ESP32-WROOM-32E internal VDD_SDIO failure
Posted: Mon Jan 31, 2022 3:25 pm
One of our custom boards with ESP32_WROOM-32E came back from user testing as non-responding. Serial port data showed that it had some issues with flash content in a boot loop containing "invalid header: 0xffffffff". We checked fuses and strapping pins according to the rare similar issue reports and found that everything is just OK with the connections. I have re-soldered the board with no results.
The module could be placed to download mode (so that we could run espefuse and esptool ) but download was always failing. I read flash id as 0xffff. I also tried to set VDD_SDIO to 3.3V in the fuses. The setting was performed OK but it did not solve the issue.
We assumed it is a flash defect.
With nothing to lose, we removed the shied from the module to make measurements to try to find out the cause of the problem. Interestingly I found that the supply voltage of the flash is 0.8V. Other pins had SPI-like comm present. As a test I soldered a cable to supply the flash from the module-external 3.3V: it helped. It works like before. No boot issues any more, flash content is there.
Seemingly the flash VDD was gone somehow (inside the module). According to the module schematics, flash is power from the ESP32 VDD_SDIO pin.
Now my questions are: what could cause such a failure? Mechanical shock? Overvoltage..? Power glitches? Have you ever met with issues like this? What are the statistics? We are just planning to go to mass production and this issue is kinda show stopper.
Where can I ask these questions if not here?
Thanks in advance,
Viktor
The module could be placed to download mode (so that we could run espefuse and esptool ) but download was always failing. I read flash id as 0xffff. I also tried to set VDD_SDIO to 3.3V in the fuses. The setting was performed OK but it did not solve the issue.
We assumed it is a flash defect.
With nothing to lose, we removed the shied from the module to make measurements to try to find out the cause of the problem. Interestingly I found that the supply voltage of the flash is 0.8V. Other pins had SPI-like comm present. As a test I soldered a cable to supply the flash from the module-external 3.3V: it helped. It works like before. No boot issues any more, flash content is there.
Seemingly the flash VDD was gone somehow (inside the module). According to the module schematics, flash is power from the ESP32 VDD_SDIO pin.
Now my questions are: what could cause such a failure? Mechanical shock? Overvoltage..? Power glitches? Have you ever met with issues like this? What are the statistics? We are just planning to go to mass production and this issue is kinda show stopper.
Where can I ask these questions if not here?
Thanks in advance,
Viktor