Protection from voltages on IO when powered off

[Anthony_Blighe]

We use an nRF52832 analog input to measure the voltage on a large capacitor. I think we have destroyed some chips when the capacitor is still charged and the power to the nRF is removed. This makes sense as inputs should not exceed VDD+0.3V.

We are measuring the capacitor voltage every three seconds or so. 

I thought a solution may be to put a resistor between the capacitor and the nRF input, and a small capacitor between the nRF input and ground.

If you think this makes sense, what values of R and C would be safe for the nRF?

Any other thoughts? Thanks

[Turbo J]

How big is that capacitor?

I'm asking because once you tried to exceed the 0.3V the diode on the pin begins to conduct and charge will flow into the VDD rail. Should not cause much damage unless the capacitor was in the F range.

Not sure how much it could affect the measurement, but you can also try just a normal diode between capacitor and VDD rail - that would conduct only if VDD was switched off at the wrong time.

[Anthony_Blighe]

Thanks. This will probably be 10F supercapacitor. We don't want to discharge it because it's the energy storage device. It feeds a DC-DC and when below 1.0V the DC-DC shuts down and the device powers down.