection by the LPF mux-filter, i.e. the POTS
splitter.

2.2.1 Significance of the Ring-Trip Event

Peaks of more than 180 V exist on the line during ringing due the superposition of ringing AC
and battery. When a phone is picked up and the ring-trip occurs during a ringing burst, the result
is the worst possible current in the POTS environment, due to the combination of high ringing
AC and the DC of the battery. The current peaks at 200 mA on short lines, but the duration is
limited normally to less than 200 ms. This can be a real challenge to an xDSL system and
particularly to the splitters.

2.2.2 POTS and in-house wiring as source of disturbance for xDSL:

The POTS signals were never intended to be combined with xDSL on a twisted pair. Therefore,
the POTS signals can be very disturbing for xDSL. The xDSL-unfriendly POTS signals are:
1) battery voltage and its transients, 2) ringing and particularly its cadencing, and 3) DC current
and ringing line current and its transients, particularly the ring-trip. However, xDSL tries to
protect itself against these strong POTS signals and transients. The splitters, the HPF part of the
mux-filters and error correcting techniques are the main elements that protect xDSL. Clearly, the
quality of the splitters is important in the protection of xDSL.

May 2012 © The Broadband Forum. All rights reserved 7 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Besides the intrinsic POTS signals and transients, the in-house POTS wiring and other
components can degrade the xDSL signal in several ways. The in-house wiring (old, poorly
installed, often with long extension cords) and protection devices can cause severe bitrate drops
due to non-linearity, or to bridged taps and unbalance. There is also pickup of RFI and EMI
noises.
NOTE: Many of the bad effects of in-house wiring on the xDSL performance can be
DYRLGHGE\XVLQJDVLQJOH³PDVWHU-VSOLWWHU´LQVWHDGRILQ-line filters.
See Section 2.4.1.

2.3 Overview of xDSL signals on the copper line

The xDSL systems put the high bitrate signals on the existing telephone cable above the POTS
signals, as shown in Figure 2. ADSL signals range between 25 kHz (ITU-T mask for G.992.1)
and 1.1 MHz. Figure 2 shows the ADSL spectrum above the POTS spectrum, together with the
Insertion Loss of the HPF and LPF. Later the bitrate was increased by widening the ADSL
bandwidth for ADSL2plus to 2.2 MHz and then for VDSL2, which pushes the frequency limit to
30 MHz.
Important in xDSL is the existence of error correction techniques, by which the xDSL tries to
fight a number of stationary noises and impulsive noises that enter its line port via all kinds of
paths.
In general xDSL is very POTS-friendly by design. Indeed, there are very few transients in the
xDSL signals, and each xDSL transceiver contains an HPF (i.e. part of the mux-filters), which
makes the xDSL transceiver virtually invisible in the POTS audio band. The HPF is explained in
the next section. Only the strong xDSL signal (i.e. above 25 kHz) on the line can affect the
POTS system, but the POTS system itself is protected by a LPF, i.e. the splitter.

2.4 M ultiplexing an xDSL System on a POTS cable carrying a POTS Signal

As introduced in Figure 2 and Figure 3, xDSL & POTS can share a single twisted pair provided
that multiplexing filters are used: a HPF and a LPF at each side of the line.
These mux-filters are indispensable to allow two parallel services on the line. Indeed, without
mux-filters a POTS system will disrupt xDSL, and vice versa.
These pairs of LPF and HPF mux-filters help in two ways:
- The two mux-filters together assure that the impedance of POTS or xDSL is only active in
the band assigned to each system. Outside its pass band each mux-filters is high-Ohmic.
This renders POTS virtually invisible in the xDSL band and vice versa.
- Furthermore, the filters also prevent that large POTS signals or residues are detectable in the
xDSL transceiver and they assure that xDSL residual signals found in the POTS system are
normally below the noise floor in the POTS system.

