Other Parts Discussed in Thread: BQ34Z100, GPCCHEM
I'm trying to understand why the learning cycle failed for the BQ34Z100. Log attached
I've followed the instructions here (page 52) and here, and believe I've followed the appropriate steps and waited enough time between charge/discharge during the test. I'm using a 4S LiPo battery configuration (each cell operates between 3V (terminate) and 4.4V (full charge), with a 4500mAh capacity. I had previously used the Chemistry GPC tool to get a chemistry ID match (1601 was the best match with 0.84 Max DOD error)
Differences observed from expected behavior:
1. The FC bit was set prior to the charging cycle finishing. I continued charging for about ~15 minutes more to wait for the charge current to fall below 100mV.
2. After charging was complete, the documentation says to wait ~2 hours for the battery to relax, at which point the RUP_DIS = 0, VOK = 0, OCVTAKEN = 1, and Learned status changes from 04 to 05. This all happened as expected, but only about 20 minutes after the charge stopped.
3. Even though it appeared ready to discharge, I waited for over 2 hours per the documentation before beginning discharge
4. I discharged at 900mA (C/5). At the very end of the discharge, the fuel gauge was indicating that the battery should be completely discharged (SOC = 0, remaining capacity = 0, etc.) prior to it actually getting there. It took close to 5.5 hours to discharge completely when 5 hours was expected (the actual battery capacity appears to be higher than the nominal capacity listed in the datasheet). Shortly after indicating that the battery should be discharged, the RUP_DIS bit became set. About 10-15 minutes later the battery pack fell below its termination voltage (12V) and I stopped the discharge. The battery relaxed, but the learned status was still stuck at 04.
According to the documentation, the RUP_DIS bit becoming set during discharge can indicate that the chemistry selection is not appropiate. I used my log from the failed learning cycle with the GPC tool, and the chemistry I used (1601) comes back in the list with max DOD error = 1.31 (there are others chemistries with lower errors this time around, but it seems like this would still be an acceptable chemistry).
Any thoughts on what I need to change to get the learning cycle to complete? I'm in the process of completing another charge/discharge cycle without changing anything in hopes that it will converge after another cycle.
Any help or ideas would be much appreciated.
