Minimal common reset between nRF9160 and nRF52832

[okwestern]

Hi.

In a prototype built for a small group of test users (20-25), we want to implement a simple tactile reset button for use in case of an anomaly (as the prototype software will be very beta). We use nRF9160 + nRF52832. The device is permanently battery powered in a closed case, hence the need for a simple reset button.

A proper way would be an NPN for each, with a 18k 3.3V pull-up for nRF52832 and none for nRF9160 (internally pulled up to 2.2V). However, with space constraints, we prefer to limit the number of components.

The reset button will most likely not be used, or maybe only a few times over a 2 month period.
Is it possible to simply bridge the 52 and 91 reset pins, and ground both signals with one button and no extra components?
From the datasheet it seems nRF52832 should accept the 2.2V as not-reset.

The other question is, does the programming depend on the hw reset signal? At the moment, the reset pin is connected to the JLink, but is this necessary?

[Kenneth]

Hello.

It looks to me that you have different power domains here (e.g. you mention both 2.2V and 3.3V), the problem here is that 2.2V is lower than the requirement of >0.7*VDD to be interpret as a logic high input. So I do not recommend that you connect the two reset pins together no. One problem you can see is escessive current into the input buffer of the nRF52832 if you do this (the input can tristate). In addition the internal 13kohm pull up on the nRF52832 reset pin will feed the reset pin of the nRF9160, which will cause the ESD diode on this pin to continously sink current, which is not good either in the long run. Is there no way you can run both IC's on 2.2V? If not, I would suggest you look into switces that possible can connect/short two pins to ground instead of one.

Kenneth