Unkwon Resets after Long runtime ; Several Hours

[krisnov]

To whom it ma concern,

we observed Unkwon Resets after Long runtime ; Several Hours.

We use the NRF52840.

After the reset bit 1 in RESETREAS is detected .

BUT investigation with a scope shows no valid change for the physical reset pin.

WE do noticed some changes in the VDD voltage but within valid working range.

I stepped over the description of the NRF52832:

Is it possible this is valid for the NRF52840 as well?

AND this behavior sets bit 1 in RESETREAS ?

Regards

Kristijan

[Priyanka]

Priyanka said:

What IC version are you using?

Could you provide your version?

-Priyanka

[krisnov]

Hi Pryanka,

Sorry I forget that

We use the NRF52840. CKAA-D0.

Regards

Kristijan

[Stian Røed Hafskjold]

Hi Kristijan,

Just to figure out if this RESETREAS =1 is valid or not, can you try to do a power cycle and read the RESETREAS register afterwards, and check if it is 1? I have seen previously that capacitance on the reset pin may cause the rise of the reset pin to lag behind the power on reset, so that a false pinreset is reported. 

Is pin reset in UICR enabled? Can you try to disable pin reset (erase UICR) and see if you still get these resets? Then we would know for sure if this is related to the reset pin or not.

The 52840 and 52832 are very similar so I would assume that the 300mV/300ms system reset is valid for 52840 as well.

Do you have any plots of the VDD fluctuations?

Are you reproducing these resets at normal indoor temperatures?

Best regards,
Stian

[krisnov]

Hi Stian,

When the reset pin ist activated the pin in RESETREAS is set as expected.

The problem is the other way around, reset pin is reported in RESETREAS but the pin was never set.

In the scope picture line 1 (yellow) is the reset, it never goes below a treshold to become a valid reset.

In the scope picture line 2 (blue) is VDD supply. It has a dip in the voltage for a short time.

[Stian Røed Hafskjold]

Hi,

krisnov said:

When the reset pin ist activated the pin in RESETREAS is set as expected.

The problem is the other way around, reset pin is reported in RESETREAS but the pin was never set.

Yes, I understand that the problem is that RESETREAS is reported even if the pin was never set. So to figure out if these resets are actually related the internal pin reset circuitry or not, it would be nice to do some tests to narrow down the problem. Hence my questions in the previous reply. If you disable the pin reset functionality in UICR, do you still get resets? And if so, what is the RESETREAS now?

Note that in order to disable pin reset, the UICR will have to be erased, and you have to make sure that PSELRESET[n] is not rewritten by the firmware.

On the scope are you able to zoom in to about 10 us/div?