• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 180 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.

BQ77915: Upper device for 8S li-ion BMS doesnt allow charging, lower device does

Thijs van der Ven
Prodigy 30 points
Part Number: BQ77915

Dear Team,

I unfortunately have encountered an issue regarding the upper device in a dual BQ7791500 / 08 in stacked configuration. The application is for a 8S1P li-ion pack, charging to 4.15V per cell at 1A and discharging at 8A peak.

Please find the schematic here:

BQ7791500 forum export.pdf


The cells for the upper device are at 4.032V and the lower device is around 3.95V. Balancing between cells on the same device is accurate to millivolts.
The CHG pin of the upper device seems to give 0-0.2V in respect to VSS2. DSG is giving 12V in respect to VSS2. This does not change after re-connecting the cells.
So the upper device is disabling charging for the pack.

As a test for the lower device, I connected the upper leg of R155 to the DSG pin of the upper device, in order to force a high signal on the CTRC of the lower device.
Now both fets turn on normally with about 12-13V gate voltage. So this makes me think that according to the lower device, the pack 'is fine'.

According to the datasheet, the only faults that can cause CHG to be off but DSG to be on, is overtemp, undertemp and overvoltage.
The NTC is replaced with 10K for the upper device, since only one sensor was needed. I checked on the scope: each second or so it sends a couple pulses on VTB, and the received voltage on TS is almost perfectly 50% of VTB. So it would seem unlikely to me that temperature protections are being triggered.

The pack was working fine before, but during a charging cycle the charge protection suddenly kicked in. Suspecting device damage, I replaced first the upper and then both IC's. Also swapped the 00 for the 08 device. This mostly gives the same results; upper device blocking charge, lower device not detecting faults.

I've read that the device defaults into OV protection after POR, but a cell voltage of 4.03 should be enough to allow for the 4.2V-100mv hysteresis to be satisfied right?
I am at a loss. Judging by the schematic, is there anything out of the ordinary that could make the upper device go into charge protection?

Thanks!

BQ7791500 forum export.pdf

  • 0
    TI__Guru* 81034 points

    Hi Thijs,

    My first thought is the OV protection after POR, but if you are seeing the same behavior with a BQ7791508 device (100mV hysterisis), then it should be satisfied. It might be good to drain the cells a little more to verify this is not the issue.

    I do notice some other small differences in your schematic from the stacking application report (https://www.ti.com/lit/an/slua906a/slua906a.pdf), but these do not appear to be relevant. The LPWR pin from the top device is not connected to the PRES pin of the bottom device, but I think your goal is to never use Hibernate mode which should be okay. The LD pins are not connected according to the app report either. 

    Regards,

    Matt

  • 0 in reply to Matt Sunna
    Prodigy 30 points

    Hello Matt,

    Sorry I kept you so long, but nevertheless, here is what I found:

    The BQ77915 was working perfectly (who'd have thought right ;)
    I didn't notice there was a difference in hysteresis between the two devices. I started with the '08' device and, suspecting the BMS was at fault instead of the charger, swapped to the '00' model for debugging. Unfortuatly I swapped to the '00' device before letting the pack discharge below 4.1V.
    Not knowing about the 200mV hysteresis of this device I falsely assumed that with the pack further discharged to still a little over 4V the OV recovery was not working.

    Like you suggested, doing a actual discharge test pointed me in the right direction. Me assuming it was not working actually made it worse!

    I made a charger out of a LM25518, with a CC feedback in parallel to the CV feedback. The CC feedback signal was however greatly overpowering the CV feedback signal due to poor choice of resistors and diodes. So it never went back to CV at the end of charge resulting in overvoltage. When the BQ77915 kicked in the charge current was forced low, and the charger would start outputting its actual intended end of charge voltage.
    That made it quite hard to detect after the fact. But this time assuming that the BMS was actually working made it far more easy to find.

    Thanks for pointing out the error in the LD part. For some reason, together with not connecting the current circuitry on the top device I also thought it would be fine to do away with the load detection. But now I put that in writing, it starts making a lot less sense.

    I'll post an updated schematic in due time. If you'd be willing to take a quick look at the updated LD part, I'd be really grateful.

    Thanks!
    Thijs

  • 0 in reply to Thijs van der Ven
    TI__Guru* 81034 points

    Sounds good. Glad to hear you've solved the issue!

    Matt