Odd problems with TDM844; all trunks/extensions answering!

Hey,

I’m helping a friend set up an Asterisk system, and am seeing some weirdness.

First of all is the classic problem of trunks not hanging up properly. The system thinks these open trunks are incoming calls, and rings back reception for every legit call that’s made. Very annoying. I’ve enabled busydetect which didn’t help; callprogress helped, but created too many other problems.

I’m using KS signalling on the trunks, and the telco (Eastlink) assures me that “cancel on disconnect” is enabled on the lines (which supposedly is the same thing as forward disconnect).

I’m checking things with a meter; Ring is -48V with respect to Tip when on-hook, good enough. During a call, Ring is about +6V (I think it’s +, I have to verify on Monday). When a call ends, it goes right back down to -48V. Is the change from 6V to -48V the “polarity reversal” that Kewlstart refers to? Or should that take a different form? I’m having a hard time finding out the specifics of the forward disconnect on loopstart lines.

(In every PBX I’ve installed to date, it was mainly a matter of hounding the telco until they found the right tech who knew what I was talking about, and then they truly turned on the forward disconnect; then things worked great.)

Even more strange, however, is that every hour or so, Asterisk things it’s seeing offhook on all extensions and rings on all trunks, simultaneously. The two things this suggests to me is a bad TDM844 card, or a floating ground somewhere. The card is behaving well otherwise, so I want to explore the floating ground possibility.

Would a floating ground somewhere cause this problem? When I put the system in, I noticed the outlet we were using wasn’t grounded (old building), so we got electricians in to ground the plug for the Asterisk system. I’m not sure if Eastlink’s telephony stuff is grounded, though. Could that cause this problem? (I would assume that the Digium card would look at the difference between the Tip and Ring, and not be too freaked at a ground difference. Or is that an incorrect assumption?)

I might try a heavier (literally!) UPS that generates its own clean power, to see if that helps, too. That’d narrow it down to a building wiring/spiking issue…

Any tips appreciated; I’m heading in monday for more analysis.