RAD problem 2

[comdata]

Does anyone know how the Xmit and Rcv pairs pinout on the RJ45 connector of a DS1 board? I,m trying to migrate a copper PRI between a Legend R7 and Magix R4 to fiber, using RAD fiber modems.

I tried to connect the RAD's to their respective DS1 board with a straight-thru Cat 5e cable. The leds on the RAD's seemed to indicate that everyting was a go, but the DS1 boards both displayed a red led. Called RAD tech support and they seemed to think that I needed to cross over the Xmit/Rcv pairs at both ends. Tried that, and the calls still didn't go thru.

Is there something I am overlooking? The only options on the RAD's I used was T1 and AIS.

 

[dagwoodsystems]

1 R Ring In RX+
2 T Tip In RX-

4 R1 Ring Out TX-
5 T1 Tip Out TX+


I've listed the pins and three different designations for those pins. Now go get RAD's information.

Our pin 1 (Ring In) should connect to RAD's Ring Out, our Tip In to RAD's Tip Out...and so on.

Once this piece is fixed, you'll need to look at the configs. Both RAD and Magix have timing parameters. Magix also configures differently for T1 than PRI.

If you're networking two PBXs together, don't assume you're wiring is incorrect just because calls don't get passed across. Networking switches requires a non-trivial amount of programming.

 

[comdata]

Thanks for your reply. I did get the pinout from Merlinman, but in trying to get back to the RAD engineer before closing, I neglected to write them down and gave him the wrong info. My apologies to Merlinman.
I know that the Magix and Legend have timing parameters. The unit I am working is FOM E1/T1 and has 5 dip switch settings; none of which address timing parameters. Do I have the wrong unit?
Just to clarify; this point to point PRI already exists and works. When I plug in the FOM units, the DS1 board goes into alarm and the error report indicates D-channel inoperative. As soon as I plug the copper facility back thru, the DS1 led goes out after a couple of minutes and I am able to call into the Messaging housed on the Legend side and make station to station calls as well.
Do I need to change any parameters on the Legend and Magix to utilize the fiber optic modems? Thanks again for you input.

 

[dagwoodsystems]

It looks like you have the right units. And now I understand your project: you are replacing an existing circuit in which two PBXs have been successfully networked.

Unfortunately, RAD doesn't publish their manuals online. Otherwise, I'd know what the heck those dip switches did. You may just have to work with RAD support unless we know more.

I've used a variety of RAD muxes and--in my experience--have needed to use T1 crossover cables between the RAD interface and the PBX (both ends of the circuit). If RAD has provided them for you, the wire jacket will be red.

Since this particular model apparently doesn't have any timing parameters, I wonder if it should be treated more like a wire than a circuit. And knowing that you can network these PBXs together with nothing more than a T1 crossover cable, consider trying a straight through cable on one end and a crossover cable on the other.

Remember too that when I speak of a crossover, I don't mean an ethernet crossover. Please don't be insulted by me pointing that out...it's just that sooooo many people do that. You need to make sure that what you think is a T1 crossover (a.k.a. rollover) cable, really is a T1 crossover cable.

An ethernet rollover cable takes 1 & 2 and crosses them with 3 & 6, and it won't work with T1 circuits.

A T1 rollover cable takes 1 & 2 and crosses them with 4 & 5. The polarity on the pairs is far less important than the position. So if you build a crossover cable with 1 to 5 and 2 to 4 (instead of 1 to 4 and 2 to 5), it's not the end of the world. It will work.

Write back and let us know what happened.

 

[comdata]

Thank you for the additional info; I have a copy of the manual on a CD that came with the unit. As a matter of fact, I have an extra CD with info on other RAD products I could mail you if you email me your shipping address.
I will be going back to the job in the next couple f days, and will post results. Email ics AT mc DOT net.

 

[kwing112000]

I have used these in the past. Do the RAD units sinc up on just the fiber. You need to flip the tansmits and recieve on the fiber itself. Trans on one is rec on the other. THen it is just plug and play. I have done it with a Merlin and a Definity. I pulled out the PSTN PTP and plugged in the RAD units and it all came right up.

 

[comdata]

Thanks for your input. I have the fiber setup as you stated, and the leds on the FOM appear to confirm that the modems see each other. Did you use a straight thru 4-pair cable from the DS1 board to the RJ45 jack on the modem at both ends of the circuit?

 

[kwing112000]

I cant remember if it was straight or not. If you need to flip it just flip one end and make pin 1 go to pin 4 and 2 to 5. THat will give you the T1 cross. The other thing you can try is to plug one PBX in and then go to the other end and put a loop-back in the rad and see if the PBX becomes green then try the other way. A loop-back will be 1 to 4 and 2 to 5. this should tell you if the connection is working.

 

[comdata]

Thanks for the input; I will give that a shot. I'm taking a Phoenix tester out to the customer premise next week to satisfy the RAD engineer that the fiber between the two modems is working properly.

 

[comdata]

We setup a Phoenix tester and successfully ran several different patterns without any errors across the point to point PRI. We were satisfied that the connections were good from the input of site A to the output of site B. We also installed a loopback plug on one end and those tests also ran without any errors. However, when we plugged the Cat 5e cables into the Legend and Magix, the red led extinquished on the Legend but wouldn't clear on the Magix system. We continued to monitor the connection and saw good data from both systems, but the two systems would not sync up. Error log showed "D-channel inoperative" and blue alarm. We tried installing a crossover and one end, both ends and straight-thru, with no success.
Close but no cigar. What am I missing? Is this a timing issue? If so, where do I need to tweak it?
Thank you.