2.4.1 6LQJOH³0DVWHU´6SOLWWHUYHUVXV0ultiple I n-Line Filters

The LPF can be implemented as a single POTS splitter. The functional diagram of the xDSL
³VLQJOH´VSOLWWHUFRQILJXUDWLRQLVVKRZQLQ Figure 4. Notice that in the case of a single splitter

May 2012 © The Broadband Forum. All rights reserved 8 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

there is a need to add an extra cable, dedicated to connect the xDSL transceiver to the telephone
line

FAX

SPLITTER SPLITTER IN-HOUSE

POTS Central Line Length WIRING
Office
LPF LPF
Twisted Pair
Modem
ADSL or
VDSL
HPF
NOTE: HPF xDSL
Extra Cable needed
Central Office or Cabinet side to connect xDSL to
+RPHZLWK6LQJOH³0DVWHU´-Splitter
the Telephone Line

Figure 4: )XQFWLRQDOGLDJUDPRIDVLQJOH['6/³PDVWHU´VSOLWWHUFRQILJXUDWLRQ

At the CPE side there is an alternative to the centrally placed splitter: multiple parallel in-line
filters can be installed as shown in Figure 5. 7KH\DUHDOVRNQRZQDV³PLFUR-ILOWHUV´RU
³GLVWULEXWHGILOWHUV´

NOTE:
Multiple In-Line Filters FAX
HDFKODEHOHGZLWK³,/)´
ILF
ILF
SPLITTER IN-HOUSE WIRING
POTS Central
LPF
Line Length ILF
Office Twisted Pair
ILF Modem
ADSL or
VDSL
HPF
HPF xDSL
Central Office or Cabinet side
Home with In-Line Filters

Figure 5: Functional Diagram of the xDSL In-Line Filter configuration

The difference between a single splitter and multiple in-line filters is the position of the in-house
ZLULQJ,IDVLQJOH³PDVWHU´-splitter is used, there is no xDSL signal on the in-house wiring.
With in-line filters xDSL is present on the whole in-house wiring, which is not favorable for its
quality.
I n-line Filter versus Splitters I ssues: Many of the bad effects of in-house wiring on the xDSL
performance can be avoided by using a single splitter instead of in-line filters.
In geQHUDODQ['6/³PDVWHU´VSOLWWHULVSHUIRUPLQJPXFKEHWWHUWKDQ['6/LQ-line filters. The
xDSL rejection of an in-line filter is often only 25 dB. This causes residual xDSL signals to be
audible in telephones. Also residual POTS noises can leak through the typical in-line filter,
which can affect xDSL, often stretching the xDSL error correction beyond its limits.

May 2012 © The Broadband Forum. All rights reserved 9 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

2.5 Disturbance of xDSL by POTS and other noises and disturbances

Section 2.4 explains the multiplexing filters are a HPF in the xDSL modems and the splitter or
the in-line filters in the POTS signal stream. These devices attempt to prevent the POTS noise
DQG3276LPSHGDQFHFKDQJHVIURP³OHDNLQJ´LQWRWKH['6/WUDQVPLVVLRQ
However, some residual POTS noise will unavoidably still leak into the xDSL transmission.
Moreover, several other sources of signals, distortion and noises are present on the in-house CPE
wiring that can affect xDSL.
- DC power (voltage and current), AC ringing (voltage and current).
- POTS Transients, due to cadenced ringing, off-hook and ring-trip.
- EMI and RFI pick-up by the in-house wiring.
- Unbalance due to in-house wiring, and capacity loss due to bridged taps.
- Distortion of protection devices, splitter components and the xDSL HPF.
Static Requirements of xDSL splitters: mainly POTS oriented
When it was attempted to merge ADSL and POTS signals on a twisted pair, the need of the
POTS splitters (or in-line filters) was immediately clear. Standardization bodies such as ETSI
and ATIS have considered a potential detrimental effect of the splitter on the POTS service and
therefore today the static requirements of splitters target mainly the audio quality of POTS.
However, the protection of ADSL or VDSL by the splitters is only specified loosely in the static
splitter standards. Mainly a strong insertion loss and good balance is mandated in the xDSL
band. Starting at around 30 kHz ETSI requires a loss of 45 or 55 dB for splitters, and ATIS
requires even more. Unfortunately the required IL of thH³XELTXLWRXV´LQ-line filters is much
less. Also, besides the loss and balance above 30 kHz, very little is required in terms of
distortion or protection against common mode noises and shielding against POTS induced noises
and transients.
TR-127 fills this lack of requirements in the static specifications of splitters and in-line filters.
Section 3 will introduce the TR-127 methodology.

