dogman dan said:
Force it to balance with bms packs by doing the charge, wait half an hour, charge again, and repeat. Ride around the block to get the charger to restart, if it won't.
Well I tried the method of charging, then resting overnight to balance, then repeating a couple times. It improved a little both times, but last night I did the forced balancing routine. I could get the charger to cycle after resting, by just flipping the breaker I used to connect the battery. After 4 or 5 cycles the voltage would barely drop and the charger would cycle off in less than a minute. This morning the battery was right where it was last night, so it seems like the balancing was successful.
One thing I noticed is that when pulling a lot of watts while bleeding off some of the charge before starting the forced balancing, even when fully charged, I'm getting significant sag, something like 4 to 5 volts at 1800 watts. I didn't notice it before, so maybe I wasn't paying enough attention. 1800 watts is above what the battery and BMS rating (30 amps), although I've pulled 2000+watts for a fair distance without having the BMS shut it down. I'm sure it's not good for the cheap Chinese cells though.
Anyway, I'm going to try a few experiments on the relationship between current draw to voltage sag since I found a nice little loop in my neighborhood that has no traffic at night and a decent short steady climb so I can hold the current steady while observing the drop, then regen down to the bottom of the loop. I have three other routes for testing regen, sag over a longer distance, and the last to test motor temperature before and after adding Statorade. I'll start another thread, but will only be editing the post, since the data is only to help me geek out and understand the various relationships (and anyone else that may be interested), but don't want the thread bumping to the top.
Here's the little loop that I'll be using for the first set of tests. I noticed last night that I could set my PAS to a certain base level, then use the throttle and pedaling to keep the watts steady. The climbing section is only 13%, but that's enough to use the battery's full range, while controlling my pedaling and throttle. I'll probably mount my phone to record the data on the display and graph the results.
EDIT: I did a pre-test last night and rode around the loop twice to see what I should expect. I forgot that to get to the loop is also a decent climb up a couple of blocks at 10%, so I let the battery rest for 5 minutes before riding the loop. So climbing up, then regening down, and letting the battery rest 5 minutes, I lost 0.2 volts per loop (58>57.8>57.6). I love the regen. Pretty consistent sag at 1800 watts, at a little more than 4 volts.