ESP32 Forum

Good day,
I would like to ask If there is anybody who can help me with problem I can´t figure out what is happening. I have several own devices on my own PCB, which are powered by Li-Ion battery and they are generaly smart sensors. Because there is need to have low power as possible, I choose special buck-boost DC/DC converter, which is making voltage from 2,7 - 4,2 V to stable voltage 3,3 V for microprocessor. Everything like programming buttons scheme is the same, as I made a lot of PCBs this way, only difference is that there is that buck boost converter instead of linear regulator, and never had these problems. I have suspicion, that this power supply has very usntable output and is causing problems with ESP32 which I am explaining in next paragraph. So I made decision that I will try power supply the ESP with linear regulator. Every devices I had made were on this pricip and never had this issue, so I hope for improvment. But the problem after change of power supply is there still and I am not sure, if there is solution, or If I have to replace ESP too and throw it to trash... I can not tell, If there is some HW demage of chip, or if there is SW problem, which is impossible to make it work again.

Problem:
After while of programing device and looking if there is everything fine, after 1 month of smooth operation there comes problems. The microprocessor is getting into flash moode while it is booting from deep sleep or after HW reset... I read many threads on this forum about pull up resisors etc... I have 10k pullup resistor and when I started to search, I measured very low voltage (1.2 V) on gpio0, which is causing the flash mode.... I can understand this... But. There is NO reason for that... That low voltage is not caused by button or any bad connection... I tried everything. I searched, that the problem is inside of microprocessor. I tried to decrease pullup resistor to value 1k, which is helping to get ESP running again - it is raising the voltage o bit, but after while, the same problem repeats... I think, that something, like registers or efuse or omething, what can control if the pin is input or output, low or high, is somehow stuck or freezed and that cause this problem. The esp is in my opinion in output mode and switched to LOW... I tried to upload code to switch on and off gpio 0 and this is totaly fine, also as input, everything in running code is fine, but while booting, when no code is running, there is something wrong. Can anyone help me and is there any solution to that?

Thank you for your ideas and I hope we can find solution.

There's nothing internal to the ESP32 that should give you the behaviour you describe. Maybe it's something simple as flux under the chip that slowly is going conductive? Just guessing here.

Yes I was thinking about that for a while, because when I replaced ESP with new one, the problem was gone for a month... So the next devices - pcbs I cleaned with IPA very carefully, but now I think, that this problem is not caused by this. Because after several uploads of code with just setting GPIO0 to output and input and cycle this procedure, the voltage on that pin changed again to normal and the behaviour was again normal for a while, so I think its something in software.

I am also facing some how same problem .
Output GPIO voltage drop from 3.3V to 1.2V after some time .
Simple reset ' start command signal connected with GPIO with very less output current requirement .

Multiple SPI devices connected with ESP32 wroom . CS of SPI devices is enabled via decoder . Enable one communication at a time .

Please suggest solution as well . Thanks