2.6 Network issues caused by poor quality filters and in-house wiring

If the distortion and non-linearity of the splitters or filters is poor, the xDSL signal will suffer an
increased BER and dropouts during ringing, off-hook and pulse dialing. Off-hook during a
ringing burst is an event that challenges the xDSL error correction at its limits. In the worst case
the xDSL will suffer a full retrain, which will last at least 10 seconds.
3DUWLFXODUO\³PHGLRFUH´LQ-OLQHILOWHUVDUHDOVRNQRZQWR³HDW´['6/ELWUDWHFDSDFLW\$OWKRXJK
the devices might pass the static ETSI or ATIS tests, a single device could already reduce the
bitrate, while every additional device reduces the bitrate even more. Drops due to poorly
designed in-line filters can be stunning. The inadvertent use of ADSL filters within a VDSL
system is known to block the highest part of the VDSL spectrum completely.
Other components known to affect xDSL aggressively are protection devices against lightning
and overvoltage. Their non-linearity can damage the xDSL signals, which causes bitrate
reduction or synchronization failure often during ringing.

May 2012 © The Broadband Forum. All rights reserved 10 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Note also that in-line filters often have only a limited insertion loss. They leak noisy xDSL
residue into the POTS phones. Certain sensitive phones make this noise audible as a hissing
sound due to non-linearity in the POTS phone analog circuitry. This effect is absent for splitters.
When in-line filters are used at the customer premises, the in-house wiring becomes the worst
element attacking the xDSL signal quality: bridged taps cause xDSL signal reflections, which
causes deep comb-shaped dips in the xDSL spectrum. The bitrate loss particularly in VDSL is
often massive. 2QO\WKHXVHRIDVLQJOH³PDVWHU´VSOLWWHUFDQmitigate this problem.
Finally the in-house wiring is shaped as a big antenna, which picks up all kinds of unwanted RFI
and EMI noises, which are transported to the entrance of the xDSL transceiver. This effect is
worst when in-line filters are used. $VLQJOH³PDVWHU´VSOLWWHUJLYHVPRUHSURWHFWLRQDQG
SDUWLFXODUO\ZKHQWKHVSOLWWHUFRQWDLQVD³FRPPRQPRGH´FRLOLWZLOOUHMHFWWKH5),DQG(0,
most effectively.

3 Testing DSL Splitters and I n-Line Filters according to TR-127

This section contains an overview of the TR-127 testing methodology. The main purpose of TR-
127 is to test xDSL Splitters & in-line Filters in a dynamic way. The splitters are tested in a
complete xDSL system, and in the presence of actual POTS signals and disturbances. The
concept of the system test is shown in Figure 6. For the particular cases of POTS noises the TR-
127 contains a complete test suite, which tests the different xDSL components in consecutive
tests.

3.1 Generic interworking testing diagram for xDSL over POTS

CO noise CPE noise

generator generator

Line length
High Pass High Pass
xTU-C Transmission line xTU-R
xDSL CO modem (either real line or line simulator sustaining xDSL CPE modem
actual ringing and DC transients)

High Pass CO noise CPE noise High Pass

xTU-C generator generator
xTU-R
xDSL CO modem xDSL CPE modem

CO POTS model CPE phone model(s)

- DC battery; Line length - DC load; AC load;
CO CPE
- Current limitation; - Off-hook switch;
Splitter Transmission line Splitter
- Ringing AC; - Ringing load;
- Voice band AC; (either real line or line simulator sustaining - Voice band AC;
actual ringing and DC transients)
- Metering injection, ... - Metering load, ...

