BQ79616: Measurement value jump between adjacent channels

Part Number: BQ79616

Tool/software:

Dear Experts,

I just observed an strange behaviour on BQ79616 board testing.

VC0-VC16 input are from resistors divider powered by a 24V AC-DC module. 

BQ79616 board can talk with a PC via modbus and below is a record for VCs value on PC. You can find over 1 hour contious measurement, from time to time, there are measurement value jump between adjacent channels and the abrupt changes in adjacent channels can cancel each other out. and it could be VC1 and VC2 some times and sometime it could be other channels. no clue.

I have double checked the voltages on channles by multiple meter which showed a flat value.

what could be the reason for this behaviour? 50/60HZ interference from AC-DC module and how to solve it?

  • Hello Ningguo,

    Where are you measuring this? Is this on our GIU?
    Do you see the same issue on our GIU?

    If you are measuring a flat line on the input via a probe, then likely this is an issue with the how you are reading the data back.

    Best,

          Quentin

  • Hi, Quentin,

    thank you for replying.

    It's not TI GIU. It;s another software like TI GIU but it talk to BQ79616 board(with a MCU together) via modbus protocol continously. the software can draw a history data as a chart which was showed ahead.

    I let the softare talk to BQ79616 board for hours and it has the data jump adjacent channels from time to time. the value jump was about 8-10mV.

    no probe on the line of all channels for this issue.

    Br. 

    Alex

  • Ningguo,

    Can you try using the TI GUI directly to the BQ79616?
    This can help us isolate if this is an issue with the BQ device or with the MCU. 

    Best,

         Quentin

  • Hi, Quentin,

    unfortunately, we are not able to use TI GUI directly to BQ on our board. there is no such connection on our board.

    SW talks to MCU via isolated RS485 and MCU on the board talks to BQ UART via a digital isolator. BQ base talks to BQs stack via daisy chain.

    no digital filter on MCU side. MCU read the value of channels' registers of BQ then fill up the registers of protocol directly. 

    this phenomenon is not reguarlly happenning. Once happened, it continued until there is power cycle the whole setup. then next time it could happen on other channels or not happen at all. for instance, it's not happended over last night which was about 8 hours.

    the amplitude is still small, 8-10mV. never see big value. no clear clue. sounds a EMC issue but not sure.

    any clues?

    Alex

  • Ningguo,
    Yes it should be lower in amplitude than that. 

    If you haven't already I would try increasing the DLPF on the ADCs.
    Otherwise, I would suspect perhaps a communication issue.

    Best,

         Quentin