BQ76942: COV Safety Alert and Safety Status Problem

Hello,

Thank you for sharing those previous forums and explaining your issue. Can you share any waveforms of this as well as your configuration file (.gg)?

Also, from those previous forum posts, there is mention of how calibration could cause the difference in ADC measurement and the hardware comparator triggering.

Have you done any calibration for the cells as well? If not, this BQ769x2 Calibration and OTP Programming Guide will be useful to follow.

Best Regards,
Alexis

Thanks for your reponse,here is my power on process,First I calibrate the current,then I use voltage meter to check the VC0-VC1,VC1-VC2 and so on,The voltage value of the AFE conversion is correct by reading BQStudio,It is within 3 millivolts of the actual voltage,I think this is no problem,so i only calibrate the current and do not calibrate voltage,Finally i write OTP and restart power.(My question1 is even the BQStudio show AFE conversion is correct, do i also need to calibrate voltage?)

Here is my test,can see the BQStudio show cell1 voltage value is more than 4200mV,The actual voltage is same to BQStudio,and COV Safety Alert register set high(this register set high mean cell voltage more than COV threshold,this is no problem),but after COV delay threshold,COV Safety Status register does not work,the oscilloscope also show CHG mosfet is still on,if i adjust cell voltage jump to about 4220mV,the COV Safety Status register work,but the delay time 9.7 seconds is too long(setting COV delay threshold about 3 seconds)，Finally if i adjust cell voltage jump to about 4241mV,the COV Safety Status register work,and the delay time 3 seconds is no problem.(My question2 is whether the COV comparator have itself reference protection voltage,and the cell voltage shown in BQstudio has no relation to COV protection?)

Finally,if i set COV delay threshold less than 200ms,the protection can be triggered normally once the voltage exceeds 4200mV(athough the actual delay may still have deviation, at least protection can trigger every times),during the test i only change the COV voltage threshold and COV delay threshold,so much that I came to a conclusion that didn't seem quite right:if the voltage is high enough the setting COV delay threshold can work normally,and if the COV delay threshold is low enough the setting COV voltage threshold can also work normally.(My question3 is whether this not quite right conclusion can explaining the other settings of the gg file and the input voltage of hardware COV comparator is no problem?because the protection logic is normally)

If the gg file and the input voltage of hardware COV comparator is no problem,what can explains this strange situation,If attributed to simulate cell voltage with resistors is unstable and noise,but then I had a similar situation with a real battery shown in the following picture,when the battery charging about cell voltage 4180mV,the COV Safety Alert register also set high,suppose the above conclusion is correct,mean that the input voltage of hardware COV comparator received more than 4200mV voltage in an instant,but cell voltage shows nothing abnormal on the BQStudio(maybe BQStudio refresh frequency is too slow),only the COV Safety Alert register set high.(My question4 is this situation seem the voltage have high or low jitter,and the frequency less than 200ms, lead to the delay counter keeps resetting or cannot satisfy the delay condition,but even though using a real battery, this situation still happens mean that high or low jitter cannot avoid?)

If this situation that the voltage have high or low jitter cannot avoid,how can my hardware design or gg file configuration settings to avoid this situation happen,because this is no mcu hardware solution,so I can't deal with it algorithmically,so that the hardware protection of afe itself is important,if protection logic is not accurate(include voltage threshold and delay threshold), protection may be advanced or delayed,it is very dangerous on the real battery,i saw the datasheet、Technical Reference Manual and FAQ file,there is no mention of such problems.Finally thanks your reply and hope to solve or avoid this problem.

BQ76942.gg.csv

Hello Alexis_H,

I noticed that Vcell and OV, UV belong to two different sections, and the accuracy is also different.Can it be understood that, for example, if the actual voltage is 4200 mV, the Vcell converted voltage is 4202 mV(show in the bqstudio), and the OV comparator converted voltage in its perspective may be any value within the range of 4185-4215mV(can not see in the bqstudio).So that's can explain this situation.

Best Regards,
CoreLam