I go down rabbit holes often and this pleasant distraction came from getting an old machine working and thinking about a related telecom project. I'm not sure where to turn for telephony advice, so I'll ask here in desperation.

I had a project sketched up at one point to build a T1/PRI interface from a commodity Maxim/Dallas framing chip - DS2155L+. It was more to just understand how trunk signaling worked from simple CAS/RBS up to SS7/ISDN. Everything was pinned to an FPGA so I could do dumps, injections, and expansion. I had worked out the needed line interface that involved a large coupling transformer, some blocking caps, and fuses, diodes, chokes, and varistors for transient protection. Then I realized a quad T1/PRI trunk card for my Adtran Atlas 890 probably used similar chips as Dallas had the main market-share. Sure enough it used a slightly older but pin compatible part. The interface circuit looked a lot like what I had drawn up. Validated - hoo-ray!

My curiosity still lies with transmission power and loss. There are no high voltage supplies going to the DS2155 - only 3.3V - and no boost provision which doesn't seem like it could drive any signal for meters let alone kilometers. There is a extremely large switching supply on the quad interface board with pads for large reserve caps. And I've seen references to the transmission side of line cards tying 100 - 130 volts to the center taps of the line coupling transformers. I also noticed on the Adtran design the line side center taps go to vias where the receive line side or interface sides do not. But I can't find where the vias come out.

Do I need to inductively couple a high voltage source to get my card to talk to the Atlas or a something like a channel bank across a room? I don't plan on connecting to any CO. But it would be nice to experiment with some of the trunking standards in the future.

Hoping someone can help.