We are having a random current leakage problem with our nRF51822 based board. Normally, our board and the application consumes around 10 uA of current. However, too often after powering it up, the current consumption stays at the level between 50...400 microamps. The level varies between the tests.
Finally we found that if the capacitors on the board are not emptied before powering up the board, that strange leakage current starts to happen. The only way to get back into the correct low-power state is to power off the system, short circuit the power-pins to empty capacitors, and then power-on the system again. This way, the leakage current always disappears and we get the expected level of around 10 uA.
My question is now, is that a known feature of nRF51822? Is there some internal circuitry that could end-up leaking current if the chip is not completely powered off before power-on? If so, what could be possible way to avoid the situation?
