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.

BQ76952: SCD, OCD1 detection delay issue

Part Number: BQ76952
Other Parts Discussed in Thread: BQSTUDIO, EV2400

Hi TI team,


We are testing protection functionality on BQ76952. A gate driver is being used to drive low side FETs using the DDSG and DCHG pins. 
On an oscilloscope, we see an additional detection delay(8ms) along with the actual fault delay programmed on the AFE.
This additional delay is causing the AFE to take a late action in the case of SCD and OCD1. We have referred to the low side FET design resources and are testing the functions using the TI sample firmware.

For e.g., if a delay of 30us is configured for SCD, then the FETs are to be cutoff after 30us. But, we are seeing a total delay of 30us+8ms for the FETs to be OFF. What could be the possible reason behind the extra 8ms delay?

SCD does not seem to be OFF in the programmed period. The FET action is taken in a programmed delay + 8ms additional delay.  This is prolonging the SCD time. This applies even to the OCD1 functionality. 
Your valuable inputs are highly appreciated. 

Note: FETs exclusively take approx. 100us to switch OFF. 
Thanks, 
Aqib

  • Hi Aqib,

    It would be good to see your register settings which will be helpful to debug this. Do you have the [FET_CTRL_EN] bit set in the FET Options register? Do you have valid settings for these registers?: 

    Best regards,

    Matt

  • Hi Matt, 
    Thanks for you reply.

    Here are the settings you pointed at:
    FET Options - 0x0D
    CHG FET Protections A - 0x98
    DSG FET Protections A - 0xE4

    As per the TRM, SCD and OCD1 protections are comparator based. Is there a delay associated with the comparator operation? 8ms is a fixed time we notice in both these protections. 

  • Hi Aqib,

    Is your issue related to this thread? https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/1189848/bq76952-bq76952-short-circuit-protection-improper-operation

    If so, can you verify the device you are using is the latest production version (check the top marking or read the FW_VERSION() 0x0002 command)? And verify the version of BQStudio you are using?

    Regards,

    Matt

  • Hello Matt, 

    Yes, my query is exactly related to the thread you pointed at. The firmware version you saw on the other thread is of the EVM. We have designed a custom board that is being used to test the SCD, OCD1 functionality. When a gg.csv file was asked to load our register configurations, we used  EVM + BQStudio to generate the same (using bq studio version - 1.3.101).  We are not using EV2400 to program the custom board with BQStudio, rather writing a code to do the same. 

    Here are the readings of FW_VERSION() 0x0002 command (data in Hex) (TRM says this subcommand reports in Big Endian format) 
    1. Firmware version - 3600  (Could be the version 0.36 as you mentioned in the other thread)
    2. Device number - 9576 (7695 as reported in big endian)

    The part looks like a production part. But we are still trying to identify how the 8ms delay is getting added. 

    SCD threshold - 0x01 (20mV = 20A)
    Rshunt - 1mOhm
    SCD delay - 30 us
    SCD recovery - 30s

    As you can see in the scope images the FET turn off happens after 8ms and not after 30us. 
    We are measuring this across SRN-SRP. 

  • Can you capture the waveforms for the DDSG pin and the DSG FET gate? Also the DCHG pin and the CHG FET gate? This will show if the FET circuit is adding any delay. If there is delay from the circuit, it may be good to reduce the series gate resistors (R98, R95, R96).

    Regards,

    Matt

  • Hi Matt

    We have seen this behavior at DDSG pin even without the gate driver in place. i.e., The FET circuitry is isolated by removing the driver and checked the OFF time of DDSG Pin along with the drop at SRN- SRP.  DDSG Pin itself takes 8ms to turn OFF. I'm quite sure that it's the delay in detection. 
    I will share the scope image soon. Do you suggest any other test apart from the one you mentioned? 

    Thanks & Regards
    Aqib

  • Can you share all of the register settings including the register addresses? I just want to double-check since an older version of BQStudio was used at one point and the addresses were different in the proto-type version. I do not have any other ideas for why the timing might be delayed.

    Thanks,

    Matt