That is the serial interface manual for all(?) TNCs, I referenced it (page 43/100) to check if I got the machine parameters right, the same information is mentioned in the post #6 in this thread (I actually found in my storage a db9-db25 null modem cable that had this exact routing), I also tried the non-block transfer mode (MP223: 0), and various other settings I could find googling for TNC RS232 issues people had, but nothing in the machine parameters suggests that it may affect the signal level, so I concluded my problem to be hardware related, not the machine parameter settings, in fact, the hardware interface manual page 6/100 shows a diagram of the expected signal levels, I attached that picture showing how signals look on scope on my control, red one is data sent from PC to TNC, blue one TNC>PC, and when I use that cable in between two PCs, then signals in both directions have the amplitude of the red waveform
Waveform looks like it tries to work. On page 24/100 there is a Note about wiring and your Model TNC155.
..
"Attention
This modification applies for hardware handshake with TNC 145 to TNC 155. In these control models, the RTS pin is not connected but tied high internally.The following modification is required for hardware handshake:Cut the line on both sides between DRS and DTR and short-circuit DSR with DTR(at the PC: pin 6 with pin 20).Without this modification, data transfer using hardware handshake is not only stopped but aborted immediately (like "power off).Do not use this configuration for TNC 335 with new hardware."
I did try that, but it made no difference at all, the signals were still -12~-3V, instead of TNCserver I also tried Hyperterminal with the correct com settings and didn't see anything at all arriving, but was able to send some random stuff to the control that it complained about.
First I think I'll just make a simple level shifter to correct the signal level and see if that makes the PC to start to receive data, and if it does, well, fixing the output buffer in the TNC shouldn't be too much of a problem, I'll update once I have some results.