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.

BQ27426: Configuration

Part Number: BQ27426
Other Parts Discussed in Thread: BQ25895, BQ2742

Hello,

I've been reading the BQ27426 datasheet, Technical Reference (sluubb0), and the Quickstart Guide (sluuah7b), as well as some threads on this forum, but I still have some questions, since I'm not 100% sure what I understood is correct.

I have a 10400mAh 3.8 Li-Ion battery with a maximum charge voltage of 4.35V. The batterie's charging cut-off current is 20mA, but my charger (BQ25895) has a minimum charge termination current of 64mA, so that's what I'm using. The accuracy of that measurement is ported by the BQ25895's datasheet as +- 20%.

With this information, my understanding is as follows:

a) I need to set

  • Chemical ID to 3230 (for 4.35V - I'm guessing this is the charge voltage?)
  • Design Capacity at 10400mAh
  • Design Energy at 39520 mWh (10400 * 3.8)
  • Terminate Voltage at whatever my system's minimum voltage is
  • Taper Rate using a Taper Current of 77mA (64 * 1.2)

b) I should save the State and RA subclasses to nonvolatile memory, and restore them to the BQ2742 on power loss if the ITPOR bit is set (meaning the BQ27426 has reset to default values). I could do this periodically (say, every couple hours) to minimize the effect of sudden power loss (our product has a power switch that can disconnect the battery from the system).

c) I can query StateOfCharge, AverageCurrent and RemainingCapacity periodically (say, every 10 seconds) to get updated information

d) Dividing RemainingCapacity by AverageCurrent will yield an estimate of the remaining time until the battery is discharged to 0%, assuming constant load and temperature

e) I should do a full learning cycle before production (as described here:), and have the firmware load the generated data (the so called "golden image") to the gauge on first-time boot. Can this process be done without an EVM?


f) There is no need to disable the battery charger once the fuel gauge reports 100% charge (which will happen because the gauge is configured with a termination current higher than that of the gauge).

Am I missing anything or is that it?

Thanks!

  • Yes, this is the correct sequence to configure and use the gauge.

    Please make sure that you use subcommands CHEM_A/B/C to select the ChemID instead of writing the ChemID number directly to the data memory.

    The learning cycle will allow accurate gauging without having to go through charge/discharge cycles on your system. The idea is to do this once with an EVM and export a gm.fs file which you will have to parse and execute in your firmware the first time you use the gauge.

    As this is a ROM gauge it will lose this information after a power cycle or reset. Periodically storing the gauge state and impedance data and restoring this after a gauge reset replaces the golden image download in your system.
  • Great, thank you very much!

    "As this is a ROM gauge it will lose this information after a power cycle or reset. Periodically storing the gauge state and impedance data and restoring this after a gauge reset replaces the golden image download in your system."
    Are you saying I shouldn't ever replace the golden image? From what I understand, this behavior (the replacing of the golden image) is correct, since the current gauge state will be more accurate (for that particular cell) than the golden image.

    Also, can you confirm point f? ("There is no need to disable the battery charger once the fuel gauge reports 100% charge (which will happen because the gauge is configured with a termination current higher than that of the gauge)")
  • The golden image is a starting point for a new cell. Each unit should be initially programmed with this golden image.

    The gauge will adjust the Ra tables and the state subclass dynamically as the cell ages during cycles.

    Therefore the golden image isn't applicable to an aged cell and you should instead make back-ups of the Ra table and state subclass in case you have to restore the gauge after a power cycle or reset.
  • Awesome, thanks for the clarification. The only thing I'm still not sure about is this (the datasheet isn't super clear on this point):

    f) There is no need to disable the battery charger once the fuel gauge reports 100% charge (which will happen because the gauge is configured with a termination current higher than that of the gauge).

    Is that correct?
  • You don't have to disable the charger once the gauge reports 100% charge. The charge termination should be tuned to a value that guarantees that the gauge measures 100% SOC as close as possible to the actual charge termination.