BQ28Z610: Learning Cycle Not Successful - Log Provided For Review

Hi Nick

The gauge does not regulate the charging current. You have to configure your charge to do that.

What chem id are you using? Did you identify the chem id using the tool gpcchem? Also, from looking at your log files, you are not charging using a cc/cv profile. That is required.

You will also need significant rest period after charge and discharge, larger than the normal 2 and 5 hours to ensure the dv/dt conditions are met.

thanks

Onyx

Hi Onyx,

I am using chem ID 0457 which we found as a good match for our cells using the gpcchem tool.

I had thought a current limited power supply set to the desired charge voltage would serve as an adequate charger and follow a cc/cv profile as we've used a similar setup for charging batteries in the past.  I have started another learning cycle using a dedicated battery charger for charging the cells.  I should be able to provide an update tomorrow with any changes to the charge behavior and learning cycle progress.

Nick

Hi Onyx,

Using the dedicated battery charger did not resolve the charging issues I was having (see attached log file).  After reviewing the log it looks like the cells might be at different levels of charge all throughout both the charge and discharge stages; this is apparent mostly at the end of both the charge and discharge stages where the cells differ in voltage by around 300mV.  I've unsuccessfully attempted in the past using the bq28z610 eval board to charge the cells individually so as to have a pair of cells that were at the same level of charge prior to commencing a learning cycle.  I'm attempting it again with a focus on the charge capacity to make sure both cells read roughly the same value before starting another learning cycle attempt.

Does this approach make sense to you?  Is there a better method for matching the cell charge levels?

Nick

Learning Cycle 110119 Log.zip

Hi Onyx,

I was able to achieve a successful learning cycle by using the bq28z610 eval board to individually discharge and then charge the two battery cells and by using the Battery Management Studio to confirm that both cells had roughly the same amount of charge passed into them before starting the 2s learning cycle with our board.  You are correct about the difficulty of using 2s LFP; I had unsuccessfully attempted a similar cell balancing strategy before.

I have enabled Lifetime Data Collection by setting ManufacturingStatus[LF_EN] = 1 and read back and saved the .srec file.  Do I only need to program this .srec file onto any other bq28z610 assembly that I would like to setup?

Thanks!

Nick