Well, it took almost 4 months, but now Paetec is insisting the cable run is too long from the smart jack to the CSUs. The run is almost 1/4 mile in an old Westinghouse manufacturing facility turned into a business park. My client, is, of course, on the direct opposite end from the Verizon Demarc.
I'll spare you-all the details, but we ran single mode fiber from the smart jacks to the CSUs. I bought the fiber from FIS in NY and, until this job, they have been right on recommending their products for their fiber. I have always used Transition Networks media converters when going from copper to fiber and then back again. They work and have never failed me. However, the going price for a SSDTF1014-120 is nearly $825 per unit and I need four of them. When ordering the fiber and the associated mounting hardware, etc. I asked the sales rep if they had T1 media converters? He said they do, but I insisted he make sure we were talking about T1-E1 to fiber and then back again. He consulted with his technical people and they assured me a SigmaMax unit would do the job. I believed him. Mistake #1. He then said the peice was about $250 per unit. I believed him. Mistake #2. Since he assured me this would work, I never bothered to call Signamax to make sure the media converters would do the job. Mistake #3. I installed the Signamax units and hooked up the fiber patch cables, but it was getting really late and I did not TEST the circuits last night. BIG MISTAKE #4.
Today was the big cut. All the dignitaries (and I use the term loosely) were there. The scheduled time for the cut was 4 am. I insisted I needed to test the circuits before committing to the cut. Well, you can guess the rest. The cut did not happen because the Signamax media converters, although they work well for ethernet to fiber, don't work, at all, for T1-E1 to fiber. After a few long calls to FIS, Signamax and Transition Networks, it was clear that I have to either find some used Transition Networks media converters or eat the $3,000 (My cost) and that's after the refund from FIS for the Signamax units to make this all come out with a happy client and two working T1 circuits.
The bottom line here is does anyone have 4 Transition Networks or, for that matter, any other brand of T1-E1 to Fiber media converters that they would like to sell? Cash and quick is how I want to do it. The Transition networks part number is SSDTF1014-120, but that is the new unit number. The old style works just fine. It's the same number with a 105 on the end. The detail is RJ-45 copper input, SC-SC Fiber Single Mode. Thank you.
Rcaman
That's one long extension...
Are those T-1s copper or fiber, 2-wire or 4-wire circuits?
The new Paetec circuits are 4 wire copper. The old Choice One circuits (which worked very well for 10 years) came in as HDSL and were 4 wire copper as well.
Rcaman
Is the d-mark area secure enough to place the routers in there and use the equipment you have now to extended the Ethernet signal from the router to the equipment in the suite?
The other question is where did Vz place the r-cards for the old Choice One circuits? If they are in the d-mark area also and then extended to the suite from there, could you possibly cut one circuit at a time and reuse the extensions from the old circuits?
Well, now you are asking for the "boring details." Yes, the Demarc room is locked and secure, however, I am trying to extend two T-1 voice only circuits and not data circuits. At the other end of the line sits a CSU not a router.
Here's the argument Paetec is making: Prior to some time in 2010, Verizon supplied smart jacks that could push the signal, almost, to the length that we are dealing with here. That is why Choice One's circuits worked and Paetec's don't. When Verizon installed the new circuits, they used the smart jacks that don't provide much in the way of conditioning to make the long hop. That's becuase Verizon doesn't support the older smart jacks.
Now, I'm an ex Bell Telephone employee and I KNOW we were expected to make the equipment work no matter how old it was or even if it was MD. So, I'm not really buying this argument. I think the circuit was not engineered properly. The problems are: Delay in voice cut through on incoming calls, cut-offs in mid call, dialing numbers and getting dead air and low receive and transmission volume. Now, the amazing thing is that these problems are not present all the time. They come and go. The circuits drop about three times a week for no apparent reason. The Mitel EL always reports loss of sync and a Yellow Alarm and if it's out for more than a few minutes, a Red Alarm. Usually, after a few miserable calls to Paetec, the circuits come back up. Occasionally, I have to restart the Mitel.
