atcomsystems.ca/forum Forums Wiring - Telecom Carriers - Trunks & Services T1's, CSU's and DSU's 2 Wire T1 vs. 4 Wire T1

Last Saturday night we had a T-1 go down. About 2:30 a.m. the AT&T tech got it back up, but as he was leaving we found out that he had converted the circuit from a 4-Wire T-1 to a 2-Wire T-1. About 5 years ago we had problems with the 2-wire circuits and made sure all our loops were a true 4-wire. We're now having trouble getting the carrier to switch the loop back.

I'd like to get some input from our T-1 gurus about this. I plan on setting up a large simultaneous upload and download sometime this week to see how it handles the transfer.

A 2-Wire T-1! In my experience a T-1 was always defined (and tariffed!) as a FOUR wire circuit, 1 pair transmit and 1 pair receive, with a bandwidth of 1.544 mbs.

Anything else, was just that - something else.

What is this, some kind of DSL type circuit?

Sam

A follow up. (I guess I hit the "Add reply" key to quickly).

I would definitely follow up with some serious stress testing. We used to use Sunsets or T-Birds, and our routine to certify a new circuit used to be:

5 minutes of all 1's
15 minutes of all 0's
15 minutes of 3 in 24
24 hours of QRS

If it passes that then maybe it doesn't matter what it looks like. But how do you interface to it? It can't appear on an RJ-48X or a DB-15. Do they use some sort of smart jack that converts it to a 4 wire circuit for you to connect to?

Sam

That was my arguement too Sam. I claimed that this was "more DSL than T-1," but that is allegedly what AT&T is installing around here. The CLEC claims that they have no power over what type of circuit the ILEC supplies as long as the data transfer meets spec. We do have a straight line back to the cloud, and right now we're getting our synchronous speeds of 1.5, (minus reasonable overhead,) but my research on these circuits leads me to believe that when the office is full, and users are both up and downloading at the same time we're going to see problems.

To answer your second post Sam, it's an integrated T-1. It comes into a smart jack outside, then in the phone room it hands off to an IAD which breaks out the voice into analog ports and hands the data off to the firewall.

It appears to me to be more of an HDSL, but I'm finding a lot of talk on the net where people are facing the same problem with them refusing the extra pair.

From what I can find it is HDSL. Like Sam in my day a T-1 was 4 wire. So this is new to me also. From reading some about it the speed and reliability is comparable to a full T-1 span. If it's working fine I'd just make sure I wasn't still paying for a full 4 wire span. Maybe one of our CO guys can shed more light on it.

We dont have T1 here but we do have 30 channel PRI, which is basicaly a 2 meg link, they normaly use 1 pair for this circuit, but if the distance is over 2 miles from exchange they use 2 pairs circuits and re-configure the modem.

But our interface to the modem is always 2 pairs

It's definitely HDSL that comes in on one pair to an Adtran (or other brand) network termination mounting. On the CPE side of the card, it's your standard RJ48X connection with separate transmit/receive.

The PRI at my office with integrated access has been here and working for over three years with zero problems. We have a sub tenant who has his own T1 that's set up exactly the same way, also with no bandwidth problems. Verizon is using ADC Higain H2TUR402 cards for both circuits. I haven't seen circuits installed any differently around here in several years.

Ed my concern was that since you don't have a different pair for both upload and download, there would be issues when a few users are running uploads at the same time others are running downloads? I suppose a single pair can handle it since speeds of upwards of 3 Megs can be obtained on a 2 wire, but I'm having trouble with calling a single pair circuit a T-1. On second thought maybe it's that I'm having trouble paying for a 2 wire circuit that's called a T-1.

All I ever see around here are two-wire HDSL T-1's. Of course they become 4-wire at the smart jack.

Do you see them being as reliable Larry? I guess this is just something newer to this area.

I know what you mean, Trace. I asked the same question the first time that I saw it. It technically does provide separate transmit and receive paths, just at different frequencies over the same pair. Very similar to ADSL, where you can have voice and DSL riding on the same pair.