 

[merlinman]

Alex, Call Me tomorrow, 1-31-07.

 

[comdata]

After a lot of turmoil we have determined that one of the fiber modems was defective on the RJ receive side; the modem was able to transmit to the Magix but for some reason, the RAD modem was not able to receive info from the Magix system. The Legend side of the circuit is functioning properly. RAD is sending out a replacement modem. I would like to thank all of you for your input and suggestions.

 

[comdata]

I just installed a new pair of fiber modems from "Transition Networks". The leds all indicated that there was connectivity between the 2 modems, as well as connectivity from each respective modem to the Legend at one end and the Magix at the other. The red leds on both DS1 boards also went away and displayed a yellow led whenever a call was attempted. I must still be missing some programming parameter. When I place a call to the opposite switch, the other tech tells me that the phone just rings one time, after which I hear a fast busy on my end. The TEI is seete at "0" at both ends. Do I need to do something with the timers?

 

[comdata]

I forgot to mention that the errorlog states, "pri svc audit timeout" error 7001 for each attempt on a channel and "pri b-chan not released" error code 7004. When I switch it back to copper, it takes at least a 15 mins to a half-hour before we can use the point-to-point again. Thank you for any suggestions or ideas.

 

[SYQUEST]

Did you review the troubleshooting section of the install guide? It shows different fault conditions and tests.

It seems the Legend/Magix is having a D CHANNEL problem. Without the DCH the switches can't communicate. All call set up and teardown happen on the DCH. Check the error log when you switch back to copper and see what alarms you are getting. It should not take 15 minutes for the switches to est the ckt back. Also try looping each DS1 module and run tests to make sure they are good. Use a physical loop also. You may have an intermittent DS1 module.

When you say "programming parameter" are you talking about the Legends/Magixes or something else?

Because your parameters for the DS1 or the DS1 module should not change for either the copper or the fiber. The Transition Networks media converter is transparent to framing and line coding.

....JIM....

 

[comdata]

Thank you for your reply. Yes, I did read the troubleshooting section of the manual. I am in the process of getting a BERT tester; however the fiber modems don't appear to have a problem. The reason we are going the fiber route is because the T1 is very unstable over copper and goes down frequently everyday. Just to re-interate; when we plug the fiber modems thru, the red led on both systems goes out rather quickly, unlike when we plug the copper back thru.
The parameter I was wondering about was the line comp setting on Legend/Magix; which set at 1, is about -0.6dB That seems a little too hot for transmitting over fiber. When I was testing with the Data Tech, he seemded to think the line comp for the fiber should be around -6.0dB. I have installed and programmed a few PRI's and T1's in my former life, but haven't had much experience in troubleshooting them. When we get back to the site, we will recheck the errorlogs before and after switching between copper and fiber as you have suggested.
I also agree with your comment on the DCH problem.
Thanks again.

 

[merlinman]

Alex - I too am curious about the DB Level of the circuit on the FIBER.

After thinking about this for a while, I think we would want to stick in some LOSS of DB Level on this circuit. (As you said, it could just be too hot)

Perhaps you will want to play around with some "BUILD OUT" resistors.

That is, start with some value of resistors (Ohms = something and WATTS about 1 or 2 - watts would really not matter much)

And then, if that didn't help, try OHMS = something + 1000, and see what happens.

Next, OHMS + another GRAND and on and on until you see something change.

Well, that's my two cents and we can only hope it works.

It really seems to me that the only difference between the COPPER and the FIBER is the transmission loss or LACK THEREOF between the two.

Now, if you had a DB Meter, something like a 15A, (most of which have long since disappeared), you probably could ZERO in on this will less experimentation.

Keep us posted.

 

[phonesrus]

You should be able to find some fiber attenuators, if you think the signal is too hot.
I have not used the TN units, but have used the RAD units you mentioned several times for Definitys and a couple of Legends, with one used to extend a Demarc for a Magix 3k fiber feet into a plant. Never had a problem with them

This may not be the best place to go, but here are the dip and other settings for RAD based on Google.

http://artofhacking.com/th99/t/P-R/53458.htm

 

[SYQUEST]

The line compensation parameter is ONLY for the copper portion between the jack on the DS1 module and the NIU/CSU/ media converter. It has nothing to do with fiber. There is a section in the Transition Networks install guide for setting the line comp or LBO. The setting has to do with the length of the copper cable between the DS1 module and the T1 device it connects to.

Merlinman, you can't use a TIMMS for DS1 ckts. They are not compatable. You need a BERT or a "T-BERD" type of test device to do any DS1 measurements. Both the DS1 module and the media converter have all the adjustments you need for the DS1. NO external stuff is needed!

Comdata, what are the model numbers of the media converters you are using?

For the fiber side of the MC, Transition Networks has a list of different models for SINGLE MODE and MULTI MODE fiber with choice of output levels also for specific distances.

What kind of fiber are you using? Have the connectors been cleaned? Dirt can cause all kinds of loss.

....JIM....

 

[comdata]

Just got back from the airport and here are the results. We ran T-BERD tests with no errors end to end across the fiber. The tests ran absolutely clean. We also monitored the receive and transmit pairs out of the DS1 board on the Legend side and the T-BERD picked a constant stream of framing bit errors from the Legend's DS1 board. We have a replacement board ordered.

Jim: the model is SSDTF1013-105. The fiber is an existing 6 strand MM fiber; customer has been running data successfully off the first 2 strands. We are utilizing the next 2 strands. Again, we are running the T-BERD error free end to end. Unfortunately, I will not be able to get back to the site for a couple of weeks.