Rcaman
Paetec might be on to something, though...any clue on what distance from the CO you're looking at?
I'd ask for a vendor meet with Verizon, and when the tech shows up have him open the pairs and test all of them with a JDSU. If all copper tests pass you move on to the next phase, which is testing the actual PRI setup.
What type of R cards are used on the old circuits, as opposed to the new ones?
How about having Peatec intall a doubler's in the 1.1 and the NIU's in the other end of the building? They can run T1's for miles outside and can't engineer another 1000' feet to the DLR because it's inside?
Ajkula66,
The smart jacks are Adtran units and I don't have the model numbers. I think they were HDLU or something like that. The CO is within a mile of the facility and these particular circuits come out of a SLiC 96 on the property.
SST,
Huh...really? You just wrote what I have told Paetec, at least, 500 times. How hard is it to design an extra 1,500' into the span? Evidently, much harder than one would be lead to believe...for Paetec, that is. We used to regularly install T-1 circuits into coal mines. Verizon would bring it to the mine property edge and it was up to us to make it the 1-1/2 miles to the bath house where the telephone equipment was installed. Funny....I NEVER had this kind of trouble with the T-1s we installed. (Over 40). Back then (20 years ago) it was just installing repeaters in the span and, as long as the span was dry, you could go on forever. I mentioned this to Paetec's engineers (I use the term very loosely) and you would have thought I was speaking Klingon.
Rcaman
If the old Adtran cards are the dark gray ones, with a long 1xxxxxx something or the other number, the matching C card can be tweaked quite a bit as far as I recall. Someone would have to do it in the CO. That might have been done when the original install took place a decade ago.
Presuming the new cards are H4TUR (light gray ones), no tweaking there on the CO side. They behave very differently in real life when compared with the old ones.
I would still get Verizon out first and prove the circuits good (or not so good) with a JDSU.
The "Klingon" suggestion is a very sane one, but I guess not for Paetec.
That being said, any time I've seen a T-1 sitting on a repeater in the field, the design involved the *old* cards even on brand new circuits...but I've never worked in PA so I have no clue what Verizon's standard practices are in your area.
Good luck, my sixth sense tells me you'll need it...
Rcaman,
With a true T-1 you could add another repeater and have it work, but with HDSL it doesn't work that way. HDSL has to have a C card in the c.o. or local vault, then it may pass through up to two doublers (no more repeaters), and finally terminate at the location with an R card that converts the signal back to t-1.
Even with the LBO set to it's highest level I haven't been able to extend a circuit much past 700' without it taking errors on the extension, so I am not sure what is up with the two old circuits. If we have an extension that is taking errors due to distance we convert the circuit back to HDSL with local C and R cards. We have to bill the carrier for the equipment since it is after the d-mark, but most carriers accept the charges.
If you still need a solution, and you have spare copper going from the d-mark to the suite, and you are sure the extension is the cause, I suggest getting some prices on the equipment needed for converting the signal back to HDSL. If it's cheaper, pull the fiber to use at another job, cut your losses and run.
Here is what you need from Adtran:
T200 H2tu-c x 2
T200 H2tu-r span power x 2
2 position t200 mounting x2
48vdc power supply to power the c-card mounting
Ajkula66,
You are quite correct. The old circuits have the dark gray Adtran cards and the new ones, the light gray. All I know is that when Bell was using Telco and Wesco cards, we could extend T-1s, easily, to 1,500 feet past the demarc without taking errors.
Whynot,
The circuits that worked for over 10 years came to the demarc as HDSL. The new circuits are pure T-1. When all this first started, I mentioned to Cavalier (that's how this whole thing started) they might need to re-engineer the circuits to HDSL and let Verizon condition the circuits to make the hop. Well, that was not in the plan, at all, and since then, we have been trying to make gold out of lead. Now that Cavalier is Paetec, it is much worse. It's almost like they have one person that kind of knows what is going on, but that person is too busy to DO anything about it.
