BQ27421EVM-G1A: About learning cycle have fail

Hello Chen,

This is a common issue seen with user with a mismatched chemID used within your system. I would redo a calibration in the calibration tab and rerun a GPCCHEM report to see which ChemID  has less than a 3% difference from your cell and stick with it.

GPCCHEM report:
https://www.ti.com/tool/GPCCHEM

Thank you,
Alan 

I use the G1-A EVB according to the document I need to use 0x0128 CHEM_ID，but our project definitely won't use this, so if I won't use this CHEM_ID means I can't learn？  This is the report about bq27421 why the DoD error is so high？

Hello Chen,

You would have to use another battery gauge that has the ability to unlock the entire chemID library, the issue is that if you had a battery gauge they used the entire chemID than you could use chemID which has a error % of 1.75 which is really good but having error % of 21.76 is not workable at all.

Thank you,
Alan

So if I keep use bq27421-G1A I still can't learning the battery right? because I can't let my VOK bit set up

Hello Chen, 

You would need to move to a battery gauge platform that doesn't only have pre-programmed chemID, you could still keep the same parameters.

Thank you,
Alan

But my problem now can't use the bqstudio learning cycle function, so I want to ask is, if I donn't want to change my F.G so I can't running whole cycle?

Hello Chen,

The reason why you can't complete a cell learning cycle is due to your battery gauge you selected. Within the bq27421-G1 you can only choose between four chemID's meaning that your best option is a chemID that has a error difference of 21% error difference which is not workable. You would need to move to another battery gauge that has access to the entire chemID platform to work with your cell.

Thank you,
Alan