nRF52832 crashed problem

[yzfy_12345]

Hi:

    MCU: nRF52832

   The device is powered by a battery. When the battery voltage drops to around 1.3V but is not completely depleted, if the device is charged at this time, there is a high probability that the device will freeze. Even after it is fully charged, it cannot be restarted. This situation occurs quite frequently.

To restore normal operation, the battery needs to be removed from the device and then power needs to be reapplied.

   How to solve this problem?

   Can it be solved by setting the BOR Register?  How to set it?

Thanks!

[Kenneth]

Hello,

We have rise time requirement on VDD here:
https://docs.nordicsemi.com/bundle/ps_nrf52832/page/recommended_op_conditions.html 

As you can see we have specified a rise time from 0V to 1.7V to be maximum 60ms. If it's slower it may not start as the power on reset (POR) signal may not propagate/triggered as intended internally in the chip. The brown out reset (BOR) will trigger if VDD fall below operating conditions, but if VDD fall low enough for the internal flip-flops to lose state, the chip will require an valid POR for the chip to start executing code again.

I think in your case you should likely monitor the power supply upon startup but also during normal operation, and if VDD get close to minimum operating conditions ensure the chip enter system OFF state until batteries are recharged or replaced. If you have the possibility to change the hardware, try to look at ways to ensure the rise time of VDD is faster, a shorter rise time will trigger the POR more easily and avoid such issue.

Hope that helps,
Kenneth

[yzfy_12345]

Hi:

Our products are battery-powered. After running out of power, it will be turned off and charged directly, and it will not start normally, just like a dead machine. But if you leave it for a while, the power will be drained and charged again, and the reverse side can be turned on normally. So we don't think it's a crash caused by a slow rise in voltage when powering up. The phenomenon is that if the power is exhausted, it can be started normally by charging it directly, but if it is not exhausted, for example, if there is about 1.3V left, charging at this time is very likely to result in it not starting normally. So I would like to confirm whether there is a function of LVD, or by setting BOR, whether this problem can be solved, thank you!

[Kenneth]

Hi,

There is no configuration that can be done in software that will change behavior here. So it seems it's some kind of electrical issue where the power falls low, but not low enough for POR to trigger when power is re-applied.

Have you done any debugging steps here? I am just wondering if you have double checked that the nRF52832 is not running, or if it's possible other ic on the board that may also be problematic here.

Kenneth

[yzfy_12345]

Hi:

After powering on again, I measured the voltage to be 3.3V, but the chip did not start properly.
Cannot operate normally. For example, after power on, the buzzer does not sound, the indicator light does not light up, the buttons do not respond, and Bluetooth radio cannot be searched.
So I think nRF52832 crashed. Other electronic components in the circuit are passive components that require control from nRF52832 to function properly

[Kenneth]

Can you share detailed chip markings?

Can you share schematic of the components close to he nRF52832?

Can you share a plot of the power supply when the issue occurs?

yzfy_12345 said:

So I think nRF52832 crashed. Other electronic components in the circuit are passive components that require control from nRF52832 to function properly

I just want to double check you don't for instance have a sensor externally that is not power cycled.

Everything point to some corner case where supply voltage drop low enough for BOR to trigger, but not low enough for POR to trigger when power supply rise again. But would be good with some more information to be sure.

Kenneth