sundance-communications.com/forum Forums Wiring - Telecom Carriers - Trunks & Services Cabling Telephones Systems Wiring Punching split-jacket DSX-1 on a 568B patch panel

25 pair is not going to be enough. I would run 24 cat5 cables to connect your smartjack with PC
Since you dont know how to correctly wire and You might need pair 1245 or 1236, so in reality you need to terminate 123456 to be safe.
Correct me if I understood you wrong.

I did, in fact, say "a pair of 25-pair cables".

I'm not a total moron. Just a partial one.

I explicitly *don't want* to run 24 jackets; I'm trying to do this "right".

IE: you understood me wrong. :-)

Let's start over:

CPE (T-1 8p8c)
|
V
568B icecube
|
V
Cat 5 bulk, 6ft long, to
|
V
Patch panel; 568B punch down
|
V
Cat 5 molded patch
|
V
Patch panel; special punch down
|
V
2 25-pair cat 3s; split TX/RX
|
V
Patch panel; special punch down
|
V
patch cord; either telco silver satin or molded ethernet
|
V
Inline loopback adapter
|
V
Jack on Westel SmartJack shelf


It's the "special punch down" on each end of the 25-pair that I'm trying to nail down. Tommy notes that it should be the wh/bl and wh/or pairs *of the patch panel jacks* onto which the matching colored pairs of the RX and TX cables go, but I'm trying to confirm whether there is crossover *in* the pairs or *between the pairs*, or nowhere.

Am I really the only person ever to try to wire T-1 patch panels this way? :-)

No, some government operations have done it back in the '80s but they have done away with most...

Hopefully this helps.

There is no way for you to guarantee 100% that you will not have to flip the pairs later.

Some equipment uses a straight thru cable and other needs to have a cross-over cable. (IE Blue and orange pairs are flipped)

What I usually do is terminate my extension straight thru. Then use the proper Straight thru or Cross-over Cable as needed. Be sure to color code the different cables so that it is easy to know which one you have and are using.


You will need to terminate pairs on the blue and orange pairs of the patch panels. I would terminate straight thru.

You'll punch both panels down the same way at each end.

For the 1st 25-pair cable, you'll punch down each pair on only the White/Blue IDC slots of each jack.

So, the cable's White/Blue pair will punch down to Jack #1's White/Blue IDC terminals.

The cable's White/Orange pair will punch down to Jack #2's White/Blue IDC terminals.

Cable White/Green to Jack #3 (still on the White/Blue terminal)

All the way to the cable's Violet/Brown to Jack #24's White/Blue terminals, with the cable's Violet/Slate pair unused.

The 2nd cable will terminate in the same order, but to each jack's White/Orange IDC terminals.

So, the cable's White/Blue pair to Jack #1's White/Orange IDC Terminals

Cable's White/Orange pair to Jack #2's White/Orange IDC Terminals

Cable's White/Green pair to Jack #3's White/Orange terminals, and so on, again all the way to jack #24 having the cable's Violet/Brown terminated on the jack's White/Orange terminals. Violet/Slate ending up being spare.

Ah. That's the answer I needed. For straight-thru wiring, assuming none of the patch cords flip over, then I don't need to flip over the in-wall wiring either. Bless you my son-who-is-older-than-me. :-)

And yes, for a couple other people, I do *not* generally need to cross-over spans that are coming in from the street; that's what the Sangoma defaults to; you have to cross it to hook it up to an FXS channel bank, and I will be doing that with patch cords at the panel.

Does anyone make molded 8p8c patch cords that are crossover *for T-1*? :-)

Baylink -

A couple of things.

When I worked for GTE we ran General "T-1" Cable. This was an 8pair cable with each pair individually shielded with an overall braid. In this cable you could mix the Tx & the Rx - because of the shielding. Since then we've been running 2 x 6 pair or 2 x 25 pair cat3 shielded cables (just an overall shield) and putting the transmits on one and the receives on the other.

Tommy has described the cutdown perfectly.

Just use straight through cords. There are manufacturers of flipped cords but if a piece of equipment requires it then the manufacturer will generally supply it (in my experience).

Instead of biscuit 568B jacks, ideally you should be using RJ48X jacks that will provide a loopback when the patch cord is pulled.

When you connect a piece of eqipment that generates a T-1 Signal ("smart" equipment) to another piece of equipment that generates a T-1 signal then you wire your transmit to their receive and vice versa. When you wire "smart" gear to a piece of ewquipment that does NOT generate a T-1 signal ("dumb" equipment) - i.e. a patch panel - you wire Tx to Tx & Rx to Rx. Only one crossover in between the two pieces of "smart" equipment. Standardize on the location of that crossover!

When we ran something like this we would connect the CPE to RJ48X jacks. From those jacks we would run 2 shielded cables and terminate them on a DSX jackfield. We would do the same from the Carrier equipment - RJ48X jacks to two shielded cables and terminate them on another DSX jackfield situated next to the cpe jackfield. Then we would patch Tx to Rx, Rx to Tx at that location. This would allow us to plug in a test set at either jackfield and let us test in either direction.

What you're talking about doing will work and will be less expensive. As an aside we would not do it that way - none of our customers (AT&T, Sprint, Level 3, Goldman Sachs, Morgan Stanley etc.) would accept it. Mostly because testing would be too difficult and time consuming.

I hope this makes sense. If it doesn't - come back and ask again.

Sam

I followed about 90% of it.

My situation is this: I have agent clusters in my call center, with FXS channel banks *at the cluster*. I extend the DSX-1 from my switches out to the clusters, but for redundancy, I have 3 identical 568B cables to each cluster: Voice, Data, and Spare. Because I can't predict what the spare will be used for, the only place where I can supply the required crossover for the channel banks is *on the front of the patch bay*... because I can't always guarantee which of the 4 ports on the T-1 card will be being used to talk to the channelbanks, and which will go to the carrier.

So it's either cut my own crossover cables, find a vendor, or use those little keychain adapters.

I would *prefer* to get molded 2' T-1 crossover cables, if there's a vendor for those.

Our general in-house DSX-1 wiring is actual both-sides-in-a-cat-5 (or--6), and for the runs we do, we seem to get away with it, but I'll quantify that when my spandy-old T-berd 224 arrives this week. :-)

Thanks for all the help, guys.

Do a search on the net for

cat 5e ds1 crossover patch cable

it will pull up several sites that have DS1 crossover cables for sell.

Try to get a differnt color for straight thru verse crossover just for easy tracing in the future

Turns out there's an *amazing* amount of noise in that search, and if I quote "DS1 Crossover", I find 3 vendors, all of whom are custom at 2 ft.

I'll make 'em myself... :-)