BQ34Z100-G1: Problems with Learning Cycle

Does the GPCRA tool replace the learning cycle?

I tried it several times and i always get following error code: 

Error: Computation was not successful, but the specific reason was not identified.

       Please check that all files are prepared according to the documentation.
       The support personnel have been informed

What is wrong with this files?

GPCRA.zip

Hi DNS,

It turned out config.txt file was not matched with the column. Column number starts from 0.

I uploaded input and output file.

1351.GPCRA.zipGPCRA-report.zip

Thanks dMax the GPCRA0-Tool works but now i am not sure if the IT-algorithmus works correct.

If i discharge till cutoff voltage the SOC drops from 10% to 0% (First discharge) and relax for some hours the SOC jumps to 24%.

Where is defined that the OCV from 22800mV is a SOC from 24%?

Attached: GPCRA-Report and the Log 

GPCRA_(2).zipGPCRA-report_OUT_(2).zipLearningCylce Tabelle.xlsx

ChemID has a OCV vs DOD information which is hidden parameter. When the simuliation happened, the voltage was placed at around 82% DOD. So, SoC would be jumped up since the gauge predicts the RM/FCC based on Current, Temperature and DOD points when OCV is taken. Would you test the discharging cycle after engaging new RA and Qmax value to see the improvement ? Please, check if you used the right chemID value.

If you post a new log, please include the row which has register name.

I checked the ChemID with GPCCHEM and changed to 0807. (0803 before)

But i have still the same problems, i discharge the battery and after some minutes the SOC jumps to ~ 15-20%.
Was there a problem with the Learning Cycle? I use the GPCRA0 with the attached log. 

In the log the OCV at 0% SOC is 22800mV. If i discharge with higher current i cant reach this Voltage (SOC 0%  23100mV) . What i can do that the IT-Algorithmus works properly? i set the cutoff voltage at a lower level to discharge longer with higher current.0247.GPCRA-report_OUT_(2).zip2313.GPCRA_(2).zip5037.CONFIG.gg.zipLOG_Learning_cycle_LC-RA1212_4.zip

The purpose of learning cycle is to update Qmax Cell 0 and Ra value.  Accuracy test should be conducted after learning cycle.

SOC 0% happens before reaching at terminate voltage since RM/FCC is predicted value under load,temp and voltage. As you use higher current, it reaches at 0% earlier at higher voltage.  SOC jump down or jump up can be happened if your chemID is not matched with your cell , resistance table is not correct, Qmax value is not matched or load select is not set properly different with your load test condition and so on. 

Did you use the aktuell.gg.csv file for your test ? It has a lot of wrong value - 175.

Here is the condition to test the cell.

1. Find the correct chemID. What was your GPCChem results ? Was max DOD error below 5% ? What was the max DOD erorr ?

2. If chemID is matched then run a learning cycle  => It will update Qmax and Ra value. GPCRA gives Qmax and Ra value matched with your device. Update status will be updated to 06

3. Once you complete the learning cycle, apply Qmax and Ra value to your gg.csv file

4. Load mode and Load select set. Does your system use constant power or current ? Most system uses constant power mode if there are many switching power devices like buck or boost converter.

 Select a proper Load select

4. Discharge cycle test after full charge termination =>  Update status should be 06.

5. SoC jump after full discharge would be happened when OCV is taken during relaxation if relax voltage is goes up a lot than termnate voltage. If the gauge was tested at higher load current condition, SoC will jump up more because OCV voltage will be higher.

Thank you for the fast and detailed answer.

Yes, i use the aktuell.gg.csv file for my test. which values are wrong?

1.) I use the GPCHEM and get the attached report.

2.) I try to run a learning cycle but i always fail, cause this i used the GPCRA tool. (report attached)

4.) The System run with constant with ~2A with some peaks ~8-10A for some seconds  (30 sec) actually i use constant current.

I use for GPCCHEM and GPCRA the attached log.3010.LOG_Learning_cycle_LC-RA1212_4.zip8780.GPCRA_(2).zip

1856.GPCRA-report_OUT_(2).zipGPCCHEM-report.zip

The wrong values was a problem with the export-file.

I follow your introductions 1-4.
I did this several times but i never reach learning status 5.

Start with default gg.csv file and modify for my battery (two serial Panasonic lc-ra1212pg) and update CHEMID to 0807. 

1.)  I charge the battery full and relax for some hours.

2.) After that im Send IT_ENABLE and RESET and start discharge with C/8 to empty and relax for 5 hours.

3.) Charge with C/5 to full and relax >5 hours

But the learning status stays 4. What went wrong? i have no ideas what i can do.

Here my log file: 4834.LC.zipLC-RA1212.gg.zip