Figure 6: Generic interworking testing diagram for xDSL over POTS

May 2012 © The Broadband Forum. All rights reserved 11 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Splitters and in-line filters are tested in a dedicated test-xDSL-system. As the splitters and in-
line filters are being tested in this system it is required to prove that all elements of the test-
system itself are working according to same minimum requirements, which will be applied to the
splitters or in-line filters under test.
If the components of the test-system fail to be good enough, the test-system will fail to prove the
quality of the splitters in the end. Indeed, if errors are seen in the splitter test, they should be
caused by the splitters and not by the other components in the test-system. Therefore, the
consecutive tests are first proving that all other parts besides the splitters inside the test-system
are working properly.
Therefore, the general test setup starts with a calibration test shown in the upper part of Figure 6.
With this test the xDSL transceivers are calibrated on the line (a line simulator or an actual
cable). All subsequent tests in the test-suite will be compared to these results.
After the calibration test the remaining tests will include a set of two splitters and the CO and
CPE POTS models, as shown in the lower part of Figure 6. However, again to allow a gradual
check of the system part, the tests with splitters will be done first with two baseline splitters and
then with one baseline splitter and one DUT (Device Under Test). The concept is that the DUT
VSOLWWHUVKRXOGQRWEH³VXEVWDQWLDOO\´ZRUVHWKDQWKHEDVHOLQHVSOLWWHUVXVHGILUVWLQSDLUVLQWhe test
setup.
The tests are a sequence that increases the difficulty for the error correction of the xDSL system
to mask the induced BER. Gradually the tests add a steady DC voltage, a steady DC current, and
a continuous ringing signal. Then transients are added: off-hook transitions, cadenced ringing
and ring-trip events. For all tests a system resynchronization is not allowed when testing either
splitters or inline filters.
For Triple play splitters or filters there is strong requirement that the achievable bitrate should
not drop significantly, that the margin must stay sufficiently high and that the BER is not
allowed to increase above 10-7.
For some splitters or in-OLQHILOWHUVQR³7ULSOH3OD\´TXDOLW\LVUHTXLUHG,QWKDWFDVHWKHQXPEHU
of CRC errors is counted and reported. This allows ranking these devices.

3.2 On-hook to Off-hook ring-trip transition: a Critical Test in TR-127

The most difficult test in the whole test-suite is the ring-trip test. During this test a series of 120
bursts of ringing sent from the CO POTS model to the CPE POTS model (see Figure 6). The
CPE model goes off-hook during each ringing burst. The CO model removes the ringing and the
CPE goes back on-hook. In spite of the 120 ring trip events, the BER must stay below 10 -7 for a
triple play splitter for an interleaved profile. This requirement shows that the splitter is capable
to sustain the strongest possible DC + AC currents on all line lengths.

3.3 The POTS models

To generate the POTS DC, the ringing and the different POTS transients, a CO and a CPE model
were constructed. By attaching these models to an xDSL system as shown in the lower part of
Figure 6, including xDSL modems and splitters, the POTS battery voltage, the off-hook currents
and all necessary transients can be generated.

May 2012 © The Broadband Forum. All rights reserved 12 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

The models are conceived as worst case situations, including (if desired) electro-mechanical
relays to cause a maximum amount of noise. According to each service operator noise and in
partiFXODUO\ULQJLQJFDQKDYHGLIIHUHQWVKDSHVDQGLVVXHVLQWHUPVRILWV³KDUVKQHVV´

4 Conclusions
A number of important aspects motivating the testing of xDSL Splitters and In-Line presented in
MR-261 are:
Successfully passing TR-127 can guarantee a virtually error free performance of a system of
xDSL transceivers including the splitters (or in-line filters) in the presence of an active POTS
service and other actual noises.
Existing static ETSI and ATIS requirements of splitters and in-line filters do not test the dynamic
behavior of splitters and in-line filters in an xDSL test setup.
An xDSL in-KRXVHLQVWDOODWLRQZLWKDVLQJOH³PDVWHU´VSOLWWHULVVXSHULRULQWHUPVRI%(5DQG
capacity of the xDSL, compared to one with multiple in-line filters, even when these in-line
filters are achieving the requirements of TR-127. In-house wiring picks up noise (RFI & EMI)
and contains bridged taps, and splitters reject POTS transients better than in-line filters.
In-line filters, i.e. not meeting the requirements of TR-127, can disrupt xDSL transmission and
cause an increase in the BER, and they can also cause a severe drop in xDSL bitrate. This might
be due to insufficient masking of noise and transients induced by POTS via the in-line filters,
which have a limited rejection (e.g. only 25 dB in ETSI). An alternative cause is the non-linear
effects in components in poor quality in-line filters.
In-line filters with limited insertion loss leak noisy xDSL residue into the POTS phones. Certain
sensitive phones make this noise audible as a hissing sound due to non-linearity in the POTS
phone analog circuitry. This effect is absent for splitters, which have a superior insertion loss.
A single master splitter can be equipped with a common mode rejection coil, which rejects RFI
and EMI noises. This prevents RFI and EMI to travel from the in-house wiring to the entrance
of the xDSL transceiver. In-line filters are incapable of preventing the RFI and EMI induced in
the in-house wiring from entering the xDSL transceiver.

