[c-nsp] Line Code Violations

[Barton F. Bruce]

*This message was transferred with a trial version of CommuniGate(r) Pro*
Do consider a simple test eliminating the CSU/DSU at the far end completely.

Just loop the circuit towards your end with a BNC jumper cable. You should 
be supplying clock at the PA-T3+ end.

You may also want to switch and try a loop towards them in a similar way, 
but they need to temporarily switch to be providing clock.

Is it cisco at both ends? Remember that the PA-T3 card can emulate several 
different brands of CSU/DSU and this is especially important if you are 
running some fractional T3 speed. I think the default is "Digital Link" 
mode. At one point, way back, even though the cisco cards do "Larscomm" as 
one of the settings, the Larscomm CSU/DSU at the far end needed some minimum 
code rev to talk to the cisco box.

Watch out for 735 cable, as it has about twice the loss of 734 and so what 
may work with 735 from a telco demarc to your equipment may NOT work when 
looped over the full cable distance! Multiply physical 735 cable distance by 
2 when setting LBO. OTOH, 734 cable matches the loss per foot the LBO 
settings in feet are designed for.

Another weird thing to check for if you have any insulated 75ohm BNC 
bulkhead fittings, is to be SURE there is continuity not only on the center 
pin (which is the obvious one to test) but ALSO on the outer shell from one 
side of the bulkhead to the other. I have almost 200 BAD ones locked away 
out of circulation that someday will go to dumpster la-la land.

You DO have real 75 ohm fittings? right! Generic BNCs are 50 ohms. You can 
cheat a little, but be careful. Cable is all 75 ohm, too? I've seen 50 ohm 
and 93 ohm all too often.

 You may find that telco folks are willing to loop towards you at various 
points. Persumably they have already run tests to loops at their demacs and 
are clean but may be willing to help you with loops at different locations.