BQ34Z100-G1: Current calibration via embedded processor

Hello Johann,

It seems like you are wanting to read current below 5mA and you can achieve this being lowering the Deadband current which is set to a default of 5mA. Typically we recommend leaving it at 5mA due it fitting most applications as is but lowering it would allow you see lower current values. You can also rerun a calibration to reach a more accurate current reading as well.

Thank you,
Alan

We can read a current of 0mA when the resolution is set to 5mA. The problem arises because of variations due to tolerances among components. We have a failure rate (reading 5mA) of about 30%. 70% of the units are reading 0 mA with no current flowing to/from the battery. What we are trying to achieve is a per unit calibration of current. This would have to be done by the unit's processor. I have used the TI Battery Management Studio to determine that a current calibration offset can be done, and that the current reported by the BQ34z100-G1 results in a 0 mA reading.

Hello Johann,

There is no decrease in resolution as you change the dead band current, currently from the response it seems that you are have a calibration issue. I would rerun a calibration within the calibration tab in BQstudio. Another thing too if your calibration was indeed the cause of the issue than you might have to redo the GPCCHEM to reconfirm that your ChemID was the right pick and also rerun a learning cycle to reconfirm. a 30% error with reading 5mA and below is not a normal tolerance within the product. In the meantime also check your scaling registers and make sure that all other registers are following the same scaling factor you had set it too.

Thank you,
Alan

I should have been a bit more clear. When I say that I used Battery Management Studio, I should have mentioned that that was on an evaluation board. The idea was to confirm that the BQ34Z100-G1 could be calibrated to read 0mA. I programmed the BQ34Z100-G1 to have a current offset, then I calibrated it using the Studio. That worked fine as a proof-of-concept. I now need the ability to transfer that functionality to the unit's processor. The processor communicates with the BQ34Z100-G1 via I2C. I captured the I2C traffic while the Studio was calibrating the BQ34Z100-G1, but it was considered (not by me) to be too labor intensive for our SW department to write code for the processor to implement the I2C traffic. That is why I asked if there was an existing algorithm or some code that could be programmed into the processor to calibrate the current measurement.

Hello Johann,

Refer to message below