...[Naturally you have to have the cover off the IPU to get at the stick pair connections. And then turn the IMA master switch ON.] You measure each pair by inserting the test + and - probes...
fyi, pack switch doesn't need to be ON to measure tap voltages. The wires ('taps') are a more or less direct connection to the sticks...
I'm not really sure
what kind of damage exists in briterobert's pack, that would prevent it from working for at least a day after the typical charges and deep discharges. In my experience it takes so little to produce a functioning pack that one that can't work at all has got to have a serious issue.
My read is that probably the most common type of out-right failure would result from a sequence like this: at least one cell has faster self discharge, cells become imbalanced, probably extremely so; and then by the time grid charging and discharging are undertaken, you end up over charging a lot of cells by a lot, but probably more importantly, you end up reversing the least charged cell a lot/for extended period/possibly at relatively high rate (whatever the full discharge current is at beginning of discharge)...
I don't know. Probably what's not fully appreciated around here, is that charging and discharging imbalanced cells in bulk doesn't mean each cell sees the same kind of treatment. For example, even 'zero volts' probably doesn't mean empty, it doesn't mean every cell is discharged deeply, I think. I think you can push single cells into reverse and then they simply never pop fully out of it during the process, even though full pack voltage could read zero. That zero could be the result of a weird combination of single-cell voltages - some still reversed and negative, some low, some probably still on the normal voltage plateau (~1.2V) and still charged. The combination of single-cell chemistries/conditions balance-out in a way that the still-charged cells don't have enough motive force to discharge through the reversed cells, a sort of stalemate...
So.... My guess is, with full pack charges and discharges, you can end up 'banging on' the same few cells a lot, such as reversing them, producing some kind of damage, and at some point they simply don't 'pop back', won't charge, or won't hold a charge... Meanwhile, the more or less more normal cells never get the 'treatment' they need, either... It's basically this weird race between these 'more normal cells', the ones that are charged the most (but probably still putzed-up), that
need full discharge, and the near empties - that need a full charge - yet keep getting deeply discharged and reversed in the full pack processes. You can get lucky for a while, you can find some wiggle room, get some usage, but it's really a crap shoot.
Anyway, not really sure of any 'easy' methods to get beyond briterobert's situation. You'd need to take tap voltage measurements under discharge load (I use autostop, but that requires a working pack; bulb discharger load might be enough), identify any outlier tap voltage, then take it from there. If say 'tap 9' had 14.20V or less and others had 14.40V or more, then I'd probably do tap shorts/discharges on all but tap 9, followed by a grid charge. It's actually more complicated - you need to measure tap voltages at rest near a known empty point (near neg recal), then measure under load some moments later, so you can calculate voltage drop and use the voltage drop as the indicator of an empty cell (i.e. least charged tap). Or measure in autostop under ~1 amp load, time 1, time 2, calculate change... It's pretty foolproof, but it takes leg work/know-how, I guess...
It's probably too much for most people to deal with, especially in this situation, where the pack/s seem completely dysfunctional, where prospects for revival are dim. IMA bypass is probably the way to go.
