ESP32 failure and repair

Other Hardware General Electronics
1

I hadn't used my ESP32 board for a while and it has been sat on the desk for a few days. Today I plugged it in to USB only to find that the power LED didn't come on and the serial port was not detected on the PC. I swapped USB cables, but the board still seemed dead. The board is an ESP32-DEVKIT V1 with an ESP32-WROOM module onboard.

Further investigation showed that it was possible to power the board via its VIN (5V) pin, although when plugged into USB there was still no detection of the serial port. Probing with a DMM and a set of SMD probes I found that there is a small diode (marked SL) in series with the V+ pin of the USB connector and the rest of the board including the VIN pin and the input to the 3.3V regulator. On the cathode side (input from USB supply) I was reading 4.99V, but on the anode (output to regulator) I was reading only around 1.0 volt instead of the expected 4.3V. This seems a peculiar failure mode as diodes usually fail short.

As I didn't have an SMD part to replace it with, I used what I had to hand, in this case a 1N4001 which I had in my spare bits box. While the result is not exactly pretty, after replacement the board seemed to be back in working order. The serial port is now being detected by the PC and I can upload sketches. I am still troubled by the fact that, at around 80mA, the board does seem to be drawing more current than the LoLin ESP8266 which draws around 60mA. I'm not sure whether this sounds right?

In any case I was wondering whether anyone else has come across this kind issue? The board worked when used last time, so how did it fail?

ESP32-repair-s.png
ESP32-repair-s.png1024×700 820 KB

2

Physical damage? Bad solder?
Was the diode removed really NFG?

3

I've seen boards (not this specific model) with blown USB voltage diodes and assumed it was because of a short to ground before or after the regulator making the diode go over current. These would fail open circuit. Perhaps the 1.0 V you measured was leakage through some other path?

4

Once I had found the thing where it had pinged to on the bench it turned out that due to damage at one end it was rather difficult to test and I got no readings, so the test is inconclusive but I agree that it is possible that the diode is open and that the 1.0V was down to leakage through some other path. I did wonder whether the diode might have run over current for some reason and gave out.

5

Thick USB cables can cause mechanical wear on the things they are plugged into. Repeated plugging in and removing components can cause mechanical stress on components.