BQ76952: VCE11 VOLTAGE IS ABNORMAL

Hi Jian,

Could you measure the voltage between VC0 and VSS as close as possible to the device's pins?

Also, is this invalid measurement consistently wrong across multiple measurements? Or is just one measurement wrong occasionally?

Regards,

Max Verboncoeur

Hi   Max Verboncoeur,

       When  cell1  is   balancing, voltage between  VC0  and  VSS  is   1500mv.  When cell1  is  not balancing    voltage between  VC0  and  VSS  is   0.

When  this  problem  occure  , I  read  balancing  state  there is  not  balancing state  in   CB_ACTIVE_CELLS  register,   at this  time  I read  vc1   voltage  is  1278mv  in progrmme.  I   measure  the  voltage   between vc1 and  vc0   by    a   tool  , I  get  3300mv.

     This  problme  occures  a  time  in  a  day. It    contiues   a bout  2  to 5  seconds,  then  it  become  normal.

Hi Jian,

The recommended operating range for VC0 is -0.2V to 0.5V. We've observed with other customers that exceeding the recommended range causes the cell 1 voltage measurement to decrease dramatically.

If a cell is balancing, then the device should not be measuring it. The device stops balancing each cell periodically to take measurements; this is shown in the Cell Balancing application note.

I'm a little confused, how are you reading CB_ACTIVE_CELLS and determining that the cell is not balancing while the actual voltage between VC0 and VSS remains high? The voltage between VC0 and VSS should pull back down to near VSS when the device takes a measurement of the cell voltage.

Also, if the device is in SLEEP mode while this happens, then this could be the behavior described in Section 10.1.3 of the datasheet.

Regards,

Max Verboncoeur

Hi   Max Verboncoeur,

      I'm a little confused, how are you reading CB_ACTIVE_CELLS and determining that the cell is not balancing while the actual voltage between VC0 and VSS remains high? The voltage between VC0 and VSS should pull back down to near VSS when the device takes a measurement of the cell voltage.

=> I  mesure  VC1  and  VC0  by  a  tool,  the   voltage   does  not  change ,  it  is  3300mv.   The data  read by bq96952    become  1278mv   through  register         0x14U at  the same time.     I  do not  measure  vc0  and  vc1.     VC2 to  VC16  are  all   3300.   This   problem  appeares   two  weeks.

  I read  balance  state  through   register    CB_ACTIVE_CELLS      0x83  in programme  every 100ms,  when  the  problem  appeares , the  data  is 0.  

 To check  the  vc0  voltage,  we only  use  multimeter  to  mesre  when  balancing.

   The  sleep mode  is  disable.

Hi Jian,

Could you capture the voltage measurements on a scope? I fear that the multimeter's measurement integration is filtering out any rapid changes to the voltage.

Regards,

Max Verboncoeur

HI,   Max Verboncoeur，

      We  make  a test.  We  make  vc0  and  ground  short,so   the  vc0  voltage  is  0mv.    The  vc1  voltge  come   1375  again. Is  there  other  problem?

Hi Jian,

Could you capture the VC0-VSS voltage on a scope at the same time as this measurement error?

Regards,

Max Verboncoeur

Max Verboncoeur,

    The  VC0 -Vss  is  short , the  voltage is always 0mv.

VC0-VSS is  short  as   VC0  is  linked to  ground  as  following: Now   we  test  in this case.

Hi Jian,

VC0 shouldn't be directly shorted to ground, but instead connected through the series filter resistor. If VC0 is directly shorted to ground then the external balancing FET won't ever be able to turn on.

It might provide more insight into what is going on with the device if you could capture a waveform of VC1-VC0 and VC0-VSS during this behavior (without the direct short to VSS).

Regards,

Max Verboncoeur

Max Verboncoeur,

    I  check     voltage  between  VC0 and  VSS .I  find   abnormal  point.   The   peak  is  1500mv.

Max Verboncoceur,

     The  abnorm point , I do not  know  where  they com.And  I  do not  confirm  it is  the reason  for  vc1 voltage  becoming  low  .

Hi Jian,

Could you explain a few things?

Which measurement is the green line, and which is the yellow line?

Also, I'm having a hard time figuring out what the horizontal time scale is in these screenshots.

Also, why is the measurement oscillating so much? It should be rather stable. If you measure the cell itself, is it also oscillating?

Regards,

Max Verboncoeur

Max Verboncoceur,

     Yellow  line  is   a  free chanel. Green line  is used  to sample vc0 to vss .

oscilloscope recorder  is  SYNERGY  made by  America  Hi-Tech. When  deploying signal, the  line  is  osillating. I  think  this do  not  affect  the data right.

Max Verboncoceur,

    Today  we   go on  sampling   data. We find  there is  no peak voltage, the vc0 to vss  and  vc1 to vcc  is  very stable.

Besides  reading data from command  0x14 to command 0x35  ,we  also  read data from  subcommand 0x75.

We do not  find  any error  reading  way .  But  the result  is error.  

 I  think  there is  nothing  with the  vc0  and  vc1  voltage. It  look like  bq76952  adc  error.

bq76952_vc1_error.xlsx

Hi Jian,

It's difficult to tell if there is any issue on the waveforms you've provided since the vertical scaling is so large. The recommended max voltage of VC0-VSS is 0.5V. At the scaling you have the scope set to, it would be very difficult to see if the voltage ever exceeds that threshold.

Also, have you tried performing an A-B-A swap to see if this behavior follows the board or the device?

Regards,

Max Verboncoeur

Max Verboncoeur，

        The  VC0 voltage  is  not the reason  to hte  problem. There is  an  othere   reason. We  are  testing. 

Hi Jian,

I'll wait to hear the results of your testing.

Regards,

Max Verboncoeur

Max Verboncoeur，

     We found  BQ76952  going to  sleep mode.In  sleep  mode  vc1 become  abnormal   sometimes. Now  we  set  register  0x9234U   with  0x2c80.