I have a class of devices that I want to connect to. They use a protocol co-developed by a couple companies back in the day and was never published. Electrically there is only one wire which wraps back around in the end cap. I broke open the end cap and there was a test point in it's PCB so I soldered something in and hooked it up to an analyzer. There's no clock line so I'm not immediately sure of the baud. It also does the weird thing where the line is high for ~61.5ms and low for the other 61.5. A message is placed on the rising and falling edge of this square wave and one in the center of each period. The shortest measured width seems to be .5833us. The async serial analyzer marks the better half of traffic with framing errors, not sure if that's more the strange inversion of line level from time to time or not being able to tell the baud.
Capture of a few tenths of a second of traffic:
Any hints on where to start would be greatly appreciated.