As far as paying, well they technically are providing you with a four-wire circuit. How they emulate it is up to them. I often say the same thing when a carrier puts in a channel bank for 24 POTS lines, but still charges for the POTS lines as if they were all on separate pairs. That's where the telcos make the real money, by reducing transport costs.

HDSL2 technology (2-wire) (whether it be Adtran or ADC the most common equipment manufactures) is just as reliable as a 4-wire T1 delivery method. As a matter of fact it’s MORE reliable than say transmitting and receiving DSX format over copper spans as was done years ago.

This is not like ADSL what so ever in that the bandwidth is synched and locked at 1.544 Mbps transmit and receive. You WILL NOT have bandwidth capacity issues with it. If the T1 can not run at 1.544 Mbps it has lost synch and it’s down.

A T1 is still a 4-wire service as far as billing and tariff goes as well… There’s a 4-wire interface on the customer side of the H2TU-R and it DOES operate “full-duplex” over the one pair of wires.

Works like a champ!

I see a lot of them here as well...HDSL and HDSL2.
Your physical connection is still the same though.

That answers a couple of my other questions Bryan. Thanks for the reply.

When AT&T first started supplying the 2-wire circuits to Birch in this market, Birch had so many problems with them that they started refusing any loop that wasn't a 4-wire in Dallas. That was five years ago however. It sounds like they have really gotten the technology cleaned up.

Well I suppose since it actually is full duplex and has been signed off on by the Tech Talk T1 Braintrust I'll stop whining about it.

Thanks for sharing all the knowledge guys.

:toast:

I like your idea of the large simultanious (bi-directional) transfer across the HDSL.

I agree that HDSL at minimum, is equally as reliable as a regular 4-wire T1. Now that I think of it, Bryan brings up a great point that HDSLs are either UP, or their DOWN.

I'd like to share a quick follow-on point about ADSL Service: Here locally, ADSL is a DSL service that is assigned to an ordinary POTS line. The POTS line is assigned no differently in the switch, with exception that the pair passes thru an ADSL shelf prior to leaving the CO.

Been waiting for you Mike. My lead tech was waiting for a switch operator to chime in before he was ready to concede.

At least from a maintenance perspective (and it's strictly my personal opinion), I think that AT&T's switching your service over to an HDSL circuit could turn out to be a plus. Was there any history of re-occuring issues with the original T1? Now, as far as the cost between the T1 and HDSL? That's definately worth looking into. It does sound like what Ed mentioned with the 24 POTS / Channelized T1 / Channel bank example. I wonder if anyone else has experienced this particular type of "automatic" service switch by AT&T. This could be a new trend..I wonder if Anthony is seing this up in his area.. or if this is standard for AT&T..I wish I knew.

Birch Telecom; I remember (without knowing the details) that Birch went thru quite a tough struggle a couple years ago. Especially with their operating right there in the back yards of SBC, BellSouth, AT&T, I know that Birch has emerged from some very, very, tough times.

Sorry for my recent lag time with responses. This is of those time periods where I have to really struggle to get here. (It's not so much as time, as it is physical access.) I'll have to join the nightowls here until my day schedule "fixes itself".. :thumb:

Darn, you caught me ramblin..

From a carrier point of view, I don't think you have anything to worry about. I work for a large CLEC and we have these installed all over our 15 state footprint. We have VERY few issues that I would say are specifically related to the circuit being a 2wire.

I wouldn't get wrapped around the axle about it being called a 2-wire circuit because whether it's a 2wire or a 4wire, 99% of the time, it will still be an HDSL circuit. IM me on yahoo if you have more questions, and I would be happy to go into more detail or answer more questiong that you have.

Thanks Mike (Megatel). We have a lot of respect for Mike (5Etek-mike) and he speaks very highly of you. Welcome to the board.



Man, it's getting so you can't tell all the Mikes around here without a program.

