BQ78350-R2-DEVICE-FW: IPSCALE Impact on Current Coulomb Deadband Configuration

Hi Mark,

The Current Deadband does need to be scaled, but not the Coulomb Counter Deadband. All of the data flash parameters that are in units of current (mA, uA) or capacity (mAh, cWh) will need to be scaled.

Best regards,

Matt

Hi Matt,

Thanks for the response. Are there any theories then why "relative state of charge" would not be changing when we drop our current draw from 110mA -> 60mA. I've attached two graphs that show the data we're observing.

SoC and Voltage with 60mA Draw

SoC and Voltage with 110mA Draw

I've confirmed our "current coulomb deadband" is configured to 38 (0x26) (reading 1 byte from address 0x4319), which I believes should put the current coulomb deadband at about 8mA, given a 1.25milliohm resistor.

Thank you!

Hi Mark,

After you update the IPSCALE and other parameters that are in units of current/capacity, it is important to run calibration. If the current offset is not calibrated, the problem is 10x worse with the scaling. When you collected this data, did you also read the Current() reported by the device to see if it was accurate?

Regards,

Matt

Thanks Matt! Yes, we do have the Current() reported. It's showing what we believe to be an accurate measurement of between 30-70mA, with most of the readings reporting 60mA.

That's good. I suggest updating all of the current and capacity parameters to incorporate the 10x scaling and then running calibration. If this still does not resolve the issue, I suggest adjusting to Coulomb Counter Deadband to see if it makes a difference. I do not expect it to make a difference if the current offset is calibrated, but it will provide additional clues to see the result.

Thanks,

Matt

Thanks Matt. I do believe we incorporated 10x scaling in all of our current and capacity parameters. I was not able to run calibration on the units yet, but we did try adjusting the Coulomb Counter Deadband. Upon adjust the Coulomb Counter Deadband, the devices did start properly decreasing State Of Charge. I will continue to collect data on that so we can be sure.

Thanks!
Mark

Just to follow-up from the previous email, here's some data showing the battery SoC vs. voltage again. This graph is showing the voltage of a single system with three different batteries placed on it. The first battery is from time 0 to the first drop in voltage (teal line) around 1300 minutes. The second battery was from 1300 minutes to second (smaller) drop in voltage (teal line) around 3150 minutes. And the third battery was measured from 3150 minutes to the end.

The first battery had "38" as the Coulomb Counter Deadband setting, and "3" (scaled to 30) as the Current Deadband.

The second battery had "4" as the Coulomb Counter Deadband setting, and "1" (scaled to 10) as the Current Deadband. It was the identical battery as the first battery, with just the two settings changed.

The third battery had "38" as the Coulomb Counter Deadband setting, and "3" (scaled to 30) as the Current Deadband. It was a wholly different battery than the first two in this test.

As you can see, the SoC seems to track well on the second battery, with the modified setting, but not track on batteries 1 and 3.

Thanks!

Thanks Mark. It will be interesting to see if the Coulomb Counter Deadband adjustment is still needed after the current offset is calibrated. If it still shows the same result after calibration, I will need to ask the firmware engineer to investigate why.

Thanks,

Matt

Hey Matt,

I just checked with our vendor, and the devices are being calibrated after the IPScale is set to 10E1 (0x1). So I believe we do have a properly calibrated devices. I'm currently unable to redo the calibration, so cannot confirm, but any other ideas what might be happening?

Mark

Hi Mark,

It may be that the Coulomb Counter Deadband does in fact need to be adjusted. I'd like to investigate this with our firmware team just to be sure. Can you send a log file from your latest plots?

Thanks,

Matt

Logs attached, hopefully this is what you need?

/cfs-file/__key/communityserver-discussions-components-files/196/voltage_5F00_and_5F00_soc_5F00_over_5F00_time.csv

What is your discharge current?

Ah, yep! This has voltage and current.

/cfs-file/__key/communityserver-discussions-components-files/196/voltage_5F00_current_5F00_and_5F00_soc_5F00_over_5F00_time.csv

Thanks Mark. I'll let you know what I find out.

Thank you!