Device locks up after BLE connection

[jeos]

We have a custom board with the NRF52840 running NCS v2.8.0.

We have a problem where our device will crash/lockup. From our testing so far, it seems to be related to Bluetooth connections - seemingly during or shortly after a connection.

The watchdog does not reset the device - instead, the device just becomes unresponsive. e.g., button events don't work. Removing the battery and re-inserting it always brings it back to life.

It's very hard to reproduce - for example, I did a test where I had 20x devices with constant BT connections happening for hours. 18x of the devices did 500+ connections successfully, and 2x locked up after a few hundred connections. The BT connections are fairly short-lived - always less than 30 seconds.

I realise I haven't provided any reproducible steps, as I'm currently just trying to figure out if this is a known problem. Perhaps something has been fixed in NCS > 2.8.0? If not, I'm curious if anyone has any suggestions on how we could troubleshoot this better.

Any help would be greatly appreciated! We've been investigating this for weeks now with no luck, so any help would be great.

[Kenneth]

Hello,

Is the device you refer to central or peripheral role?

Have you been able to measure the current consumption or power supply when the issue occurs? To get some indication what state the chip is in (e.g. is some code running, or is it some kind of reset state).

What is the chip markings on the devices you refer to?

Are you using external or internal 32kHz crystal? 

Anyone have reviewed your schematic?

The watchdog does not reset the device

Odd indeed.

Kenneth