BQ76952: Safety Shutoff only triggering once

Hello TI Team,

short update: This issue seems to only happen with the current based safety shutoffs. If i simulate a low voltage on a cell, i can reenable and trigger the condition again without issue.

But for all current based safety shutoffs, i only get one shutoff event after powerup. If i then reenable  the FETs i only get the alarm signals on the general alarm register and in the safetyAlert registers. But no shutoffs or any bits in the safetyStatus registers.

Hi Dominik,

Can you give a step-by-step of this behavior or a log file if possible? It would also be good to see your register settings. The behavior you describe is not the expected behavior.

Thanks,

Matt

Hello Matt,

I made a dump of the current setting i use in this test. See the following link. First value is the address and the second value is the value as hex ( apart from floats, which are printed as float). All missing settings are left as default.

https://pastebin.com/nJ83i3UF

I just noticed that this issue is somehow related to the CONFIG_UPDATE mode. If have the board in a situation where the bms does not switch off anymore, if i then enter and exit CONFIG_UPDATE mode, it will work again. So it is not that it works once after power on, it works once after exiting CONFIG_UPDATE mode.

So now i will try to track/log all writes to the BMS chip, to find out which command/memory write causes this issue.

Hi Dominik,

I cannot access the shared drive link you sent (we are blocked from shared drives from our network). 

Is it possible that you have accidentally entered CONFIG_UPDATE mode but did not exit? That might explain the behavior you are seeing. The protections are not active while you are in CONFIG_UPDATE mode. This mode should only be used to load the register settings initially after the device powers up.

Regards,

Matt

Hello Matt,

No, i can not be in CONFIG_UPDATE mode, as in this mode the FETs are switched off and in my scd and overcurrent tests the fets are back on. I normaly only enter CONFIG_UPDATE mode on the first boot (power-on) or when my code detects that the bms chip has restarted and we want to reload the config. As this switches off the FETs I definitly cant use it in normal use.

But i think i somehow issue a command which needs the CONFIG_UPDATE exit to get it working again. I am still tracking all the commands and read/writes and eliminating most of them to find out more on when this issue exactly starts. 

In my last test the issue started to occur after my code detected a bit in the PF Status B (SCDL was set) and then my code issues a PF_RESET command. And I just noticed that in the newer TechnicalReferences this command was removed, whith the comment it is not recommend ? So maybe this is the issue ?

Also i am not sure on why i got the SCDL bit, because i think it should be disabled in my current settings. But i will do further tests and keep you updated.

As you can not open the link, i attached the long list of paramters here.