BQ34Z100-R2: GPCRA0 Report Errors

Hello Jordan,

That is a strange issue especially since all the chemID and especially chemID 1224 is easily workable within your system and typically the warning you are receiving would correlate to the wrong chemID. In terms of your setup are you using the same cell part number and assured that there is not deviation from cell to cell. My best guess would be that there is some differently aged cells within you pack or that your cell vendor has poor quality control. 

Thank you,
Alan

Hello Alan, sorry for the delayed response.

That is very unfortunate to hear that this is a strange issue. We matched all our cells when we received them from our supplier and each pack has cells matched within 1.5% capacity of each other, and the same for internal impedance. All 16 cells are the same part number and there are slight deviations from cell to cell once discharging is complete, which we expect with these multi-cell applications. We cut discharge off once the lowest cell hits 3.0V to prevent over-discharging and damaging cells, so the highest voltage cell might be at 3.5V. Would that be enough to throw the tool off?

I generated a new set of chemID's using the GPCCHEM tool using a full pack instead of a single cell, assuming that we'd like chemID's for our actual application. I've attached the new chemID report. I tried using a couple of different ID's, trying to choose ones with similar capacities from the chemistry list in BQStudio, and have been getting the same result with the same errors. the only time I was able to get a "gg_out.csv" back was when I didn't scale the voltage in the 'sysrate_rel_dis_rel' file (meaning the voltages were all doubled), but that led to the SOC reporting low, around 30%, when we expected it to be closer to 80% (supplying 64V when 67.2V should be 100%)

I also read in SLUUBC9 that its recommended to use a single-cell voltage in the log file. I tried that, but to no avail. I've attached that submitted zip file as well.

I know the current gets scaled in submitted data, but should the voltage get scaled too based on whats in our gg.csv file? Is there any other route to take other than programming and trying different ChemID's?

Thanks for your help

Jordan

Griff_ChemID-report.zip       Griffon_RA0.zip

Hello Jordan,

Thank you for attaching the .gg file and the update chemID report and the team will review your parameters and chemID. But before we do I wanted to clarified that you did the calibration before the GPCCHEM report, how many cells in series and parallel, the cell datasheet and the warning below is the only issue you are currently seeing and it happened during the learning cycle?

Current calibration:
https://www.ti.com/video/5841163892001

Voltage calibration:
https://www.ti.com/video/5841160918001

Jordan F said:

Warning: DODdifference before and after discharge less than 70%     Qmax accuracy reduced: 
Warning: Voltage under load 2021 mV is higher than OCV 2011mV. Please check correctness of your chem ID.

Thank you,
Alan

Alan,

We have a custom test fixture for battery packs at our facility that lets us monitor individual cell voltages as well as chg & dsg current, so no calibration was done to the fuel gauge, since all the data gathered and submitted came from just the cell stack. we then adjust the current and voltage values based on our scale parameters before submission (if applicable). I've had more luck using the Ra0 and Rb tools with just a cell stack rather than a traditional learning cycle, since our circuitry we use in our system prevents us from fully discharging our batteries to their minimum voltage.

And yes, those are the warnings I am getting from the report sent back from the Gauging Parameter Calculator, with no "gg_out.csv" in the file sent back to me.

Our system is a 16S1P configuration. cell datasheet is attached below

Thanks,

Jordan

Fullymax evaluation spec. sheet_240343-A0_FBC966195-1S1P-5040mAh(PXT-50404741-1101-A0,with tape)-2.pdf

Hello Jordan, 

I've had a chance to review your .gg files and updated the "critical Parameters" for the gauge (document attached below for it as well). 

How to Complete a Successful Learning Cycle for the bq34z100:
file:///C:/Users/A0542843/Downloads/How%20to%20Complete%20a%20Successful%20Learning%20Cycle%20for%20the%20BQ34z100%20(8).pdf

-Design Voltage = Nominal Voltage * cells in series 
-Taper Current = Nominal Capacity / 10 
*Values I saw that needed updating 

Jordan F said:

so no calibration was done to the fuel gauge,

-Calibration is critical for the performance of the gauge and for the ChemID I would recommend using a single cell to get your chemID rather than the entire pack.

Thank you,
Alan

Alan,

the design voltage field has a maximum of 5000mV, so I think that is for one cell then it takes the amount in series into consideration to calculate the actual design voltage. We are scaling our voltage because the systems max voltage is over 65535mV. Should the number entered in the design voltage be a scaled number? For instance, the nominal voltage of our cell is 3700mV, if we scale our voltage by two should that be 1850mV?

Sorry if I wasn't clear, I will list the exact order in how I went about generating the file used for submission.

1. Generated chemID using a single cell. recieved ID 1224 as the best match.

2. Powered on our circuit board with Fuel Gauge on it using a cell stack emulator. updated our data flash parameters, calibrated fuel gauge, and programmed chem ID 1224.

3. Exported gg.csv file from data flash parameter tab in BQ studio AFTER configuring data flash parameters, calibrating, and programming ChemID. Fuel gauge is still being powered by cell stack emulator, not actual battery pack.

4. Ran Ra0 cycle according to SLUUBC9 document using only a battery pack. Fuel gauge is not connected to battery at all.

5. Recorded voltage, temp, and current using custom battery test fixture. updated the measurements to reflect what scaled numbers would be. We are scaling our current by 6 and voltage by 2, so in the log file, it is saying we are charging and discharging at 840mA (5040mA actual) with max voltage of 33600mV (67200mV actual) and min voltage of 24000mV (48000mV actual).

6. submit to gauging parameter tool the configured, calibrated, and chemID programmed gg.csv file, the data log with scaled values, and the config file

When I submit the log file with scaled values, that is when I am receiving these errors:

Warning: DODdifference before and after discharge less than 70%     Qmax accuracy reduced: 
Warning: Voltage under load 2021 mV is higher than OCV 2011mV. Please check correctness of your chem ID.

When I submit with unscaled voltage in the log file. I don't get any errors, but the parameters I get back from the report are out of bounds and wont allow me to write to the fuel gauge

Seeing this, I think I am still confused about when and what parameters get scaled in the submitted log file

Thank you for your help in clearing this up. I really appreciate the time.

Hello Jordan,

We are looking into this problem and will response with the answer as soon as possible.

Thank you,
Alan