BQ34Z100-R2: Current Scaling Confusion- CurrScale does almost nothing?

Steve Tecza

Part Number: BQ34Z100-R2
Other Parts Discussed in Thread: BQSTUDIO, BQ34Z100

Tool/software:

1. What is CurrScale (aka Curr Scale, AKA Currscale()) supposed to actually do?

In my observation, changing the value of currscale in flash (page 48, offset 31) has no effect on the reported current from Current() or AverageCurrent(), but it DOES for some reason MULTIPLY the value of AveragePower().

This is the ONLY change I can observe from the gauge. I have tested this using bq studio as well as host communication.

My chip FW version is 0x0202

To add confusion, changing the value of EnergyScale (page 48, offset 32) only seems to DIVIDE the reported value AveragePower(), and no other changes are observed.

Voltscale (page 48 offset 30) produces the expected results- the reported values of Voltage() and ChargeVoltage() are divided by voltscale.

What is going on here?

8 months ago

0 Diego 8 months ago

TI__Mastermind 26175 points

Hi,

CurrScale will divide the reported current by whatever you set as your scaling factor. In addition to setting the scale factor all parameters in BQstudio reported in mA/mAh/A/etc will need to be divided by the scaling factor.

Are you sure the scaling factor was written to dataflash?

Are you using a voltage divider?

Did you calibrate the device?

Regards,

Diego

0 Steve Tecza 8 months ago in reply to Diego

Prodigy 130 points

This is not what I can observe.

Currscale has no affect on Current, but it does Multiply average power and available energy. (Why multiply?)

I have tested this on a chip running FW 0x0201 and FW 0x0202, and the results are the same.

I have tested this using BQ studio and by using my host controller.

I have reflashed one chip with the .srec for 0x0202 and repeated the test.

I tried setting the SCALED bit in Pack Configuration, even tho it's deprecated. No change.

Can you please confirm the expected behavior with a real BQ34Z100 chip?

Here is my procedure:

Connect the BQ34z100 to bq studio. no problem, all values are updating. scan is enabled. current scale is 1.

Change current scale to 2 on the data memory tab. return to the Registers tab, current scale now reads 2.

The reported current remains unchanged, but both average power and available energy have been multiplied by 2, proving that current scale was changed in flash.

I have attached 2 screenshots to illustrate.

currscale is 1:

currscale is 2:

0 Steve Tecza 7 months ago in reply to Diego

Prodigy 130 points

Bump. Can you help please? look at my screenshots

0 Steve Tecza 7 months ago in reply to Diego

Prodigy 130 points

Hello? Can anyone help with this?

Does this chip work as advertised at all?

Is there a better way to get support?

0 Diego 7 months ago in reply to Steve Tecza

TI__Mastermind 26175 points

Hi Steve,

Respectfully, this behavior is not seen on my end. When I change Currscale to 2 my reported current is halved.

I have report this to SQM for further analysis.

Is your system discharging more than 32A, if not current scale is not needed.

Regards,

Diego

0 Steve Tecza 7 months ago in reply to Diego

Prodigy 130 points

My system is designed to work with current up to 250 amps in normal operation.

Please, what firmware version is your chip running? I can reinitialize mine by flashing with the .srec for version 0x0202, and this failure persists. I also observe the same behavior with a brand new -R2 chip fresh off the reel with FW version 0x0201.

How can I explain the discrepancy? What version of BQ studio are you using?

This issue is very urgent for me, I need to get this into production very soon and the scaling simply does not work. What can I do?

Do I perhaps have the wrong .bqz file? I had to add it manually. Can you provide the correct files for the BQ34Z100 with FW 0x0202?

Can you export the golden images from your BQ34z100 so I can flash my chips and try again?

0 Diego 7 months ago in reply to Steve Tecza

TI__Mastermind 26175 points

Hi,

I was able to recreate the problem and have reported it to SQM.

However, since this is an urgent issue, we do not have time to wait.

I am not sure what sense resistor you are using but as long as the max discharge of 250A does not exceed the SRN/SRP range the device will not be damaged. So please select a sense resistor that can meet these requirments.

Then use the calibration tab to scale the current, see thread below.

(18) [FAQ] BQ34Z100-G1: Current/Capacity Scaling - Power management forum - Power management - TI E2E support forums

Regards,

Diego

0 Steve Tecza 7 months ago in reply to Diego

Prodigy 130 points

I have programmed an external scaling factor in my host controller, which will work for now, but I'm disappointed by the loss of precision.

Do you have any idea when the problem can be corrected? What will it take, a new firmware revision for the BQ34Z100?

My actual sense resistor is 0.5mΩ, and I am using a 20x scale factor for current and energy to accommodate design capacity up to 640 amp-hours.

0 Diego 7 months ago in reply to Steve Tecza

TI__Mastermind 26175 points

Hi Steve,

I have alerted the FW and SQM teams about this, it will be fixed on the next FW revision.

Regards,

Diego

0 Steve Tecza 7 months ago in reply to Diego

Prodigy 130 points

That's wonderful. Do you have any idea when it will be ready?

I have invested a lot in my product design around this chip based on advertised features which don't work.

Why did you first say that you couldn't reproduce the issue? Are you using a software simulator instead of an actual EVM?

How did the -R2 version of this chip make it into production with such serious flaws?

I'm very disappointed in TI, I used to think your products were top shelf.