So...after all that, the battery current sensor looked really good inside:
I cleaned up every connection on that side of the battery pack, took quite a while, put it back in and...nothing different.
With that sensor looking good inside, which I did not expect, that points back to the first sensor. Either it's toast, or something else is going on.
It's supposed to get 5V:
And is getting 5.08.
It's supposed to put out 2.5V at 0 current:
Which should be variable by checking against the 5V source:
Close as it was, I chalked it up to my meter not being all that accurate. But...maybe not.
I have this sensor kicking around, left over from another project:
Which runs on the same voltage and principle...it says right on it that +/-0.65V output = +/-200a current being sensed. If the one in the Insight is the same - and at this point, why not? - then at 2.61V, it already thinks 20+ amps of current is already passing through the sensor. Which would explain why the IMA system is reading a difference between the two currents and having a meltdown over it.
So...next step is to bench test the 200A sensor. Feed it 5.08 volts and see what it outputs. Feed it exactly 5V and see what it does.
Then, depending on the results, I'll decide between trying to feed exactly 5V to the one in the insight or else hooking the 200A one up in place of it.
Not much chance that the one in the insight is 200A - it's probably only 100A - but if it outputs 2.5V exactly, then it might be enough to convince the IMA not to spaz out.
Sigh. Better than throwing $200 trying to replace the sensor in there, only to find no change.