Four wires on a T1 question.
Sean Lerner
Question:
I understand that a T1 has four wires:
Transmit - Tip
Transmit - Ring
Receive - Tip
Receive - Ring
My question is, when hooking the T1 up, does it matter if Transmit -
Tip and Transmit - Ring are reversed? Ditto with Receive - Tip and
Receive - Ring?
Is it okay to have them reversed if the T1 is B8ZS, but not okay if
it's AMI?
Thanks!
Sean
Yoram Henik
Sean,
No problem with crossing the Tip and Ring.
The T1 line code is always AMI. B8ZS or B7 are zero suppression codes
that are used to ensure a minimum 1's density on the line.
With AMI a "1" is transmitted as a positive or negative pulse on the
line (Alternatly), a "0" is transmitted as "no pulse". As you can see
from this definition, there is no problem with inverting the signal
(=crossing Tip & ring) on the line.
--
Yoram Henik TEL: +972-3-6455452
RAD Data communications FAX: +972-3-6475924
31 Habarzel st. Email: he...@radmail.rad.co.il
Tel-Aviv 69710, ISRAEL
Al Gillis
Hi Sean -
I don't believe most DS-1 equipment (CSUs, PBXs, muxes of one sort or
another) are polarity sensitive - that means tip-ring reversals won't
cause any operational problems (Transmit and receive must be wired
properly, of course).
The consternation these reversals WILL cause, however, is that normal
wiring conventions will not be observed if you wire things up with
reversals. That means jumpers will have reversals, plugs or jacks will
be wired "wrong" or cables will be cut down on blocks wrong. This will
only lead to problems in the future when someone not aware of the first
reversal tries to duplicate, replace or copy some of the weird wiring.
I'd stick to doing it the conventional way and you'll avoid problems
tommorrow (but it will work!).
Wayne E. Amacher
In CCITT recommendation G.703 there is a provision recommending that the
transmit coax shield (ring) be earth grounded at the source. There is an
option to earth ground the shield at both ends. I know that this holds
for E1 lines. I don't have a copy of G.703 handy to confirm that it
holds for T1 also. If both ends have the grounding jumper installed, tip
and ring cannot be reversed. If both ends are not grounded and the
inputs are transformer coupled, then it should make no difference that
tip and ring are reversed.
Wayne E. Amacher
wama...@rahul.net
Yoram Henik (he...@radmail.rad.co.il) wrote:
: Sean Lerner wrote:
: >
: > Question:
: >
: > I understand that a T1 has four wires:
: >
: > Transmit - Tip
: > Transmit - Ring
: > Receive - Tip
: > Receive - Ring
: >
: > My question is, when hooking the T1 up, does it matter if Transmit -
: > Tip and Transmit - Ring are reversed? Ditto with Receive - Tip and
: > Receive - Ring?
: >
: > Is it okay to have them reversed if the T1 is B8ZS, but not okay if
: > it's AMI?
: >
: > Thanks!
: >
: > Sean
: Sean,
fon...@esslink.com
what is wrong, and in checking the wiring finds that his/her signals trace out wrong.
Best bet is to do it right the first time!! As a field service tech and trainer, I find the
biggest problems in the field are caused when some installer got frustrated and
couldn't figure out how to do the job correctly (following the standards) and decided
to just, "make it work." In most cases, it will not cause a problem, but if there is ever
a problem, having additional variables (such as miswires) makes troubleshooting
and diagnosis more time consuming, and thus more expensive for the customer. (of
course with the overall decrease in service & equipment costs, hell they can afford
it) [hey, easy, it was a joke]
Regards,
TimP
robert milton
HEY,LOOK HERE NOW.idisagree that all T1 ckts are AMI they are either
AMI whice is old technology or B8ZS whice is now the most widely
used.With ISDN ckts most are configured B8ZS/ESF .B8ZS ensures 1s
density requirementson the network by replacing an all 0s byte with a
specific byte containing two intentional bipolar violations.the framing
must be compatable with the carrier provided service.Most CSUs cam
convert AMI line coding to B8ZS to be compatibale with older COs.
Hope I haven't bored you to much. Good luck, Robert Milton
Steve Pinkston
se...@ican.net (Sean Lerner) wrote:
>Question:
>I understand that a T1 has four wires:
>Transmit - Tip
>Transmit - Ring
>Receive - Tip
>Receive - Ring
>My question is, when hooking the T1 up, does it matter if Transmit -
>Tip and Transmit - Ring are reversed? Ditto with Receive - Tip and
>Receive - Ring?
>Is it okay to have them reversed if the T1 is B8ZS, but not okay if
>it's AMI?
>Thanks!
>Sean
There is absolutely no problem with reversing Tip and Ring when
hooking up T1/DS1 wiring. Make sure you don't reverse transmit
and receive, or split the pairs!
The 3 biggest problems I see when people are hooking up customer
premises wiring or extended demarcs for T1 are:
1. Use of unshielded twisted pair (UTP) wiring. This is the
easiest way I know of to create bipolar violations (BPVs)
on the line. Deadly to data, annoying to voice. Use AT&T
type ABAM cable or its equivalent; this type of twisted
pair cable has separate shields for the transmit and receive
pairs. The shields should be grounded at one end only.
