Hi, I am not sure where to start this, it could be related to several things.
I am trying to setup an ESP32-c3 with a wake up from two different pins.
The esp32_ext1_wakeup feature is not too well documented, so first, this could be something to improve in the doc.
After some research, I finally found how to write the code without getting errors at the yaml validation.
Alas, at the compilation level I got errors, Okay… some more research, then I find a post where it says, that, finally it would not work on the C3 devices ???
Here is the link to the github post:
The fatal answer right at the bottom of the post.
Hmm, Where are the limitations listed in the doc ?
And now the big questions:
Is there a work around (apart from adding some more circuitry outside).
Is there a list of ESP32 devices that really support this feature in ESPHome ?
Will there be a change at some point for the c3, or should I request a change ?
From the Expressif doc, I do not see that the limitation is the chip itself (I may be wrong), so I assume this is rather in the ESPHome implementation of ESP32-c3.
Sounds like a hardware limitation. Personally I’d just get another board.
Not that I’m aware of. I think part of the issue here is that the new esp32 variants (C3, S2, S3 etc) have added more variety /complexity to the hardware landscape, and the ESPHome docs have struggled a bit to keep up.
Well ssieb who closed that issue knows the project very well. Sounds like it needs a docs update to explain it’s not possible plus some config validation to flag it to users if they try. If you want to contribute to the above I would ping ssieb on Discord, making reference to this thread, and ask if you can be of any help (including confirming if it’s a hardware limitation and raising any change requests wanted).
Thanks a lot for this detailed answer. I already tried to find out if other types would compile without throwing insults at me. Looks like esp32-s3 is a good candidate. But I have already a few c3 boards and if I can avoid buying again, all the better.
As I looked at the Arduino side of things(documentation), I was under the impression that this advanced option for deep sleep is available with c3 chips.
In any case, I will do as you suggest and contact ssieb.
Thanks for the additional tips. I sent a direct message to @ssieb on the discord server, but he did not seem to reply. For the time being, I will try to stick to my c3 boards and I will just split the two inputs to one GPIO pin with schottky diodes and try to workaround my problem that way. Not ideal, but I prefer that to buying, again, some new hardware that I do not know yet.
I will update this thread later when I have something relevant.
Something interesting, but maybe it is already known: I was programming a board, ESP32-c3 super mini. First, I did not use my antistatic mat, and I probably fried one that way.
Now the second was showing inconsistent results regarding the wifi connection. This was until I noticed that it was bad whenever I put it on the breadboard. Just dangling it on a USB cable, everything was fine.
Apparently, the breadboard has a big dampening effect on the signal. So I bent the pins 90 degrees on one side and installed the circuit vertically. Now everything is OK, I can experiment some programming.
I looked at my c3 boards, and yes, they look like the ones with the small gap, but I suspect this is more complicated than that. If they do not lay flat on a breadboard, they seem to work ok.
Anyway, I ordered s3 boards now and also c6, as I wanted to look at the zigbee
option and realized that it needs again another type.
Btw, I think this tread can be closed. The solution is probably to use an S3 board.
Thanks for the help.