BQ40Z50: BQ40Z50R2 SOC updated wrongly form 100% to 80+ % after long rest

Hello,

Can you provide the configuration file of your data memory (.gg file), as well as a log of when this happens?  Is there any load during this time? Was a learning cycle done, and the ChemID updated to match your specific cells?

Cell balancing can be enabled for RELAX, CHARGE, or SLEEP modes depending on the configuration.  Refer to section 7 of the TRM (SLUUBK0A).

Hi Steve,

There is no load at the moment when SOC update taking place. The pack has gone through a learning cycle and we have selected ChemID 0x0444 for Heter 3200mAh cell.

Attached is the gg file and data log when SOC update happens.

Thanks

Bat gg.xlsx

7888.S2 charge to full then rest.xlsx

Hi Steve,

Is there any update? I have attached the log file in previous reply for your reference. Your assistance in this matter is highly appreciated.

Thanks,

Shirley

I have been looking at your data, and when the OCV update happens, it is updating the RSOC/DOD0. This is not unexpected based on the voltage and temperature plotted as well. Are you using a Heter 3200mAh cell, or something similar? LiFePO4 cells have a very flat discharge profile, so small voltage changes make a big difference. You have the settings right for DOD0EW and LFP_RELAX. Can you send a log of the full charge (the segment before what you sent) to make sure you are actually doing a full charge and not pre-maturely stopping charge?

Hi Steve,

The balancing algorithm of the chip will balance the cell with highest capacity instead of cell with highest voltage?

How do we calculate what's the correct balance time for each cell?

Thanks

Hi Steve,

Is there any update on question above?

Thanks

Hi Steve,

When we reflash the .srec into new pack, the initial balance timer is generated based on the OCV?

We have seen balance timer not calculating as expected in some packs that has gone through many cycles. In case if the balance timer is not calculating as expected, is there any way to correct it?


Thanks

Can you provide more detail about your expectations, and the conditions when cell balancing is not behaving as expected. A log showing the behavior would be useful. Please ensure that the registers for cell balancing are set as you need to get the expected behavior. This is what you have set in the .gg file you sent previously.

Advanced Charge Algorithm Cell Balancing Config Bal Time/mAh Cell 1 41 s/mAh
Advanced Charge Algorithm Cell Balancing Config Bal Time/mAh Cell 2-4 42 s/mAh
Advanced Charge Algorithm Cell Balancing Config Min Start Balance Delta 3 mV
Advanced Charge Algorithm Cell Balancing Config Relax Balance Interval 18000 s
Advanced Charge Algorithm Cell Balancing Config Min Rsoc for Balancing 40 %
Advanced Charge Algorithm Cell Balancing Config Start Rsoc for Bal in Sleep 60 %
Advanced Charge Algorithm Cell Balancing Config End Rsoc for Bal in Sleep 40 %
Advanced Charge Algorithm Cell Balancing Config Start Time for Bal in Sleep 100 hrs

Sorry for the delayed response. I am still looking at your charging and discharging data you sent, where cell 2 is behaving differently than the other three cells. There seems to be a more fundamental setup issue. Can you send a full charge, relax, discharge, relax cycle log, along with the configuration (.gg) file?

Hi Steve,

Sorry for late reply too. We are planning to redo the GG file for the pack, i will send you the data log once its done.

At the same time we also try on another chip bq4050 which doing balancing according to cell voltage difference. I find the default CEDV reading is too high for our battery cell, do you have the CEDV profile for Heter cell? For your information, below is the cell specification.

 

Also appreciate if you could enlighten me on the proper procedure to perform learning for this chip because I could not find much information from the forum.

Thanks

I am still not understanding why the balance timer is not calculating as expected. Have you been able to capture the full logs?