High Sleep Current Usage When BLE/WiFi is Untouched

[coop92]

Hello,

I am currently using the ESP32 (ESP-IDF v4.1) for a project and have been experiencing high sleep current consumption (~1.7mA) when I enter deep sleep without having used the BLE or WiFi radios. Whenever I do use either of these radios, I get ~.17mA during deep sleep which is much more acceptable for our use case.

As an example, there are times when we are just trying to wake up the device to grab a sensor reading and then quickly go back to sleep without turning on any radios, but during these times are when our current readings are exceptionally high. We made a workaround to essentially start BLE and then stop BLE before deep sleep and this fixes our issue, but we are wondering why this is even necessary. In the same vein when we grab a sensor reading and then communicate it out over WiFi on the same cycle, we get our acceptable ~.17mA.

Shouldn't BLE/WiFi never need to be turned off if it was never turned on in the first place?

Thank you for any help/feedback you can provide me with this issue!

[BB1964]

We're having a similar problem as well. When our system wakes up, checks in to our server using wifi, and then goes back to sleep, we get the low sleep current that we expect. When using an external radio(i.e. 900 MHz), the ESP32 would have a high current draw when it goes back to sleep, never using the wifi or BLE. When we experimented, we turned on the BLE on and back off briefly while using the external radio and gave us back our low sleep current. I have looked to see if something is not getting turned off that should but, can't find it. My guess was the RF circuitry on the ESP32 is getting turned on boot up and does get turned off until the BLE is shut down. Just a guess. The difference between a good sleep current and bad is around 1.5 mA.

[WiFive]

Something similar https://www.esp32.com/viewtopic.php?f=13&t=18529

[coop92]

Yes that 100% fixed my issue. Thank you so much!

[BB1964]

That did it, thank you for your help.

[william.ferguson.au]

I can confirm that for ESP32 IDFv4.3 that deep_sleep consumes

• 1.89 milli amps when wifi not started (and not shutdown)
• 0.04 milli amps when wifi wifi started and then shutdown

This seems a real issue with the firmware, albeit one that we can work around.
But the difference between 1.89mA and 0.04mA is vast.