2. Incorrect pinning of the RJ-48 (USOC 8-pin) jack. The
standard way of connecting T1 is:
1 - rx tip (from telco)
2 - rx ring (from telco)
4 - tx tip (from CPE)
5 - tx ring (from CPE)
3. Less-than-ideal connections: when in doubt, reterminate!
- steve
----------------------------------------------------------------------------
Steve Pinkston (pink...@kentrox.com) | sup...@kentrox.com
Technical Support Supervisor | http://www.kentrox.com
ADC Kentrox, Portland, Oregon, USA | 1-503-643-1681 ext 6341
----------------------------------------------------------------------------
Karl Denninger
In article <1996Jul8.2...@kentrox.com>,
Steve Pinkston <pink...@kentrox.com> wrote:
>
>There is absolutely no problem with reversing Tip and Ring when
>hooking up T1/DS1 wiring. Make sure you don't reverse transmit
>and receive, or split the pairs!
>
>The 3 biggest problems I see when people are hooking up customer
>premises wiring or extended demarcs for T1 are:
>
>1. Use of unshielded twisted pair (UTP) wiring. This is the
>easiest way I know of to create bipolar violations (BPVs)
>on the line. Deadly to data, annoying to voice. Use AT&T
>type ABAM cable or its equivalent; this type of twisted
>pair cable has separate shields for the transmit and receive
>pairs. The shields should be grounded at one end only.
Nonsense. I have run HUNDREDS of feet of quality UTP on DS-1s without
trouble. T1s are delivered as a BALANCED signal. There is no problem
with running this over UTP; in fact, balanced signals should generally
*be* run over UTP (10BaseT Ethernet is a balanced signal as well; it is
spec'd over UTP precisely for noise rejection reasons).
Look at the riser cable in 95% of the large buildings out there. Its UTP
(and crappy Cat-3 -- maybe -- at that) and yet it still works just fine for
DS1 signals. Just don't stick it on top of a fluorescent light fixture.
ABAM cable is good too, but its MUCH larger and more bulky. The primary
problem with ABAM cable is that people FREQUENTLY ground both ends of the
shield, and that can cause serious problems (ground loops, and in the case
of runs between points not at a common potential, VERY significant currents
can be developed.)
I dislike shielded cable for balanced signal use because of its much higher
capacitance; this tends to "round" the corners on signal transitions, and
that, over long distances, leads to degredation of the waveform. At DS1
speeds its not significant, but this is why STP *will not work* for
10BaseT Ethernet.
Our entire plant here is CAT5 UTP terminated in accordance with the Cat5
standards (no more than 1/2" untwisted at connectors, etc). We have *NEVER*
taken a single BPV or framing error on most of these circuits (we go
directly from a Fuji FLM150 to Cat5 to CSU RJ48 plug.)
>2. Incorrect pinning of the RJ-48 (USOC 8-pin) jack. The
>standard way of connecting T1 is:
>
> 1 - rx tip (from telco)
> 2 - rx ring (from telco)
>
> 4 - tx tip (from CPE)
> 5 - tx ring (from CPE)
I'll add to this: LEAVE THE OTHER PINS OPEN UNLESS YOU *KNOW* THAT THE JACKS
YOU ARE PLUGGING INTO HAVE NO CONNECTION ON THE OTHER PINS INTERNALLY. It
is very common to have some of the other pins shorted to ground; you can
EASILY create ground loops inadvertantly this way if you use common CAT5
10BaseT jumper cables.
It is generally safe to use these as a connection from the telco Demarc
on a RJ48X to a CSU/DSU; the reason is that the telco almost always leaves
all other pins open on the RJ48X. If in doubt, however, pop the cover and
LOOK.
>3. Less-than-ideal connections: when in doubt, reterminate!
Yep. Bad crimps on the connectors will hose you good. Use a QUALITY
crimp tool and quality RJ connectors, and be careful not to nick the
wires when removing the outer jacket on the cable. The most common cause
of an open pair is a nicked wire during jacket removal; as that flexes
it will eventually fracture and leave you with an open circuit.
--
--
Karl Denninger (ka...@MCS.Net)| MCSNet - The Finest Internet Connectivity
| T1 from $600 monthly; speeds to DS-3 available
| 23 Chicagoland Prefixes, 13 ISDN, much more
Voice: [+1 312 803-MCS1] | Email to "in...@mcs.net" WWW: http://www.mcs.net/
Fax: [+1 312 248-9865] | Home of Chicago's only FULL Clarinet feed!
Cliff Liles
ka...@MCS.COM (Karl Denninger) wrote:
Karl,
Your explanation of how you do your T1s identifies why you don't have
problems. It is still not good T1 Engineering however. The ballanced
nature of T1 is important, and the reason you can reverse tip and ring
with impunity.
