In this gist, I report some current consumption measurements I had performed on DevKitC-02 and DevKitM-1, the official devkits from Espressif based on MCU ESP32-C3. The following measurements are far from comprehensive, but they may give you an idea about these devkits.
A bench power supply regulated at 5V powered the boards. I didn't use the USB connector, but direct pins. I used Arduino Core for ESP32 (version 2.0.0-alpha1). The firmware starts by disabling WiFi and BLE (they will be investigated in the future).
| CPU @160MHz IDLE | CPU @160MHz* | CPU @80MHz IDLE | CPU @80Mhz* | CPU @40MHz IDLE | CPU @40MHz* | Deep sleep | |
|---|---|---|---|---|---|---|---|
| DevKitC-02 (ESP32-C3-WROOM-02) | 25.8 | 41.8 | 22.6 | 30.3 | 11.0 | 14.7 | 1.2 |
| DevKitM-1 (ESP32-C3-MINI-1) | 25.4 | 40.7 | 22.2 | 29.7 | 11.1 | 14.4 | 1.2 |
All the values are expressed in milliampere (mA).
*The CPU was keep busy running the following:
for (volatile int i = 0; i < 100000000; ++i) ;
You can note that switching from 80MHz to 40Mhz halves the power consumption, but you lost the connectivity (remember that connectivity is available only at 80/160MHz). However ESP32-C3 has Dynamic Frequency Scaling (DFS) so you can effectively alternate between high and low power modes
About power consumption while executing code, those values can be higher during more complex operation (i.e. running crypto algorithm), since they are observed during integer sums (a very narrow operation).
There are few differences between the 2 boards: the DevKitM-1 seems to be a little more efficient, but I cannot state if it is due to the intrinsic differences of components or because the ESP32-C3-MINI-1 is slightly more efficient than ESP32-C3-WROOM-02. A completely different devkit board is needed to investigate this question.
Since the RGB LED has a high idle current, I had tried to desolder it and measure current values. Without it, the values reported in the table diminish about 0.4mA.
These devkits are very basic, and they are intended to play with ESP32-C3. You shouldn't consider these boards to build low-power projects. Since I had used the unstable core v2.0.0-alpha1, these values should not be considered as definitive and future releases may improve the power consumption. Finally, these webkits are not ideal to perform advanced tests on ESP32-C3, since they are stuffed with many additional components (LDO, LED, Serial-USB converter) whose preclude the real measurement of ESP32-C3 power consumption.