OK one of you CO Mike types. (Like Trace said) Any difference in trouble shooting these?

No real difference troubleshooting them. We don't do anything differently for looping and sending patterns across them. Effectively there is no difference between them. The 2wire and 4wire only refers to the span between the HTUC card in the office and the HTUR card at the site.

Thanks. I assumed that was the case, but we all know what assuming can do.

One notable difference is that there are no repeaters on HDSL circuits. So in that sense, they provide one less point of failure over a traditional 4-wire T1.

In the Chicagoland area we refer to the AT&T Techs.that service these circuits as "DOGS",
as they are in the Digital Operations Group.
I've often heard them refer to the single pair T1
as a "high voltage T1", I have no idea why as it's not like the old Western Union Trunks that could knock you on your butt.
I do know that when bundled with DSL it can knock out the DSL. We found that out the hard way when they ran a single pair T1 through the same 25 pair UG that had several DSL lines in it.
This was to a small strip mall.
Had to set up a vendor meet.
Disconnected the T1 at the CO and the 3 DSL lines came back up, tied it down the 3 DSL
all went out again.
Changed the T1 to a 2 pair all was happy.
It was an old time Tech. who suggested this as he had seen it before.
The CLEC, (XO) claims it's cheaper for them to use the single pair where they can, I'm not sure if this is true but I was also told by the AT&T Tech. that they do have to be fairly close to the CO for them to use a single pair.
I thought all this was interesting until it cost me 2 days screwing around on my time.
---------------------
MrG

Good points... True there are no repeaters on the circuits, but they do have something very similar called a doubler. Same beast that does the same thing, just works differently.

What they are referring to as "high voltage" means that there is "span" power on it. Off the top of my head I'm going to say about 130VDC. Pretty hot and will knock you on your butt. Distance from the CO for these is not really an issue because of the doublers. I've seen circuits that had 5 or 6 doublers on them.

As far as the cost, I don't think there is a price difference on the circuits. IMHO, it's just a matter of preference or availability of pairs or equipment for the carrier.

The term “high-voltage” is referring to T1 that has a repeater. The voltage is for powering the repeater in the span.

The circumstances you talk about in knocking down ADSL service, was exclusively a problem when trying to REPEATER 2-wire HDSL loops. It has since been determined that if loop length is too great to reach the H2TU-R from the C.O. with 2-wire, (roughly 9 Kft of 26 gage) then using a repeater is NOT an option. (Adtran actually discontinued producing the H2RE ADSL2 repeaters for this reason.) Adtran has since developed and starting producing the 3rd generation of HDSL … HDSL4 which can reach farther and be safely repeatered.

Here’s a quicky look at HDSL technologies out there:

1st generation:
HDSL
4-wire loop
Max loop length 9 Kft of 26 ga. or 12 Kft of 24 ga. without being repeatered
HRE repeaters can interrupt ADSL signal in the same binder group if the ADSL loop is greater than or equal to 10.5 Kft.

2nd generation:
HDSL2
2-wire loop
Max loop length 9 Kft of 26 ga. or 12 Kft of 24 ga. without being repeatered
H2RE repeaters can interrupt ADSL signal in the same binder group if the ADSL loop is greater than or equal to 10.5 Kft.


3rd generation:
HDSL4
4-wire loop
Max loop length 12 Kft of 26 ga. or 16 Kft of 24 ga. without being repeatered
ADSL spectrally compatible with repeaters.

Span power on Adtran units is 190 vDC.

I’m in the position where I’m the one to choose what T1 technology is deployed on a per circuit basis in my areas for the lil’ ol’ LEC I work for… I ALWAYS try to use HDSL2 first if it’s an option. That saves a pair (which is a huge deal) and with only one pair, it halfs the probability of a pair going bad.

As far as what the hardware costs… The HDSL4 is slightly more expensive per unit, so that’s another win for HDSL2 when is a viable option.

