fixed. the shunt transistor was turned on slightly. it was stuck at -1.38V between the base and emitter. normally when the shunt transistor turns on the base is driven all the way to about -3.8 or so because it is the full cell voltage that is pulling the base down through the comparator on the underside of the pcb.
speedfreke sent me two 20S v2.5 signalbs that had the led on constantly. both had the led lit on channel #11. after i built up a 20S pack from scraps, i measured all the base voltages and found out how it worked, and when the cell drops below the 3.60V level then the base is off and the voltage drop from the emitter is only about 10mV or less.
so i worked on the BMS that was dead first to expirement. i replaced the 5 pin comparator on the underside by desoldering the old one with solder wick and replacing it with one i swapped out of one of my parts BMSs. it worked right away. the led was off and the base was only about 10mV off the emitter.
so i worked on his good one next, figured i had it down pat since the first one had worked first time.
i replaced the comparator on the underside as i had with the first one, then plugged in the sense wires and it still had the same problem of the led being on all the time, and the base was stuck at -1.38 as before. so then i swapped out the shunt transistor, and it still did not fix it, the led was still lit and the base was at -1.38V.
so i decided to replace the comparator on the bottom again!!!! with another one from my parts BMS, and this time it worked!!! i am charging that cell up now since it had drained down with the led partially on. but i am fairly certain that was the problem so you will get your BMS back as soon as i can ship it.