• TI Thinks Resolved

BQ27426: Settings for Low-Power Application

Expert 4205 points

Replies: 6

Views: 112

Part Number: BQ27426

Dear Team,

our customer would like to use the BQ27426 in a low-power application. The battery is rated with 120mA and the average current in active state is 1mA, and in stand-by mode <0.01mA. For higher accuracy, the customer chose a scaling factor of 60x (following slua792).

> How does the customer need to set the discharge, charge and quit current thresholds?

Specifically, the customer sees that the Quit Current would be around 2400, whereas the TRM says a max. of 1000.
> What is the impact of having the values wrongly set up?

Thanks and best regards
Martin

Does this post answer Your question? Please mark it with the "Verify Answer" button.

  • The impact of having it wrongly set up is that you will have erroneous gauging accuracy.

    Why not use 20 as your scaling factor instead of 60?

    Since your discharge current in active state is so small even with scaling, the gauge would like just be in relax mode for the most part and make adjustments based on voltage look up. They should experimentallly determine what mode works best.

    ie  setting the gauge dsg curent  and quit current above their load current. With a 1mA current, and a 20x scale factor, discharge current can be set to 60mA while quit current can be set to 25mA. They should also test a case where dsg and quit are set to below the load current, so dsg should be set to 15 and quit to 10mA. I am leaning towards the second option more.

    thanks

    Onyx

  • In reply to Onyx Ahiakwo:

    Hello Onyx,

    thank you for the reply. Here the customer's feedback:

    The quit and dsg current thresholds are specified in .1 Hour rate units (i.e. relative to the design capacity). Therefore they are independent of the scale factor.

    The Quit current threshold can be set to a minimum of C/100.0 = 1.2 mA (with our 120 mAh battery)

    And the Dsg current threshold can be set to a minimum of C/200.0 = 0.6 mA

    Only during powerup we will enter discharge mode for 'dsg relax time' (by default 1 minute), after which we will enter relax mode for several days (load is then 1 mA).

    This leaves us with the first mentioned option and only depend on voltage look ups. The question would be how badly the accuracy will be affected (after 300 cycles / 3 years of operation). Badly enough to consider a different fuel gauge perhaps?

    Thanks in advance for your support

    Best Regards
    Martin

    Does this post answer Your question? Please mark it with the "Verify Answer" button.

  • In reply to Martin Seeger:

    Hi Martin,

    That would be difficult to say. If the resistance gets updated, then even the voltage lookup will be somewhat correct because the voltage lookup is not really a simple voltage-soc correlation. It is a simulation where the gauge notes the point on the OCV curve you are and then using the average I last run and the resistance profile simulates how much capacity will be left. Then using that capacity, determine the SOC.  

    But if your current is small such that you are not having resistance updates, and if the gauge ages significantly such that the resistance is significantly different from the resistance on the gauge, then you will have error. So the question is if your gauge is how does the resisance of the cell change with time. The qmax as well. They could opt to run periodic conditioning cycles so the gauge stays accurate.

    thanks

    Onyx

  • In reply to Onyx Ahiakwo:

    Hello Onyx,

    If I understand correctly, it is unsure whether the resistance tables will get updated in the field as we are only in discharge mode for 1 minute and our currents are small. It might be an option to increase our current consumption from C/120 to C/85 (just above the quit current threshold of C/100). Would this be enough for proper resistance table updates? Otherwise we can also consider periodic conditioning cycles at C/10 to keep the gauge accurate.

    Anyway, we will get in contact with our battery supplier to get more information on the change in cell resistance and qmax over time.

    Thank you very much for your help on this.

    Best regards,

    Marco

  • In reply to Marco Kruijswijk:

    Hello Marco,

    Thanks, please let us know if you have further questions.

  • In reply to Kang Kang:

    Hello Kang,

    We may have the option to increase our current consumption to C/85, which is just above the quit current threshold of C/100. But the product will never consume more than C/85. Is this enough current for the bq27426 to properly track the battery ageing?

    Regards,

    Marco