Hey Trace now you know who to call when you have problems. Good information guys, thanks. :thumb:

That's exactly the same thing I was thinking. Trace not only got a bunch of good answers, but he asked for the time and learned how to build the watch. Good job with some really high-quality answers gang.

Well guys, since this thread is going to become the definitive reference, I want to get a response on two other points.

First, since there are no repeaters, is this distance sensitive?. For example where I live, I have a T at my house because that's all I could get from a distance standpoint. Can this other circuit be ran to a cabin on the top of a mountian many miles from the CO?

Also, speaking of the high voltage issue. I have been told in the past that on a circuit such as this, it could be put in today and work fine. Then as the area around you built up, if you ended up with several of these in the same cable, your service got degradated. Is there any truth to that?

Trace, I can only speak from personal experience. My office is located about 9 miles from our serving CO and there is one underground "hut" along the cable route. Chances are, there is a doubler there. We lost our circuit about a year ago and I was told that it was due to a power supply failure in this hut. This implies that there must be some type of electronic device associated with our circuit in the rack that lost power.

From what I've been told by Verizon technicians, our circuit is on a copper pair the whole way. I know of at least six of these circuits that are working in our office park cul-de-sac, all riding on the same 900 pair distribution cable down our street with no issues whatsoever.

The feed to this general area is comprised of two aerial 2,700 pair 26 gauge cables and one 2,400 pair, 26 gauge cable. There are certainly visible repeater cans and load coils along the route, but they are obviously for unrelated pairs to mine.

I've tried to beat up our circuit a couple of times by shorting or grounding the pair and it always comes right back up. Our distribution cable is 1970's era PIC, so it's not in the best of shape. With that being said, I'd say that it's safe to assume that these circuits are fairly bulletproof. :rolleyes:

Guys -

A lot of good information on this post! This one is a keeper.


Sam

Trace,
HDSL whichever form/generation you use IS distance sensitive. For HDSL2 (2-wire) what sparked your original question, the maximum is 9 Kft for 26 gage; 12 Kft for 24 gage, and 22 Kft on barbwire (19 gage.) Other factors DO play a role in it too. I.E. How much is buried (constant 70 degree F temp) and whether the cable is pic (plastic insulators) or pulp (paper insulation.) So your cabin on the mountain “could” get T1 service on 2-wire if it had good 19 gage cable and was within’ 4 miles or so of a remote. … BUT more than likely I’d be using HDSL4 with up to two repeaters to get to your country home with a view. As long as you was within’ say 7.5 miles from a remote (calculated on 24 gage)

Is there any truth to service being degradated if other similar services were added after yours? …. Uhhhh… My professional opinion is, “kind-a.” LOL. For your HDSL2, no. You really can run a sheath full of those. When applying HSDL 4-wire technology you’re supposed to apply separations to the pair assignments. (Keeping transmit and receive pair out of the SAME 25-pair binder.) If you start running a bunch of Tx’s and Rx’s together in the same 25 count… there’s a CHANCE things will get a little “flakey.”

Ed,
At nine miles (roughly 40,000 feet) there is NO way you’re on copper from the main C.O. There has to be some kind of MUXed T3 carrier or a DLC at the remote (or “hut” as it’s called in your area) where your T1 is placed on copper from.

To achieve THAT kind of distance on copper (which “can” be done) you’d have to still be on old T1 repeater technology… about the only thing that WAS GOOD about those was the ability to use one repeater after another, after another on a copper span. … but if you fatred on those things you’d knock ‘em down and they are NOT auto-realigning.

BTW, you also mention load coils in your post… Load coils are digital signal KILLERS. They HAD to be cut out at the cans before you ever got a digital circuit.

Regarding the degradation of service caused by numerous DSL loops within the same cable: Trace, whomever told you this in the past, was correct. :thumb:

Particularly when more than one pair within a cable's binder group carries high speed data, interference between these data-pairs may increase the error rate for these high speed circuits. Normally, these error rates are maintained at acceptable levels by reducing the transmision level of various services such as ADSL. Unfortuantely, in my experience, this is not possible on HDSL circuits.

