• State TI Thinks Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 1925 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

BQ76940: MOSFETs burnt after connecting charger

SHASHANK SAGAR
Prodigy 80 points
Part Number: BQ76940
Other Parts Discussed in Thread: BQ78350-R1, CSD19536KTT,

I have fabricated a BMS using BQ76940 as AFE and BQ78350-R1 and Companion controller according the EVM's user guide. The system worked fine for a week for small discharge currents of 5A. Then I connected a 9A charger when the MOSFETs were off. Then the mosfets on the charging side, were burnt (source and drain was always conducting). After replacing the MOSFETs and the ESD suppressor diode, the system worked fine for some time and now the AFE reads only the first 5cells only. It's a 13S pack but only five are being sampled. 

MOSFETs -  CSD19536KTT

ESD Suppressors - SMCJ75A

Clamping Diodes - SMBJ30A-13-F

AFE - BQ7694001DBT

I have an urgent requirement, can I use a BQ76940EVM for 100A current with total 4 Mosfets ?

  • 0
    TI__Guru 54715 points
    Hi Shashank,
    Burning the charge FETs is not usual. The charge FETs should be kept off by the RDS for the FET on the application. It is this resistance which typically turns off and keeps off the FET. You can adjust the value from the 1M R9 to more effectively hold down the gate, it will take more current when on. You may want to look at your circuit topology to see if there was some other issue which caused the FET to be damaged, for example be sure your equivalent of the EVM Q5 is operational, otherwise the gate of the charge power FETs can not go low and the FETs may overheat.
    If only the lower 5 cells report voltage it may be a configuration issue of the bq78350-R1 AFE_Cell_Map, but that would not have changed unexpectedly. An XREADY fault on the top cell group of the bq76940 will also cause the missing voltages from the middle and top group. The gauge may be busy trying to clear that fault. Check you power connections to the top cell group, CAP3, and be sure the thermistor or resistor is attached to TS3.
    The EVM was not designed for 100A. If you connect from the EVM to another board with FETs the drive capability may be compromised by the distance. Also check the terms and conditions for the EVM, EVMs are designed for evaluation, not as a product or part of your product.
  • 0 in reply to WM5295
    Prodigy 80 points
    Thank you, the cells are all being sensed by the AFE, but cell 10 in the AFE_Cell_Map is reading 4.024V where as the actual voltage is only 3.593V.
    I have checked the diodes, they are all good.

    What else could go wrong ?

    Thanks in advance
  • 0 in reply to SHASHANK SAGAR
    Prodigy 80 points
    Also the DEV_XD register is set when I checked through I2C in the bq76940 evaluation software.
  • 0 in reply to SHASHANK SAGAR
    Prodigy 80 points

    The system is detecting a charge current as you can see in the above image. And the FETs are not responding to the commands too. Please help. I have soldered SQM120N10-3M8 Mosfets after the previous ones were burnt. Please help

  • 0 in reply to SHASHANK SAGAR
    TI__Guru 54715 points
    Hi Shashank,
    It would seem you have fixed the XREADY issue as you now have 13 cell voltages. The odd voltage at cell 9 (VC10-VC9) would seem to indicate a power problem at the VC10X power pin. Some problem there could be the cause of the XREADY. Check to see if the pin is soldered to the board. A missing resistor seems more likely to draw supply current through the diode pulling down the voltage.
    The FET_EN command is a toggle command. If you send it a second time it will turn off again. Since FETs are off in Operation Status but there is not a XCHG or XDSG it would seem the FET_EN shown in the log panel must have disabled the FETs. Look down in the bit registers area to the Manufacturing Status, the FET_EN should be set there to enable the FETs.
    FET damage is more likely due to some analog signal. Check for proper component installation & solder. If you have an EVM compare signals. As mentioned before check and consider the size of the Rgs resistor. You might test with low current first to confirm switching before using full load currents.
  • 0 in reply to WM5295
    Prodigy 80 points
    Thanks for the reply. Even though the FETs are not ON, there is a charging current detected by the system. And nothing is connected to the pack.The FETs are not responding to the toggle command. So I disconnected the BMS for further diagnosis.
  • 0 in reply to SHASHANK SAGAR
    TI__Guru 54715 points
    Hi Shashank,
    Diagnosis is a good idea.
    Current without any current might be some voltage leakage to the SRP or SRN pin, or some ground offset. It could be a damaged bq76940 or an incorrect calibration value stored in the bq78350-R1.
    Unexpected charge current in the FETs or charge FET damage as noted in the post would seem to be voltage leakage to the gate of the charge FET, or some leakage or resistance to GND if the PACK- is below GND.
    You might check that the bq76940 VSS and related GND is connected to the battery negative.