Page 1 of 1

ESP32 power usage at deep sleep

Posted: Sun May 09, 2021 6:01 pm
by pietnl
For a custom board I'm building I'm experimenting with ways to make the power usage as low as possible when it's in deep sleep.
My goal would be to keep it running for a few weeks on a 800mAh lipo.

Currently the power usage of my board is 3mA at deep sleep, which would deplete my battery in about 11 days (and that would be if it never would wakeup in that period)

According to this guide: https://diyi0t.com/reduce-the-esp32-power-consumption/ it should be possible to get a much lower energy usage in deep sleep, they even tested a board that wend down to 0.011mA, I would already be happy if I can get it down to 1mA

The best performing board they found was the FireBeetle ESP32, so I checked that board out to see what components they used, apparently the voltage regulator can make a difference, I was using the AP2112-3.3, they where using the RT9080-33GJ5, which I could use as a drop in replacement, but I'm not sure that that would be enough.

I did give the FireBeetle ESP32-E a try (could not get a hold of the regular FireBeetle ESP32), but I can't get that one lower than 1.8mA. Possible it's using a different voltage regulator, but it's hard to identify what's on there.

Now my question is: Would changing the voltage regulator make a huge difference? or are there any other considerations I should take into account to make my board more energy efficient?

Re: ESP32 power usage at deep sleep

Posted: Mon May 10, 2021 2:06 am
by WiFive

Re: ESP32 power usage at deep sleep

Posted: Wed May 19, 2021 6:59 pm
by pietnl
Very interesting tread, but I've by now been able to get hold of a regular FireBeetle ESP32 module, and I can now confirm that that's actually using 0.02mA in deep sleep using exactly the same firmware (I'm actually using the simple timerWakuUp sketch that comes with the ESP core for arduino) that I tested my board with, so it seems I'm facing a hardware problem, not a software problem.