This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BQ28Z610: Calculated by the chip does not match the capacity actually measured

Part Number: BQ28Z610
Other Parts Discussed in Thread: GPCCHEM, BQ40Z50, BQSTUDIO

Hi Team,

Could you help our cutomer's concern below. 

We have developed a battery with the IC BQ28Z610 And have the problem that the remaining capacity calculated by the chip does not match the capacity actually measured. We use LiFePO4 cells with a nominal voltage of 3.2V and 3.6Ah in a 2s2p configuration.

We have already carried out the following points to solve the problem:

 - We determined a chemical ID using the "GPCCHEM" tool

- We have carried out learning cycle according to document "slua777_learning_cycle".

Attached is a graph showing the problem.  

Thank you so much for your support.

Best regards,

Jonathan

  • Hello Jonathan,

    Which customer is implementing this design?

    The BQ28Z610 does not support LFP battery chemistry and it will lead to errors during relaxation period, like seen in this testing. It appears the gauge most likely took an OCV in relaxation which corresponded to the wrong DOD point. Can you share the logged data along with their .gg file?

    Gauges like the BQ40Z50 have gauging features that help mitigate the errors associated with LFP batteries.

    Sincerely,

    Wyatt Keller 

  • Hello Mr. Wyatt Keller,

    We have checked the available data sheets from the BQ28z610. There the support cell type Li-ion and Li-polymer are mentioned.

    LiFePO cells are Li-ion cells.

    My question is:

    Which data sheet or website explains that the BQ28z610 does not support LiFePO?

    Best Regards,

    Ilias Majid

  • Hello lIlias,

    When you search for the gauges on out website there is a separate search box for LFP batteries. LFP batteries are notorious for gauging and require extra functions in order to help gauge them properly. It is outlined in the datasheet and product pages, if you go to the BQ40Z50 product page you can see LFP chemistry is supported.

    https://www.ti.com/power-management/battery-management/fuel-gauges/products.html

    Sincerely,

    Wyatt Keller

  • Hello Mr Wyatt Keller,

    I read the following excerpt on the Texasinsturms website:

    e2e.ti.com/.../how-to-choose-a-multicell-gauge

    This shows that the IC BQ28z610 is suitable for LiFePo4 cells.

    Furthermore, I could not find any reference in the data sheet of the BQ28z610 that this would not be suitable for LiFePo4.

    I would like to add that the electronics regarding the cut-offs in the upper and lower range work perfectly. The problem with the calculated remaining capacity is just a blemish.

    Is there a chance to increase the accuracy by adjusting parameters or chemical ID?

    Thank you and Best Regards,

    Ilias Majid

  • Hello Ilias,

    I think there may be some confusion, technically you can get the BQ28Z610 working with LFP battery chemistry, the reason I say it is not supported is because the BQ28Z610 does not have functions like the LFP_RELAX that the BQ40Z50 uses, which makes gauging more difficult for LFP cells using the BQ28Z610. There won't be anything mentioning it is not supported, but there are no additional features like in our other gauges which help support it.

    That's why I recommended not using the BQ28Z610 if possible with LFP chemistry.

    The chem ID is proprietary, we cannot open it for editing by customers. You can tweak the Impedance Track parameters to get better results during active periods. During long relaxation periods it may not be possible due to how flat the voltage is of the battery. If you can share the .gg file of your configuration and SREC, along with the BQStudio log, I can help make some suggestions.

    Sincerely,

    Wyatt Keller

  • Hello Illias,

    From reviewing the data it is hard to tell exactly why the gauge is scaling the current. RemCap should basically be a coulomb counting value unless the smoothing engine is affecting the results.

    It would be most helpful to see the True FCC and True RemCap along with the filtered RemCap and FCC (basically what is reported by the SBS commands when smoothing is enabled).

    I think what occurred is the gauge took an OCV which change the Qmax/FCC/RemCap so the smoothing engine tried to spread out the jump during the whole discharge. by spreading out the jump over the discharge period it lead to this behavior.

    Sincerely,

    Wyatt Keller