When providing HDSL across copper, like Bryan also suggested, try to ensure that the HDSLs are assigned to pairs within separate 25-pair binder groups.

Oh, boy. Its been a while, but out comes the soap box:

I don't know guys..... A binder group is still within the same cable jacket. I don't see how a simple plastic tape strip to bind 25 pairs for identification purposes is going to have any electrical effect on the pairs themselves. Twisted pair cable is designed with varying twist lays so that there's little likelihood of crosstalk or other interference between the pairs. The two circuits could still be right next to each other as the binders are side-by-side within the cable sheath. Everything spirals in a twisted-pair cable of any size, so it's inevitable that the pairs, no matter which ones, are going to come in contact with each other at some point.

Think about this: The cable feeding my building is using pairs 51-75 and my circuit is on pair 51. My neighbor orders a circuit too, sharing the same cable count. The telco isn't going to do an LST so that we are on different binders. His circuit will simply end up on pair 63; the next available pair in the count. I really, really doubt that any thought process goes into assigning circuits to the proper binders.

On the other hand, what if there's not a choice? Plenty of times, a 25 pair cable runs down the road feeding multiple customers. If that theory was accurate, then only one customer could have a Hi-Cap circuit for that entire distribution segment.

What about when a 3,600 pair cable leaves the CO and branches out two miles down the road? In a manhole, this cable splits in two different directions:

To the left is branch cable A, an 1800 pair cable is spliced in as pairs 1-1200 (the remaining 600 pairs in this branch cable are just spare for future use).

To the right is branch cable B, a 1200 pair cable is spliced in as pairs 1201-1500 because only 300 pairs are needed for now. In that cable, the remaining 900 pairs are also left as spares for future use.

Later on, the 1,200 pair count is full in cable A and they need to get 200 more pairs. Bill schleps down into the manhole and cuts the next available pairs through. The first branch cable now becomes 1-1200 and 1501-1700.

How many times have you seen something like this written on a building entrance terminal?

CA 236, 401-600, BP 1-200
CA 117, 201-225, BP 201-225
CA 209, 1801-1900, BP 226-325
CA 9, 2201-2275, BP 326-400

That's right, hook or crook, multiple feeder cables are merged together at some point into the single cable feeding the building. All of these different cable counts end up neatly terminated as binding posts (BP) or pairs 1-400. There is absolutely NO rhyme or reason as to how the binders are assigned with any telco. It's all about throwing the pairs where they are needed and usually a case of borrowing from Peter to pay Paul.

Next time you are driving down the road, look up at the telephone cable splices. You'll frequently see how a cable jumps over to tie two cables together. That's because of the pairs in one cable were running low and the other one had spares, so the original cable got a "refill".

Moral of the story: Binders really don't mean a hill of beans in the cable world. All of those pairs resided together within the same cable jacket for miles before they went in different directions. They touched many, many times over the two-mile route.

I'm all in favor of keeping circuits separated as much as possible, but I truly believe that the "separate binder" thing is more of a psychological issue. We tend to associate pairs 1 and 25 as being 8" apart since that's how they end up on a 66 block. They appear to be miles apart, but they really aren't. They are still tightly bound together for hundreds of feet or even miles before they end up terminated.

Is it possible for Hi-Cap circuits or even DSL to interfere with each other in the same cable? Probably... Is this binder-specific? That is impossible. It is just a coincidence that something like this happened.

Boy Ed you really cause a guy to think. I remember the T-carrier, where transmit and receive were separated by a metallic barrier. I think the separation of groups is for a reflective type interference. As you know as well as pair twist, the groups also twist which reduces reflective interference. So there you have my theory for what it's worth. (I about wore out spell check on that one)

Great Discussion! I'm sure glad you found that box again, Ed. And yes, you've got me thinking too!

