SWDIO Pin shorted to ground

[jarhead]

Hello

I am having rather odd problem with SWDIO pin shorted to ground. I am using Fanstel BT840x BLE module and I have never seen this issue on my custom board until last week. 

Basically, the SWDIO pin seems shorted to ground even though board is not powered up by any source and I can not reprogram the chip ever again. Weirdly the MCU is working with the last firmware I flashed onto chip but not getting any new code through SWD pins.

I am checking the short circuit with Multimeter by touching swdio pin and ground pin. and I can see that there is a short circuit. I even put pull up resistor from VDD to SWDIO but the result was the same. SWDIO pin sees GND voltage.

I came across similar issue on this thread

https://devzone.nordicsemi.com/f/nordic-q-a/70654/esd-protection-for-swd-line

but changing the regulator did not solve my problem. Could it be that I might have damaged the SWDIO pin through ESD?

I bricked at least 5 of custom board within a week and that is pretty annoying.

Thank you in advance,

Eren.

[Turbo J]

jarhead said:

Basically, the SWDIO pin seems shorted to ground even though board is not powered up by any source

Note that absolute maximum limits in the PS. You cannot put voltage on SWDIO without powering the board first.

That is why all the commercial SWD programmers have target voltage as an input.

And yes, if you tried powering a circuit via SWDIO or another GPIO pin, the high current flow can permanently damage the pin. Note that there is a diode between (almost) all pins and VCC (except NFC pins).

[oversc0re]

Hello, 

We are experiencing the same issue. Custom board, using Ublox Nina module, nrf52840. The SWDIO is shorted to ground, can't program but it runs normally. The pin is routed directly to the programmer, nothing else. We are using NRF51-DK for programming. 

This is now the third bricked module. We bricked two modules at the beginning, then we added 220ohm resistors in series with programming wires and it worked this way for over a month. Today, the same issue again. 

DK has VTG-detect and even if there would be power applied to the SWDIO before powering the board, it would go through the protection diode to VCC. If diode current would be exceeded it could short SWDIO to VCC and not to GND which happens in our cases.

We have analyzed the board but were unable to figure out how a negative voltage could be applied to the SWDIO pin (or ground lifted above SWDIO)

This obviously does happen on these chips but we can't seem to get to the bottom of it. Any suggestion would be appreciated. 

Matic