However, the reason for shielded cable on T1 is a different matter
entirely. One of the most common causes of T1 line errors is "Near
End Cross Talk" (NEXT). This is generated when a high level signal in
the transmit pair is not shielded from the low level (attenuated by
distance traveled) signal in the receive pair. The level differences
of these signals can be 15db or more appart. When this occurs, it
doesn't take a very long unshielded cable run (a few feet) for noise
to propagate into the receive (low level) signal. Using any
unshielded cable (even CAT 5) WILL cause NEXT if the signal level
difference is greater than 7.5 db, over any significant cable length.
The riser cable is done by isolating the transmit pairs (high level)
and receive pairs (low level) into seperate bundles. This keeps the
pair to pair level differences low, and minimizes the NEXT.
In the case you describe, the signal coming out of your Mux is approx
0db, and the signal coming out of the CSU is approx 0db, hense little
or no level differences. This permits the use of UTP wiring, which is
how DSX (Digital Signal Cross Connects) are wired. The DSX is a
special case that allows almost unlimited numbers of DS1s to come
together and be connected without NEXT. This is due entirely to the
characteristic that all signals coming into the DSX are the same
level.
However, most metalic (not delivered by multiplex system) T1 lines
will have significant signal level differences, hence the standard
that T1 lines use shielded cable. Short of physically isolating the
transmit and receive pairs (seperation), shielding is the only way for
T1 users to guarrantee noise immunity.
Cliff Liles
Fred R. Goldstein
>HEY,LOOK HERE NOW.idisagree that all T1 ckts are AMI they are either
>AMI whice is old technology or B8ZS whice is now the most widely
>used.With ISDN ckts most are configured B8ZS/ESF .B8ZS ensures 1s
>density requirementson the network by replacing an all 0s byte with a
>specific byte containing two intentional bipolar violations.the framing
>must be compatable with the carrier provided service.Most CSUs cam
>convert AMI line coding to B8ZS to be compatibale with older COs.
>Hope I haven't bored you to much. Good luck, Robert Milton
Well, I hate to disagree with your disagreement, but it comes down to
semantics & terminology.
In practice, all T1 is AMI. However, there are two flavors. "Classic" AMI is
usually treated as Zero Code Suppression (AMI/ZCS), though it's really a rule
against more than 15 consecutive zeros, not specifically the all-zero octet.
The preferred form is AMI/B8ZS, where AMI is still the discipline but it has
the B8ZS enhancement. The former is usually abbreviated "AMI" and the latter
"B8ZS" but they're both forms of AMI. And if you wanna be picky, they're
AMI/RZ (return to zero). The MDB code used on the BRI S/T bus is close to AMI
but inverted and not the same RZ waveform.
--
Fred R. Goldstein k1io fgold...@bbn.com
BBN Corp., Cambridge MA USA +1 617 873 3850
Opinions are mine alone; sharing requires permission.
Fred R. Goldstein
>However, the reason for shielded cable on T1 is a different matter
>entirely. One of the most common causes of T1 line errors is "Near
>End Cross Talk" (NEXT). This is generated when a high level signal in
>the transmit pair is not shielded from the low level (attenuated by
>distance traveled) signal in the receive pair...
>The riser cable is done by isolating the transmit pairs (high level)
>and receive pairs (low level) into seperate bundles. This keeps the
>pair to pair level differences low, and minimizes the NEXT.
That's a good point. In practice, the rules for installing T1 wires on
outside plant local loops (aerial/underground wire) do the same thing in order
to minimize NEXT. You can't run T1 on 25-pair cable. Ordinary phone cable is
bundled 25 pair at a time. You run the transmit side in one bundle and the
receive side in another, so you need a minimum of 50 pair on the pole.
T1 is a very primitive technology (invented 1960) compared to, say, ISDN's
local loop which is full duplex on one pair, or newer xDSL technologies.
Stephen Satchell
In article <1996Jul9.2...@kentrox.com>, lil...@kentrox.com (Cliff
Liles) wrote:
> However, the reason for shielded cable on T1 is a different matter
> entirely. One of the most common causes of T1 line errors is "Near
> End Cross Talk" (NEXT). This is generated when a high level signal in
> the transmit pair is not shielded from the low level (attenuated by
> distance traveled) signal in the receive pair. The level differences
> of these signals can be 15db or more appart. When this occurs, it
> doesn't take a very long unshielded cable run (a few feet) for noise
> to propagate into the receive (low level) signal. Using any
> unshielded cable (even CAT 5) WILL cause NEXT if the signal level
> difference is greater than 7.5 db, over any significant cable length.
Now wait a minute -- part of the problem is pair choice. If you choose
two pairs that are twisted in opposite directions, that takes care of most
of the NEXT problems that you see. Of course, if you have multiple
signals in the same cable, you have to make sure that the transmit use one
twist direction and receive use the opposite twist direction.
I've measured crosstalk in 10 base T cabling, and it's somewhere around 50
db attenuation between alternately-twisted pair. I haven't tried
same-direction pairs yet...
--
Stephen Satchell, Satchell Evaluations
http://www.accutek.com/~satchell
Fanaticism consists of redoubling your effort when you have forgotten your aim -- George Santayana