ESD Causing Radio Failure

[Dave_couling]

I've noticed three PCA10059 Dongles have failed in the past week.  One on my desk and the two on a co-worker's desk.  Its the same failure mode where everything appears to function properly MCU-wise, however there is no Bluetooth Activity.  The steps I've tasked to debug the hardware are as follows:

1)  Check if any GPIO pins are shorted to ground or low impedance

2) Tested with different LFCLK settings incase it was a crystal issue

3) Tested reverted code and Nordic example code

4) Replaced crystals

5) Heatgun on the board to reflow

All firmware runs fine, LED sequences shows operation appears to be normal.  However there is no radio activity as all.  Its not a huge issue as I can order more dongles, but I've never seen this particular failure mode as on reference designs at that.  Is this a failure mode anyone else has seen before?  

Dave

[Kostic]

I recall a similar issue on custom board with 52840: everything except radio was functional.
In my case, radio initialization in FW could never complete. I took 2 identical boards, put the same code. One was always passing radio init, the 2nd was always failing (hang somewhere in radio init routine). We sent this part for screening, 1 GPIO was found to have high leakage, which suggested ESD damage. This GPIO had nothing to do with radio, but I guess ESD could damage inner circuits of the chip. 

[Ivan Herrera]

I've had two dongles and one of them almost instantly died, and I'm highly suspecting it was because the casing around my USB-hub is metal and it shorted things out on my first one. Always using the second one with a plastic USB-hub and it seems to be doing OK. 

[Kenneth]

You may also check out this thread (in case it is firmware related):
https://devzone.nordicsemi.com/nordic/short-range-guides/b/getting-started/posts/nrf52840-dongle-programming-tutorial 

See in specific: Recover after an accidental UICR erase

Though I don't think it is related if everything but the radio works.