nRF52833/52840 reset pin in brownout

[dmleone]

We have 2 projects. 1 using an nRF52833 and the other using an nRF52840. In both circumstances we can cause a brownout condition. In order to remedy this situation i wanted to use the RESET pin and pull it low momentarily after Vcc rail returned from 1.6V to 3.0/3.3/3.6V depending on circuit. For the project with the nRF52840, the system can trigger a reset while in brownout when i pull the RESET pin low for a period and let it rebound to 3.6V.

The nRF52833, with the same circuit and trigger on the RESET pin, does not cause the system to go into reset and the 833 remains in brownout condition.

Is there a reason why the 840 can pull itself out of brownout using the reset pin and the 833 cannot under same conditions?

Here is the 52840 trigger sequence

Here is the 833 reset sequence

[Priyanka]

Hi Dominic,

Apologies for the delayed response. If you power the chip through VDDH (High-Voltage mode), the BOR comparator still watches the internal VDD (REG0 output), not VDDH. So the internal VDD may still be <1.6 V on the nRF52833 (e.g., regulator hasn’t restarted yet, or is loaded), so BOR is still asserted. Could you check the actual VDD pin at the SoC and not just the 3.3/3.6 V rail ?

-Priyanka

[xgroleau]

Hi Priyanka, 

We've been experiencing something similar with the nRF52833, it seems that when we are using the internal DCDC.
We can't recover from a brownout when lowering VDD to about 1.6V, raising back the voltage the chip doesn't startup, we are unabled to connect a debugger to the chip and a reset doesn't seem to fix it. Note that we are not using high voltage mode in this product.

Disabling the DCDC and using the LDO instead seems to solve the issue, but I'm not conviced if that just makes it less likely or actually solves the problem.

If internal DCDC enabled

• Power starts at 1.8V
• Lower power source at 1.6V: brown out
• 1.85V Still not alive, need complete power cycle to start properly

If internal DCDC disabled (thus using LDO

• Power starts at 1.8V
• Lower power source at 1.6V: brown out
• 1.65V Comes back and the MCU starts properly

Before issuing that fix to our customers I would like to confirm if that is an actual fix or that just makes the issue less likely. Also if that is a known/reproducible and what is the source of the issue in the nRF, if it's an issue with the startup power path or something else. 

Thank you
Xavier