Placing the fiber was an arduous job and, after taking with Transition Networks engineers, I think their T1-E1 MUX is the way to go. I ordered two of them yesterday and expect them to be on site by Wednesday. I have accepted that I will have to eat the cost of the MUXs, but, now it's more a principle thing than anything else.
Here is a bit of knowledge that I gleaned from this train-wreck: The devices to extend, condition and convert T1-E1 media have all but gone away. There are a very few companies still making the devices and those that do can charge anything they want. All the other companies are churning out media converters for ethernet only. The way of the T1-E1 media converter is slowly slipping away. Heck, I worked on N screen and T1 circuits when we had vacuum tube equipment to condition the circuits. Back then, Bell got the circuit to the telephone equipment and it worked. Alas, those days are gone. The Verizon techs today have their hands tied and can only replace cards and do little else. Recently, I had a Verizon tech out to test another T1 circuit and he couldn't get anything to work. His T-berd was working, but he just did not understand why he couldn't see any patterns. After a half hour of watching him do everything he could think of, I finally asked him if he knew he was trying to test the wrong circuit. He looked at me like I had snakes crawling out of my ears. He never thought to look at the smart jack to see if there were any lit indicators.
Rcaman
Rcaman,
If the new circuits are running on light gray cards (H4TUR) they are likely provisioned as HDSL as well. That being said, I'm not keen on any newer Adtran cards, be it 2-wire or 4-wire.
I miss the "old" cards and setups...
People get very limited training these days, and have no desire to go past the norm in most cases, no curiosity to find out what the alternative solution might be...I count myself lucky for having some really great older techs show me stuff in the field, otherwise I'd be deaf, dumb and blind like most of my colleagues choose to be...
Good luck with this project and keep us posted on its progress.
Rcaman, Huh....yea really, your post was concentrating on extending the d-marc and how many mistakes you made during the cut-over not how Peatec didn't listen you telling them how to re-engineer a T1 span.
Ajkula66,
Will do. I should have this under control by Wednesday.
SST,
I left a lot of background out. Several months ago I had a thread open that gives the background and I thought repeating it would be too boring. My initial hope was that some of us here might have some media converters that they would like to sell quickly and I could resolve the problem in a matter of hours instead of days. That not being the case, I will report back the progress after I install the MUXs. To be honest, I don't really think it should be the client's responsibility to get this circuit working. I'm doing it because Paetec has literally given up. That's just not the way telephone people react to a problem. Call me old, but there is still something called ethics and responsibility that is sorely missing with a great many telecommunications companies today.
My comment about Paetec and re-engineering the span comes from experience as this is what I did for Bell when I first started back in 1968.
Rcaman
I'd like to help. Could you restate the issue in simple terms. That is all my wife says I can understand.
Thanks
Is it safe to assume that in your case the Telco demarc is about 1500 feet from the customers equipment? If so, the limit for the length of the extended demarc is about 955, give or take a few feet beyond the smart jack. The power level at the end of the extended demarc cannot be lower then -22dbm to have a chance of working. In HDSL it usually isnt a problem since the power level at the output of the smartjack is around 0dbm and won't lose that much to the extended demarc.
I would not eat anything. Have they asked you to? Tell them you are on a diet.
When we've faced these same circumstances, we use Patton 2115 T1 extender - uses one copper pair and extends 3500 feet. Of course, that assumes that you have a copper pair available...
Mike
For those of you who are awaiting the results, we have been waiting for Transition Networks to ship the units to the local Graybar. It seems that:
a. these are the most popular pieces of equipment Transition makes - or
b. they make them as needed as there was no stock in the US or at the factory.
Yes, the problem is I have 2 T-1s 1,500+ feet from the smart jacks to the customer's Mitel PBX. The carrier is doing nothing to make the circuits work so we have to make a solution to the problem.
I'll give everyone an update as soon as I have them.
Rcaman
A thought. Are you using -021 T1 cards? If so, check the distance settings on them. They 'may' work if you set them for MAX distance, even though the CSU's are close. Also, how about moving the CSU's closer to the d-marc? They are, after all amplifiers as well as signal conditioners.
