FWIW, I've been sitting on a lazy 36V 10Ah pack I got from evcomponents. Travis Gintz has been pretty helpful, my changes to the battery have been far slower.
I consistently got 2.6Ah out of it before it cut out.
First I replaced the BMS, which was a learning experience -- this is turning out to be both a frustrating and fun hobby. New BMS made no difference.
Travis suggested I had a weak cell. I opened the pack again and found one cell that measured 2.4V (!!!) while others measured 3.27V +/- 0.03V. I bought a
single cell charger, and foolishly charged the one cell while the rest of the pack wasn't totally charged.
I got 3.2 Ah the first ride after that, put it on the smart charger over night, after seeing the fully-charged LED light up and turn back off in approximately 5 second intervals (question, was I watching top-end balancing happen?). Took another test ride last night, I voluntarily stopped after 4.4Ah (out of boredom and it being near midnight) and was still drawing 7 to 10 amps with no voltage drop below 36V (usually hovered around 37) -- left it on the charger again last night and will give it a longer test ride soon.
If I have problems, I'll show up again with pictures, as I was quite puzzled about the wiring when I replaced the BMS. I thought it's supposed to go like this, according to the diagram I got from the new BMS:
Connect "B-" on BMS to battery cell 1 negative pole.
Connect "1" on BMS to battery cell 2 negative pole, then follow a pattern: connect "N" on BMS to battery cell "N+1" negative pole, if less than 12.
Connect "12" on BMS to battery cell 12 _positive_ pole.
...for a total of 13 leads from the battery pack to the BMS. Is that right?
If that is right, I'm not sure that cell 7 was wired correctly (from the day I first received it).