I wasn't trying to imply that interference between DSL circuits assigned to copper pairs cannot occur across binders within each 100 pair count.

However, in most instances I've experienced, interference (even crosstalk) between 2 circuits assigned to copper pairs occur between pairs within the same binder. Thus, if tested vacant pairs within an adjacent 25 pair count terminating at the same location are available, I'd assign a second HDSL to pairs within the adjacent count, versus within the same count.

With a regular 4-wire T1 being sent down copper? I'd separate the xmit and receive pairs by one or 2 vacant pairs (I know..keywords= "if available")

Bill, I know what you mean. T-Screen cable is rarely seen these days, yet it made sense for analog T1 circuits. These days, T1's are just emulated using digital circuits. I agree that on analog T1's, all of the transmit pairs should be in one compartment of the cable and the receive in the other for long-haul 4-wire circuits.

But when we have the technology to place both services over one pair, what happens? Is it a transmit or receive pair? Therein lies the question. The proverbial chicken or the egg thing for sure.

Digital is magic. That's why you see the south end of the north bound horse us old phone guys are riding.

Ed -

When I was with GTE we always ran T-Screened cable. General Cable had a type that had a foil shield on each pair and an overall braid. We never had a problem. Transmits and receives under the same sheath, but it was enormous - an eight pair cable was the size of a standard 50 pr!

When I did carrier work for my current employer, we ran either 2 @ 6p 22AWG shielded for small installs, or 2 @ 25 pr 22AWg for larger installs, or 2 @ 30/32 pr 22 AWG ABAM cables.

All the cable was shielded and we ran the transmits in one cable and the receives in the other. If you tried to run them in the same cable it would fail the stress tests every time.

The only exception was short (<50') runs of Cat 5 that some customers spec'd and we tried to keep away from that.

I always felt that the Transmit & Receive pairs of a carrier circuit were like Boys and Girls at a High School dance. Always trying to get together and make trouble. You had to keep them apart and when they did come together, get that shield on!

Sam

Ed,
As always I appreciate your post and even the soap-box rant… There’s more to it than a “psychological issue” with binder group separation… Though you correct that the overall proximity of say cable: A pr:23 and cable: A pr:32 are very close together when looking at a cross-sectioned cable with your eye or even a physical measuring device for the proximity of pairs. The biggest concerns that binder separation accomplishes is really eliminating capacitance between pairs.

Now I’ll admit the theory get’s to be a bit more complex than I ever cared to ponder on, but the rather miniscule amount of separation is pretty large when applied it how a capacitor works with distance between insulator plates (for cable pair that’s the paper or plastic insulators.) I have a feeling this is WAY too simplified of an explanation but it’s how I try to understand it myself … The less amount of capacitance between pairs there is, the less the pair will know the other even exists. (That could be so over simplified to the point of not even being accurate but that’s how I get by.)

Bryan:

I didn't think it could make it that far either. I'm sure that the Verizon technician who told me this didn't know what he was talking about.

As for the load coils, about three miles of the 2,700 and 2,400 pair cables are brand-new (less than two years old). So are the load coils. I'm sure that only certain binders are actually loaded for the POTS services. :rolleyes:

Our ADC HDSL shelves provide separate amphenol connectors at the rear of each chasis; 1 for the transmit, and 1 for the receive. The two 25 pair cables are terminated to a horizontal block at the MDF, by combining both cables to form each HDSL pair on the block.

Wow... great information guys. I came here today to see if I could find answers to the same questions. Our CLEC just moved a T1 for one of our customers and this time they installed a single pair T1. I was worried because we've had other issues with this CLEC and I figured this would just be something else to figure out.

I understand the concept of separating the transmit and receive for a T1, but sometimes I see it come in on separate binders, but go to the customer's RJ21X on 2 pairs together, such as 24 and 25. Is this a bad thing?

