BQ78350-R1: BQ78350 SOC function

Hello,

This question has been assigned within the team and will be reviewed and followed up with another application engineer when possible. In the meantime please attach any associated .log/.gg files associated with the projects

Thank you,
Alan 

#1: You do not need a different ChemID if you restrict the voltage range. You must make sure that the gauge detected a full charge with your limited charging voltage so that the gauge learns what 100% SOC means in your application.

#2: This depends on how well relaxed the cells in your battery are when you get the gauge out of shutdown mode. This is a fundamental dilemma for gauging as the gauge (not limited to TI) needs to establish a baseline and depending on how much current you are drawing (or just drew), the measured cell voltages may not be indicative of depth of discharge (because they aren't relaxed) and then you will get an error. --> Make sure that you don't use shutdown mode unless you can guarantee that the bq78350 is taken out of shutdown mode when every cell in your battery is absolutely relaxed. Or you will get an initial SOC error.

Hello Dominik,

Thank you for your reply.

#1: I will redo the Golden file learning to make sure the Battery can trigger VCT == 1 and SOC = 100%.

How about the current during learning, is it OK to charge\discharge the battery with a high current (~60A) which the same with customer side? Of course, the current will be decresed to lower to meet the VCT and EDV2.

#2: The battery is NMC, the cell voltage will be stable after 30min relax based on the test data. But we will redo the test after the Golden file with longer time relax.

#3: Is it OK to keep the BQ78350 always wakeup? I mean that Keep the BQ78350 always in operation mode or sleep mode, but not shutdown, maybe for many years. Is it OK?

I know the consumption of the battery will be increased ~300uA or ~600uA, not sure any other risks.

Hello Dominik,

I did the test as below today. The battery is already full discharged

Test steps

1. Full charge to 4.1V to trigger VCT == 1,

2. Still for 1h

3. Full discharge to 2.9V, the SOC == 2%. But didn't trigger EDV2 point because of high current. (This step is a mistake)

4. Still for 1h

Then redo the shutdown and wakeup test.

Test Result

1. Shutdown then Wakeup, the SOC == 15%, Vcellmax == 3.066V, Vcellmin == 3.049V

2. Charge the battery with 60A for 15min, the SOC = 46%, Vcellmax == 3.569V, Vcellmin == 3.557V

3. Still for 10min, then shutdown and wakeup, the SOC == 42%, Vcellmax == 3.558V, Vcellmin == 3.547V

Question

1. Why the SOC is 15% when Cell Voltage is [3.049V, 3.066V]?

2. What standard steps for the golden file learning?

I know that the test steps above is only 1 cycle and didn't trigger EDV2 point. So, I'm continue the test tonight by following steps

Steps

#1, Full charge the battery with 30A, and drop to 1A for 2min to trigger VCT flag

#2, Still for 2h

#3, Full discharge the battery with 30A, and drop to 12.6A to trigger EDV2 point

#4, Still for 2h

#5, Jump to step #1 for 2 cycles in total.

These SOC jumps are due to the fact that this particular algorithm is based on previous FCC measurements and not on prediction of FCC (like Impedance Tracking, another TI algorithm).

FCC is a function of load and temperature (among others) so if these conditions change then the previously measured FCC may be significantly off and you will get an unavoidable SOC jump until the gauge measured FCC for the present conditions (and the conditions persist through future discharges - the idea is that there is some consistency in an application). If conditions change all the time then this algorithm will produce unavoidable SOC jumps because of the fundamental fact of this algorithm relying on previously measured FCC.