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: 2S1P system (not pack) application

Part Number: BQ28Z610
Other Parts Discussed in Thread: BQSTUDIO, BQ40Z50, BQ34110, BQ34Z100-G1

Hello,

I am using the BQ28z610 in a system (not pack) application.  The battery is external to the main PCB which has the gauge on it.  The pack already has charge/discharge protection therefore I'm not using any of the pack protection elements so the schematic is simplified.  Schematic attached.  The pack is 2S1P configuration 8.4V (4.2V/cell) 600mAh.  The battery will be embedded in the device and is not user replaceable.  

The Battery voltage gauge in BqStudio always shows 0V and Cell 1 voltage and Cell 2 voltage are always 0, but Bat pin voltage and Pack pin voltage are correct.

I have disabled any protections in the protections registers, since I don't need them.  I have also raised the limits on the over-voltage and under-voltage protection.  I have cleared the CC0 bit to enable "single cell" mode.  I have also disabled balancing.  

If I use an external power supply to drive a little voltage to the pins instead of using the battery, I get a good voltage reading when less than 6V.  When the supply is set to > 6V, then BQstudio shows 0V as the voltage in registers 0x08/0x09.

Picture of the schematic is attached.  Is this device just not capable of such a configuration or is it just a matter of getting the register configuration correct?

I have followed the recommendation in this thread: 

  • I think that scales only to 6V or so. I'll have to talk to the hw engineers to find out. I'm sure this is not possible on the bq40z50 series, with this being similar to that, I'm sure that has the same limitation. Allow me a week to confirm. You have 2 threads asking the same question. I will close one out.
  • The device is designed to measure cell voltages individually in differential fashion. So in a 2S configuration, it measures VC2-VC1, then VC1-VSS. When in 1S configuration, VC1-VSS is the cell voltage. The cell voltage measurement is done by the ADC with fullscale range set to ~6V, which is why you cannot measure a voltage higher than 6V. If you can wire out the midpoint voltage from your pack to VC1, such that VC2-VC1 < 5V and VC1-VSS < 5V, it should work ok.

    Other options you might consider are bq34z100-G1 or bq34110. These are designed for use in a configuration like you have, where the gauge measures the top-of-stack, without needing contact to any of the intermediate nodes in the stack. They also don't integrate protection, so that also fits your case.

    Thanks,

    Terry
  • Hi Terry,

    The BQ34Z1x0 was my first choice however being a TSSOP it was just too large.  I was just too space constrained and it required too much support circuitry for what I was getting.  I could make the whole circuit of the BQ28z610 fit in the same space as the courtyard of the BQ34Z1x0 IC itself.  It's too bad the BQ34 series doesn't come in smaller packages.  Anyhow, I saw another thread where someone asked the question "can I use this with a 2S pack without the center balance tap" and they got a resounding yes, just put it in 1Cell mode and change your registers.  There are other reasons why we can't bring out the center balance connection.  Anyhow, water under the bridge.

    Luckily I don't need "excellent" gauging and I ended up using a resistor divider (47K x 2) from VC2 to GND with the midpoint of the divider going to VC1.  That gets me in the ball park.  At least I'm reading "something".  Whether it is super accurate or not doesn't matter for my application.  As I go through the learning process, I'll find out how good or bad the workaround really is.

  • Hi Jason,

    Ok, the resistor divider should definitely work for you - the gauge will assume each FET is half the full stack voltage, which is probably reasonable if you don't get significant imbalance. The only downside is that your resistor string will consume current continuously, you'll have to evaluate whether that's ok in your application.

    Thanks,

    Terry
  • Hi Terry,
    One other thing. There seems to be a bit of current draw that is pulling the VC1 pin down every second or so. I've added about 10uF of capacitance on this pin in an attempt to balance it out, but it's just too much current. I have a hard time thinking it's the AFE doing it's measurements. Would this be the "Cell Detach" feature? Is there a way to disable that feature?

    Best Regards,
    Jason
  • Hmmm, that sounds odd, I don't know why it would be drawing that much current.  Let me ask around, but offhand I'm not sure what might be causing it, or how to reduce it further through settings.

  • How much current are you seeing? The voltage measurement might take ~2uA for ~30ms every 250ms, which would cause ~40mV of variation on your divider. If more than that, you might make sure you have cell balancing disabled.
  • Hi Terry,

    I don't see the current on VC1 until I do a register read from BQStudio.  I see about a 2V deflection from 4V down to about 2 volts when a do any sort of reading of registers in BQStudio.

    HA!  I just figured it out.  The I2C lines are pulled up to VC1 net when P2 and P3 are shorted!

    I changed to using external I2C pull-ups since I'm using an EV2300 and that fixed it!

    Thank you for your help!