Jeff -

It's not the best practice. The big thing is to keep the transmits and recieves apart. The longer they're running side by side, the more likely that there will be problems. If the circuit comes in on different binders and just goes to the demarc with a x-connect then there probably won't be any trouble - as long as the customer separates it again from that point on.

But it's still not the best practice.

Sam

How bad is it for the T1 to be in a single cable from the smartjack to the equipment, about 10 feet long?

Rob, it's not bad at all. In general, if it's under 166 feet, the circuit is relatively tolerant. Now if that cable is wrapped around a fluorescent light fixture, then you might have issues.

Rob -

10' should be no problem at all. Not to disagree with Ed (Which is ALWAYS a losing proposition) I would keep a single cable at 50' or less and stick to Cat 5 or better for it. That's just based on my experience. Your mileage may vary.

Sam

Here is a quick writeup on HDSL:

https://www.arcelect.com/HDSL3698.htm

I know that this topic was originally started late last year but as a Telecom tech, I would like to advise the readers of my own recent experience with 2 wire vs 4 wire T1 circuits for voice traffic.

One week ago, I had to de & re a phone system for a customer that relocated to another city. At their old location they had 2 T1 circuits that were 4 wire and we never had any problems with them for almost 5 yrs that they were in that location. At their new location, the T1 circuits were 2 wire circuits.

About 3 days before the move, I brought out a test system to the new location for the turn up and test. No, I don't have a T-bird for testing. At first, I kept getting "Out of Frame" errors on my system on both T1s, which the CLEC was able to stabilize after making adjustments somewhere and then the turn up and test went smoothly. On the day of the move, after re-installing the system, sporadically, I would get Yellow Alarms, which sometimes would last anywhere from 10 seconds to about 2 minutes.

Opening a trouble ticket with the CLEC, they looped and tested to both the smart jack and to the system's CSU for about 30 minutes each on both T1s. After the testing and unlooping, again we started getting hits on both T1s. The local Telco, which is AT&T, came out and moved both T1s to new pairs and they then appeared to be working properly. I asked the AT&T tech how far it was from the CO and he said it was less than 4000 ft.

The next business day, which was the customer's first work day, the sh*t hit the fan, both T1s kept going down all day. Again they tested them and they ran clean but the customer still had problems with the circuits. The tester for the CLEC tried to claim it must have been a problem with the customer's equipment but I explained to them that this equipment had been in and working for over 4 years with out any problems. By the way, this customer has had this same type of equipment at other locations without any problems either.

I wasn't involved in any other conversations over this problem but finally, just yesterday, the CLEC sent out a tech with a T-Bird for another vendor meet. He tested the first T1 for 20 minutes with only 2 errors, tested the second T1, which he said was clean. Then he tested the first one again and almost immediately he got over a million errors and at first he wanted to put the blame on the cabling for the extended T1s from the Smart Jacks at the NIUs, which are on the first floor to the customer's suite on the fourth floor. He went downstairs and tested it directly at the NIU and got similar results.

The CLEC opened a new trouble ticket with the local telco, why it was only put in for the first T1, I don't know but last night they reconfigured it as a 4 wire circuit and it has been up all day with out any problems, the second T1 though, which is still a 2 wire circuit, went down 6 times today. They called the CLEC to have AT&T to convert the 2nd T1 to a 4 wire circuit too.

I'm convinced that 2 wire T1s just aren't as stable as 4 wire circuits. At least not T1s for voice channels. Maybe data only circuits have the same problem but it isn't apparent because they only think their network is slow when it's down. Of course with VOIP being used more and more, they will realise the problem if they any VOIP connections over those T1s, like remote extensions, networked systems or use SIP trunks.

Now they are going to have to figure out who is going to be responsible for all my extra time I've spent on this, which will cost them somewhere between $800 and $1000 more than the original estimate for the de & re!

Chuck,
There was something else going on there... From what you've written I can't figure out WHAT is was… But the HDSL2 technology has proven itself VERY reliable for several years and it is actually more resilient than 4-wire due to only needing two conductors.