May 2012 © The Broadband Forum. All rights reserved 13 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

5 References and Terminology

5.1 References

Document Title Source Year

[1] TR-127 Dynamic Testing of Splitters and in-line Filters BBF 2010
Issue 2 with xDSL Transceivers

[2] G.992.1 Asymmetric Digital Subscriber Line (ADSL ITU-T 2001

(transceivers.

[3] G.992.3 Asymmetric Digital Subscriber Line ITU-T 2009

transceivers 2 (ADSL2)

[4] G.992.5 Asymmetric Digital Subscriber Line (ADSL) ITU-T 2009

transceivers - Extended bandwidth ADSL2
(ADSL2plus)

[5] G.993.1 Very high speed Digital Subscriber Line ITU-T 2004
transceivers (VDSL)

[6] G.993.2 Very high speed Digital Subscriber Line ITU-T 2006
transceivers 2 (VDSL2)

[7] ETSI TS 101 Access network xDSL splitters for European ETSI 2009
952-1 deployment; Part 1: Generic specification of
xDSL over POTS splitters.

[8] ETSI TS-101 Access, Terminals, Transmission and ETSI 2012

952-3 Multiplexing (ATTM); Access network xDSL
splitters for European deployment; Part 3:
Generic specification of static distributed filters
for xDSL over POTS

[9] ATIS- Network End POTS Splitter Requirements. ATIS 2010

0600026.2010
[10] ATIS- Remote End POTS Splitter Requirements. ATIS 2008
0600016.2008
[11] T1.421-2001 In-Line Filter for use with Voiceband Terminal ATIS 2001
Equipment Operating on the Same Wire Pair
with High Frequency.

[12] ATIS- Splitters Used for Line Splitting and Line ATIS 2009
TRQ.10.2009 Sharing Applications

May 2012 © The Broadband Forum. All rights reserved 14 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

5.2 Definitions

Baseline Filter A splitter that is used to establish the reference levels before adding the DUT.
Distributed filter Alternative name for In-Line Filter.
In-Line Filter Individual filters with a splitter function, but protecting each individual
telephone device at the Customer Premises.
ISDN Integrated Services Digital Network, and narrow band DSL system, used for
digital telephony and switched digital 64 kb/s channels.
Micro-filter Alternative name for In-Line Filter.
Master Splitter Splitter that is used to isolate xDSL frequencies from POTS frequencies at a
VLQJOHSRLQW RIWHQFDOOHG173RU1,' DWWKHFXVWRPHU¶VSUHPLVHV.
Multiplexing Merging two services onto a single transmission medium (e.g. a telephone line)
at one end transmitting the together over the medium, to separate them at the
other end.
Network Interface The NID is a device that serves as the demarcation point between the carrier's
Device (NID) local loop and the customer's premises wiring. It is placed at the NTP.
Network The NTP is the demarcation point between the access pair and the in-house
Termination Point wiring, where often the master splitter is placed. The NID is placed here.
(NTP)
Off-hook State of the POTS equipment at either end of a loop connection when the CPE
equipment is in active state, to make a telephone call.
On-hook State of the POTS equipment at either end of a POTS loop connection when the
NTP terminal equipment is in the quiescent state.
Plain Old Telephone The traditional Public Switched Telephone Network.
System
Ring-trip The process of stopping the AC ringing at the Central Office when the telephone
being rung is answered.