AND, it is possible that you have MORE than enough level @ the CSU output and are overloading the input to the 021 card. So, try setting the distance switches for CLOSEST.
Of course, if they are not -021's ..........
Lightninghorse,
Thanks for the reply. Yes, they are 021 modules and I have switched every setting so many times I fear the switches may go bad. We are using Adtran ACE CSUs and they, also, have switches. Same deal. The very first thing we tried was to move the CSUs to a mid point, more or less, to see if anything would happen. That did improve the situation, somewhat, and that is what lead me to install the fiber and put the whole distance thing to bed. Moving the CSUs also caused some other minor problems (slip and ber errors started to climb and, every now and then either T1 would lose sync) and moving them back resolved those problems.
When the circuits were first installed over ten years ago, we made sure the xmit and recv T&R were in separate binders and even went so far as to install another 25 pair PE-39 alpeth to further isolate the T1s. That worked well, for over ten years through two carrier transplants. However, Paetec is insisting the hop is too long and that is another reason we installed the fiber.
Rcaman
Here, as promised, is the result of this ardious journey.
The MUX units were installed on Tuesday. Right out of the box and connected, they worked perfectly. All I had to do was adjust the CSUs back to 133' (minimum) and cancel out the yellow alarms. It was the slickest and easiest equipment installation I have ever done in 45 years of telephony.
I was working at the PBX end and another one of our techs was at the smart jack end. We had a phone connection going so we could report back to each other the condition of the smart jack, the MUX and the PBX. While connecting the MUX at the smart jack end, my tech said "The smart jack LEDs indicate AMI SF. The MUX fiber link LED is green and the T1-E1 ports are green. Good to go." With the exception of the smart jack, I had exactly the same indications. I told him I was going to place a call to him on his cell phone. I did and he answered and said "did you do something in the switch? The smart jack LEDs indicate B8ZS." I said, no, but when I checked the channel status, it said B8ZS. I checked the T1 forms and the T1 condition was programmed AMI.
So, I waited two days and called the customer and asked how things were? She said "The delay in answering is gone. The volume levels are great, both ways. But, we are still getting random cut offs and wrong numbers dialed." I said uh-huh.
So, she asks me to talk to the person at Paetec that's supposed to know all the answers. As usual, she was arrogant and insisted we needed to change out the CSUs because that's what her testing has determined was the problem. Now, she maintains, distance was NOT an issue! I mentioned the smart jack indications and she said, that doesn't mean anything!
So, just to prove a point, I installed two brand new CSUs AND replaced both T1 modules in the PBX. SAME PROBLEM! Huh....really? The customer knew I replaced the equipment but Paetec didn't. When I called Paetec and asked about the condition of the circuits, she blythely stated: "Until you change out the CSUs, we aren't going to do anything more. All our tests are conclusive that it's your CSUs that are at fault." I asked her when she did her latest test? She said "This morning." (Friday). It took all I had not to scream LIAR in the phone. All I said was "I doubt your testing showed anything but a lot of errors and you really don't know where the errors are coming from." Then I said "Please tell your circuit engineer to check over his/her work and make sure it wasn't engineered as B8ZS instead of AMI. If it was engineered as AMI, then you need to have a chat with Verizon."
So, my fine feathered friends, let my misery be for your profit. NEVER trust the carrier and don't ever let them know you know more than they do. I know, you are all thinking, did the LEDs on the smart card show AMI or B8ZS BEFORE the fiber was installed? Yes, they did indicate B8ZS on the smart jack and in the link status on the Mitel. But, Paetec's "brain trust" said that was because the line length was too long and that's a "perfect indication" of line loss. Really? Huh!
By the way, the Transition MUXs are fantastic and work beautifully. They come as 4 T1-E1 x 1 serial port or 4 T1-E1 x 1 ethernet port all over one pair of either single mode or multi-mode fiber with your choice of ST, SC and LC connections. The data rate between MUXs is 155 Mb/s, plenty wide enough for 4 Ts and a 100 Mb/s ethernet.
Rcaman