There really is no difference either on voice or data… 1.544 Mbps is what a T1 one is. What’s IN the train of bits does not matter to the Telco equipment that’s used in HDSL2 it keeps the training a rolling at a fixed (timed) 1.544 Mbps.

Hi all,

I'm new to this forum, just wanted to say what a great place with knowledgeable people.

I'm a relative newbie (3 yrs), started with FIOS splicing and have been doing copper most of my 3 yrs. I work for Verizon in NE Indiana and nowadays we do EVERYTHING from the CO frame to fixing cable to CPE to computer work (DSL contracts). A handful of us install and repair special ckts (T1, analog, etc).

Since we have a large area, some of the old independents Verizon (GTE) acquired used what we call traditional 4w T1s where you had your pr separation with shielded cable and long hauls (most of them are for CO or RSU (huts) to Muxes).

In the city its fiber (lite-span) muxes. The T1s are 4w HDSL 1st gen, HDSL4, or HDSL2. The prs assigned are whats left in the count, or what we can use that's good - without any consideration for what binder generally. HDSL4 is used when HDSL2 reaches the maximum attenuation spec (length, gauge, etc).

HDSL4 is definitely more robust than 2w, but I think the main problem with all of these flaky ckts is:
1) Are the prs REALLY clean? (faults, dirty splices, bridge tap, etc.)
2) Is the ckt falling well within specs (dB, S/N, atten)
3) Is the HTU-C (CO card) any good? Sometimes when ckt goes down when all else has been done and the remote testers say its good to the office the card is erring on the cable pr side of itself.
4) Have the protector modules (5 pin) been changed, and are they the right type rated for hi-cap ckts? If the demarc screw in tube type and they're carbon, they can break down and throw errs.

The ADTRAN cards that have the T-SCAN diagnostic built in can clue you in on what is going on alot of times. We plug our laptop in to the craft interface (RS-232) and look at the Margin (of S/N ratio) fluctuation, Atten (dB), history of err/sec, severely err/sec, Unavailable (down) sec.

The nice thing about working on HDSL4 with repeaters is loopable repeaters to isolate what section pr faults are in. The traditional T1 you gotta goto each can, pull the repeater, next can, next can, put em all back in....Plus with HDSL I can loop myself w my laptop (which should of been done before ticket sent outside).

Sorry for the long story, and believe me I respect you guys and your decades of experience, I just wanted to give you my perspective of how I work ckts.

And THANKS again for all the info, I've sure learned alot in the posts you "Old-timers" have done!

Welcome to the board!

You've given us a lot of good information - Thanks! Stop back regularly, ask for help if you need it, give help if you can and have a good time always.

Sam

HDSL2 (ADC/ADTRAN) is held to the same standards as a repeatered T1 facility by ILEC's. HDSL2 modulates a 4-wire DS1 payload onto a boradband/analog carrier between 2 digital transceiver units(HDSL Line Unit/HDSL Remote Unit)over 1 2-wire cable pair. HDSL2 does not work as well as a repeatered T1 facility on marginal OSP facilities due to it using only a single conductor for its simplex power loop between the HLU & HRU. Light ground-faults and high-resistance opens and shorts that may not typically cause troule or go unnoticed on a T-1 facility can play havoc on HDSL2/4. As long as it is carried over healthy OSP facilities between the telco and customer, HDSL is just as reliable as any other DS1 transport technology.

And on a side note. Don't bother wasting your time plugging a T-BERD or any other T1 test set up to an HDSL span with the intent on it taking errors by sending a certain stress pattern. Those stress patterns were developed to troubleshoot a purley digital facility with its associated loop electronics and wiring(ie* T-1). Sending differnet patterns of 0's and 1's isn't going to make a difference to the analog waveform of the HDSL span. The only reason to connect a BERT device to a DS1 circuit being transported on HDSL 2/4 is to troubleshoot the wiring and electronics on either side of the HDSL span and line-coding and framing options.