5.3 Abbreviations

ADSL Asymmetrical DSL

ATIS Alliance for Telecommunications Industry Standards
BER Bit Error Ratio
CO Central Office
CPE Customer Premises Equipment
CRC Cyclic Redundancy Check
DSL Digital Subscriber Line
DSLAM DSL Access Multiplexer

May 2012 © The Broadband Forum. All rights reserved 15 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

DUT Device Under Test

EMI Electro-Magnetic Interference
ETSI European Telecommunications Standards Institute
HPF High Pass Filter, i.e. a filter blocking DC and low frequencies
IL Insertion Loss
IPTV Internet Protocol Television
ISDN Integrated Services Digital Network,
ITU International Telecommunications Union
LPF Low Pass Filter, i.e. a filter blocking the upper part of the spectrum
MUX Multiplexing or Multiplex
NID Network Interface Device
NTP Network Termination Point
POTS Plain Old Telephone System a.k.a. PSTN
PSTN Public Switched Telephone Network = POTS
RFI Radio Frequency Interference
RMS or rms Root Mean Square
TR Technical Report
VDSL Very High Speed DSL
VoIP Voice over Internet Protocol
xDSL ADSL or VDSL

May 2012 © The Broadband Forum. All rights reserved 16 of 17

Motivation for Testing xDSL Splitters and In-Line Filters MR-261 Issue 1

Notice
The Broadband Forum is a non-profit corporation organized to create guidelines for broadband
network system development and deployment. This Broadband Forum Marketing Report has
been approved by members of the Forum. This Broadband Forum Marketing Report is not
binding on the Broadband Forum, any of its members, or any developer or service provider. This
Broadband Forum Marketing Report is subject to change, but only with approval of members of
the Forum. This Marketing Report is copyrighted by the Broadband Forum, and all rights are
reserved. Portions of this Marketing Report may be copyrighted by Broadband Forum members.
This Broadband Forum Marketing Report is provided AS IS, WITH ALL FAULTS. ANY
PERSON HOLDING A COPYRIGHT IN THIS BROADBAND FORUM MARKETING
REPORT, OR ANY PORTION THEREOF, DISCLAIMS TO THE FULLEST EXTENT
PERMITTED BY LAW ANY REPRESENTATION OR WARRANTY, EXPRESS OR
IMPLIED, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTY:
(A) OF ACCURACY, COMPLETENESS, MERCHANTABILITY, FITNESS FOR A
PARTICULAR PURPOSE, NON-INFRINGEMENT, OR TITLE;
(B) THAT THE CONTENTS OF THIS BROADBAND FORUM MARKETING REPORT
ARE SUITABLE FOR ANY PURPOSE, EVEN IF THAT PURPOSE IS KNOWN TO
THE COPYRIGHT HOLDER;
(C) THAT THE IMPLEMENTATION OF THE CONTENTS OF THE MARKETING
REPORT WILL NOT INFRINGE ANY THIRD PARTY PATENTS, COPYRIGHTS,
TRADEMARKS OR OTHER RIGHTS.
By using this Broadband Forum Marketing Report, users acknowledge that implementation may
require licenses to patents. The Broadband Forum encourages but does not require its members
to identify such patents. For a list of declarations made by Broadband Forum member
companies, please see http://www.broadband-forum.org. No assurance is given that licenses to
patents necessary to implement this Marketing Report will be available for license at all or on
reasonable and non-discriminatory terms.
ANY PERSON HOLDING A COPYRIGHT IN THIS BROADBAND FORUM MARKETING
REPORT, OR ANY PORTION THEREOF, DISCLAIMS TO THE FULLEST EXTENT
PERMITTED BY LAW (A) ANY LIABILITY (INCLUDING DIRECT, INDIRECT, SPECIAL,
OR CONSEQUENTIAL DAMAGES UNDER ANY LEGAL THEORY) ARISING FROM OR
RELATED TO THE USE OF OR RELIANCE UPON THIS MARKETING REPORT; AND (B)
ANY OBLIGATION TO UPDATE OR CORRECT THIS MARKETING REPORT.
Broadband Forum Marketing Reports may be copied, downloaded, stored on a server or
otherwise re-distributed in their entirety only, and may not be modified without the advance
written permission of the Broadband Forum.

The text of this notice must be included in all copies of this Broadband Forum Marketing Report.

End of Broadband Forum Marketing Report MR-261

May 2012 © The Broadband Forum. All rights reserved 17 of 17