We’re using Nordic Semiconductor nRF52840 and we may have some problem with RTC when we simulate a brownout. In our application we have both coin cell battery and external power (of bad quality). Occasionally the external power generate voltage drops that causes the coin cell battery to take over, but in some rare cases this may not happen and we get a brownout reset.
We’re using an external oscillator as clock source for both high- and low-frequency clocks, to get a better RTC tolerance. The problem we’re having is that sometimes when we get a brownout reset, the RTC jumps into the future approx. 3 to 8 minutes.
Our expectation is that the RTC is unchanged during the reset (i.e. able to continue counting). Is this a known problem, or is it stated that we can’t trust the RTC if a brownout reset occurs, i.e. we must explicit reset the RTC?
