BQ34Z100-R2: Wrong SOC during discharge

Hi Nevena,

Did you select a ChemID with a DOD% error of less than 3%?

Can you share the log file of the SOC jump.

Regards,

Evan

Hi Evan,

Yes, my ChemID is 1452 and in GPC report DOD error is 0.85% for this ChemID.

Unfortunately, I don't have appropriate log file now because I haven't used BQ studio for logging. I will perform another cycle and send log file back to you.  

I forgot to add that description for this ChemID is LiCoO2/carbon 11 and the specified battery model has capacity of 16800 mAh.

I'm using LIPO battery with a capacity of 4500 mAh. Could this be a problem?

Best,

Nevena

Hi Nevena,

It should not matter as long as you followed the steps in the document below to obtain the chemID.

Simple Guide to Chemical ID Selection Tool (GPC) (Rev. A)

Please share the log file once the cycle is complete.

Regards,

Evan

Hi Evan,

I had communication issues when I tried to log data with BQStudio, so I implemented my own data logging by using MCU on our board. I haven't logged DOD0PassedQ parameter. I didn't know how to read the value.

I attached log file of charge-disacharge-charge battery cycle, .gg.csv file and data log that I used to determine ChemID with received GPC report.

0363.chg_dsg_chg.csv

2438.bq34z100-r2.gg.csv

2438.chem_id.zip

Best,

Nevena

Hi Nevena,

Thank you for sharing these files.

After reviewing the data you have sent, I do not think your gauge is reading temperature correctly, this is an important parameter for impedance track and will lead to an error in SOC. 

Are you using the internal thermistor or an external thermistor? Are the bits configured properly?

Regards,

Evan

Hi Evan,

We are using external thermistor. In the Pack Configuration register TEMPS bit is set, so I think that is fine. 

We performed another test, but this time we put thermistor between the battery cells. Previously, thermistor was just soldered to the board and didn't touch the battery at all. I have to mention that we had to disconnect battery from the board when installing the thermistor.

Can you look at the attached log file and tell me what do you think about temperature readings now? We still have wrong SOC prediction during discharge. Shold we perform 2-3 additional charge/discharge cycles in order to get valid SOC prediction, because we disconnected the battery from the board before the test?

Could anything else be a potential problem? What are the next steps that you suggest to us?

chg_dsg.csv

Best,

Nevena

Hi Nevena,

The temperature reading looks better.

Was a successful learning cycle completed? If not this can cause unexpected errors in SOC?

Achieving The Successful Learning Cycle (ti.com)

Regards,

Evan

Hi Evan,

Yes, we followed instructions from this document and as I wrote before got update status 0x06. Can you explain in more detail why the cycle could fail despite the received update status 0x06 and the instructions followed?

Can you please answer to my previous questions? What happens if we disconnect the battery? Can we expect valid SOC after learning cycle, or we need to perform 2-3 additional charge/discharge cycles? 

To sum up, we configured gauge registers, performed current and voltage calibration, determined ChemID and performed learning cycle.

What other informations would be useful to you to help us solve this problem?

Best, 

Nevena

Hi Nevena,

What I meant was a successful learning cycle completed with the temperature reading fixed.

If you reached 0x06 you completed a success learning cycle. If you do a second discharge you should get to 0x0E which means the gauge is now ready for the field.

The error in SOC is not due to a failed learning cycle. Why are you suggesting the learning cycle failed? 

If you remove the battery, you will not need to do a new learning cycle once reattached. However, if the cell is replaced with a different one then the learning cycle will need to be recompleted. After completing a successful learning cycle, you should see accurate SOC if you are not there is a problem somewhere.

You previously said "disconnected the battery from the board before the test", I am not sure what you meant by this but to clarify the learning cycle test should be performed on the battery.

Can you share your board schematic of the sense resistor?

The log file you previously shared "chg_dsg", this was taken after the learning cycle with the thermistor configured properly or was this of the learning cycle?

Is Impedance Track enabled?

Is it possible to share the .srec file?

Regards,

Evan

Hi Evan,

We will perform another learning cycle with fixed temperature readings. I didn't understand that we should do the learning cycle again.

In user manual for bq34z100-R2 I can't find update status value 0x0E . Are you sure that we need to get to 0x0E update status value after second discharge?

I don't understand what do you mean exactly when you said "if the cell is replaced with a different one then the learning cycle will need to be recompleted"? When we replace the battery with a different one, but it is still same battery type, we need to perform discharge/charge in order to be ready for the field?

"Disconnected battery from the board before the test" means that we disconnected battery from the board because we needed to install thermistors between the cells and connected battery to the board again and performed the test. I thought that it was important to mention that gauge was disconnected from power supply for a while.

I attached board schematic.

Impedance Track was enabled.

Best,

Nevena

Hi Nevena,

0x0E is achieved after the gauge takes a second Qmax update. 0x06 is achieved after the first Qmax update. After consulting with my team, you are correct that 0x0E is not required for the BQ34z100.

This is mostly dependent on the battery manufacture, If the batteries have inconsistencies from battery-to-battery SOC will be affected when they are swapped. If the batteries are consistent, you can expect reasonable performance with the old batteries learning cycle.

Understood that should not be a problem.

Is SRP connected to BAT-?

Regards,

Evan

Hi Evan,

I am a hardware engineer working with Nevena. Just to help regarding your last question. 

On the image above, battery array is connected to the BMS board "from the left side" meaning, BAT+ goes to the PAD1 and BAT- goes to the PAD2. 

Right side of the schematics goes to the Charger/Load. 

J2 is used to connect single cells to the balancer. 

Hi Ivica,

If that is the case, then I see no obvious errors in your sense resistor set up. 

Regards,

Evan

Hi Evan,

After another successful learning cycle (with fixed temperature readings) we are still facing issues with wrong SOC during discharging.

I attached log file of chrage-discharge battery cycle.

Do you have any other suggestion?

chg_dsg_fixed_temp.csv

Best,

Nevena

Hi Nevena,

Can you share your .srec file. So, I can look at your full gauge configurations. 

Also, what is you load select?

Regards,

Evan

Hi Evan,

I attached .srec file.

Load select is 1, load mode is 0.

0100_2_02-bq34z100.zip

Best,

Nevena

Hi Nevena,

Me and my team are actively looking into this for you, please allow us some time.

Are you using the most recent BQ34z100 FW? If not, can you please download the most recent version and see if anything changes.

Regards,

Evan

Hi Evan,

We are using the latest version 2.02.31. 

Best,

Nevena

Hello Nevena,

We will let you know if there is an update, it may take a couple days for the team to review the data.

Sincerely,

Wyatt Keller

Hi Wyatt,

Any update? 

Hello Nevena,

We have been running tests to confirm this behavior on our side but have not been able to recreate the issue. We are still looking into this.

Sincerely,

Wyatt Keller