• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 63 subscribers
  • Views 666 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

bq34z100 reports wrong SoC when in standby

Peter Marinov
Prodigy 245 points

Other Parts Discussed in Thread: BQ34Z100

I have a 11Ah 10s battery, using bq34z100 as a FG. The battery has three operation modes:

 - normal operation - charge with up to 4A or discharge with up to 20A current everything looks ok

 - standby - battery is off, but the microprocessor is operational for an hour, before going to low power mode - BMS draws about 15mA

 - low power mode - after an hour of inactivity, the battery goes into deep sleep mode with only 300uA being drawn by the BMS

The problem is, that the SoC reduction reported by the FG when the battery is in low power mode is not accurate - it looks like the FG still reads the standby current of 15mA and in a matter of 10 days the SoC drops by 15-20% although the actual capacity has not. Resetting the FG brings the SoC back to normal.

My guess is that the standby current is not measured correctly, but since the CC Offset is calibrated automatically I do not see how to overcome this issue. I have checked that it is changed from the value of the golden sample and it is in fact. My question is, is it possible that it is done during the standby mode of 1 hour and once the battery goes into a low power mode, the error is significant enough to introduce wrong reading over the time of the coulomb counter?

I am trying to attached a gg file but unfortunately don't see how.

Any advise is appreciated.

Regards,

Peter Marinov

  • 0
    TI__Guru 73830 points

    Peter,

    The device does run a CC Offset auto calibration routine when it the device enters its Sleep state. We usually do not see offset calibration issue unless there is a grounding or noise issue. Is the VSS connection from the device connected on the BAT- side of the sense resistor? If so, the device should not measure the quiescent current from the device when in Sleep.

    Tom

  • 0 in reply to ThomasCosby
    Prodigy 245 points
    Tom,

    Thank you for the quick answer. The FG is connected in exactly the same way as in the reference schematic and the Eval Board. Also the microprocessor circuitry is referenced to the same (-BAT) GNS, so the current supplied to it is not measured by the FG. My concern is not the actual current measured in standby, because it should be zero, but the error due to offset or perhaps noise.
    There are two possible reasons for this error:
    - we never put the FG to sleep, this is to prevent issues we had in the past when it wakes up and also because we monitor the temperature of the battery continuously, even in deep sleep mode
    - we are turning off the I2C pull up voltage to reduce the current consumption when the battery goes to deep sleep, so the FG is still powered up, but the I2C communication is disabled; would that mean that every time this happens the FG runs auto calibration or it is done only once and there is an internal flag that is triggered not allowing another auto calibration?

    Kind regards,

    Peter Marinov
  • 0 in reply to Peter Marinov
    TI__Guru 73830 points

    Peter

    The device run autocal when it transitions from normal mode to Sleep mode. There are a few qualifiers that must be met, before it updates the CC Offset in the data flash and these are covered in the Autocalibration section of the datasheet. You could check for a zero current reading and then place the pack into the state where the problem occurs. If the current is no longer reporting zero current, then check the CC Offset value in data flash and change it back to the original value. See if this returns the current to zero. If it does, then then is probably an offset getting introduced by your setup. 

    Tom

  • 0 in reply to ThomasCosby
    Prodigy 245 points
    Tom,

    The current reading in standby is always zero, apologies I have not clarified that from the beginning. We have 10x activated and I guess it's impossible to display currents lower than 5mA. Perhaps I have been barking the wrong tree, I think the issue is the value we've put for Initial Standby - 2mA (actually 20) as this is the consumption initially, but afterwards it drops below 1mA. According to the datasheet this value is updated only if the measured standby current is above the deadband (2mA in our case) and less than 2x Initial Standby. Well this would never happen as our standby current is zero, so I suspect it stays with -2mA all the time, hence the wrong SoC indication after 10-15 days.
    Is it ok to put 0mA as Initial Standby? I believe this would sort out the issue.

    Regards,

    Peter