• State Answer Suggested
  • Locked Locked
  • Replies 5 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 1364 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

BQ78350 ASCD/ASCDL faults loop even after load removal

Fred P
Expert 2235 points
Other Parts Discussed in Thread: BQ78350

I have experienced what looks to me like a BQ78350 bug: an ASCD / ASCDL loop without any load connected.

The pack(15S4P) was discharging on a 300W constant power electronic load.  
Near the end of discharge, protections tripped 3 times for about 1 second.
The pack would discharge again for about 6-8 seconds before the next event.
I suspect a repeated cell undervoltage.
At the fourth trip, discharging never resumed.
Plugging EV2300 to the pack, I could see ASCDL and ASCD flags were set.
Short-circuit tripping is definitely possible. When resuming discharge on a constant power load, a current spike can occur and be seen like a SCD (settings are delay=70us and threshold~15A).

However, with the load removed, ASCDL and ASCD flags would never clear.
ASCDL should have reset after 15s but did not, most probably because ASCD was set and never cleared.
Increasing ASCD latch limit from 0 to 1 followed by a reset command lead to a set-clear flag loop.   ASCD would clear, then ASCDL but ASCD was set again along with ASCDL shortly after. Clear-set-clear-set...
From a multimeter standpoint, DSG FET was never activated.
There was no load connected.

A shutdown and reboot brought back the pack to a proper behavior.  I can discharge it again.

I fully discharged the pack that way a few time, except that I have changed for more powerful cells since the last time I did it.
It is the first time that I experience that fault loop.

I could not find any way to break the loop except full shutdown.  
This is not an acceptable workaround for a pack in the field...


Any thoughts on that?

Regards,

Frederic

  • 0
    TI__Expert 6470 points

    Hi Frederic,

    Most likely, some kind of EMI or electrical overstress is being generated from the short circuit current loop and can upset the logic in the IC. Look for a firmware upgrade to improve short circuit recovery later this summer.

  • 0 in reply to Doug Williams4811
    Expert 2235 points
    Hi Doug,

    Safety on short-circuit events is one of the critical functionality of a BMS.
    I need to evaluate how far we can proceed with our tests while this bug is around.

    Can you imagine any safety issue related to that susceptibility?

    While waiting for the FW update is there any hardware modification that could help?

    Do you have any projected date for that FW update?

    Regards,
  • 0 in reply to Fred P
    TI__Expert 6470 points

    There is no safety issue, just a nuisance factor. Its a case of a register within the IC being corrupted by some EMI or electrical overstress which happens during the short circuit. The firmware fix, which should be available in August, is to modify the firmware to look for such corruption and fix it if found.

    In the meantime, you might try looking for the sensitive pin in the AFE that might be seeing some kind of illegal current (in or out) during the short circuit event and taking some steps to control it better.

  • 0 in reply to Doug Williams4811
    Expert 2235 points

    I experience the same loop on two different units but with the AOLD and AOLDL flags. Shutdown is the only way out.

    It happens sometimes during a smooth constant power discharge (BK8500 electronic load).

    EMC may have an important role in this issue since, in the AOLD case I experience, there is absolutely no overstress event happening. I'll look into that but that's quite a challenge...

    Just to let you know, our monitoring (MSP430 hooked to the BQ78350 SMBus) would see the flags loop but the current reading was stuck.  However, I don't know if the BQ was "nacking" all requests or if its reading was really stuck at the level when the problem occurred.  The reading was stuck at about 5A while the overload is configured for ~28A (see screenshot, sense resistor = 2mOhms)

    Anyhow, needless to say that we expect much from the next firmware release about this bug. EMC or not, such a loop is a problem.

    Hope the release is coming soon!

  • 0 in reply to Fred P
    Expert 2235 points
    R1 release did not solve that issue