I should have called the non-volatile storage RTC RAM as it is a small area of memory that will survive a reset but not a Power Down or EN cycle. Some people use it to preserve some state during deep sleep if using the reset to wake up.
I think the best way is to add one (or maybe even two, for both inputs) mosfets to switch them low very hard. It doesn't need to be a power mosfet, it can be something simple like a bs120. It's Rds-on is a couple of Ohms which should be sufficient. Othere you could use a power mosfet whose Rds-on is well below one Ohms. Something like a simple (non-"power") darlington transistor would work very well too, I think.
btidey wrote:No. SPIFFS ( or the recommended replacement LittleFS) is stored in a region of Flash and will survive any type of reset or power down.
I should have called the non-volatile storage RTC RAM as it is a small area of memory that will survive a reset but not a Power Down or EN cycle. Some people use it to preserve some state during deep sleep if using the reset to wake up.
Ok, I'm not using that storage area.
So...I tried to set the serial monitor to 74880 and also "Serial.begin(74880)", then I wrote a simple sketch to connect to my WiFi and to put ESP into deep sleep for 30 seconds.
The result is the following (from the serial monitor):
11:50:05.788 -> Connection established!
11:50:05.788 -> IP address: 192.168.1.26
11:50:05.788 -> entering deep sleep...
11:50:34.584 ->
11:50:34.584 -> ets Jan 8 2013,rst cause:2, boot mode:(3,6)
11:50:34.584 ->
after the 6th line nothing else happened, I tried both with EN pin and RST but the results are the same.
I have to power off the ESP, wait about 10 secs. then it boots again.
It's weird why 2 (apparently) identical ESPs behave so differently.