WiFi switch with low power consumption (0.03W) by piotr_go
Recently, when building a controller for my cooker, I decided to experiment with something simpler. Admittedly, I'm not keen on Wi-Fi control, but I was curious to see how low the current consumption could currently be. In the off-the-shelf solutions I tested, the consumption ranged from 0.3 W to 1 W. So I bought some Zigbee modules, a couple of Wi-Fi modules, a couple of different transformerless inverter ICs and bistable relays. A bit of coding and. A pleasant surprise. 0.1 W in the first tests, and after a couple of changes in the software + enabling power saving less than 0.05 W.... With WiFi... Eeeeee? It didn't stop there. By modifying the power supply slightly, I managed to get down to around 30 mW. I even assembled my own meter to confirm the results. I was expecting something in the region of 0.1 W. I decided not to use the Zigbee version.
I wrote the softwares using ESP-IDF 5.5, without any bytes, apps or clouds. Control from a browser on/off + simple configuration. There is also the possibility of updating the FW.
To make things less rosy, the whole thing took a while. I am completely unfamiliar with web development, some 90% of my programming time was taken up with it.
There was also a bug that was causing all sorts of things to happen, I got stuck until I found it. Reset! Apparently the capacitor + resistor in the module wasn't enough and ESP was able to crash at odd times. I connected a reset to the GPIO and set it to high after startup, helped.
Used Wi-Fi module ESPC2-02 from DOIT, 230 V-->5 V inverter BP2525D, 5 V-->3.3 V inverter M3406, bistable relay control GM8023B.
Schedule:
Building: Building:
Dimensions: 44x44x22 mm
Current consumption:
Measured with a plug-in meter and an average hourly measurement when powered from 5 V.
Consumption measured with the control side running.
Control: Control:
WiFi Configuration:
Updating FW:
You can either press and select the file, or drag the file onto the button.
Info:
Various helpful information.
Commercial use prohibited.
Comments
Nice one. ;) The case is PLA and the black jumper is a pen? [Read more]
PETG, higher softening temperature. Although the chip is unlikely to heat up. ??? aaaa WZorek, printed in black (on grey film) PETG on the first layer. [Read more]
Could you write a bit more - how did you manage this? 30 mW is usually what the ESP alone draws and without a radio, and you have a radio (quite responsive) and an inverter. [Read more]
Inaccurate meter. :) [Read more]
I initially tested on an ESP32-C3 (I was making a cooker controller on it so it was handy), C2 performed slightly better (can't remember the exact numbers). R10 I changed to a zener. The original manufacturer... [Read more]
Great respect! "The Chinese don't like it, see why". All in all, Zigbee also consumes 30 mW during transmission. There the standby is implemented a bit better and I think it closes in 1 mW ? But I guess... [Read more]
I was already waiting for you to post this. Congratulations, it came out great. Just one thing puzzles me - wouldn't it be possible to optimise the circuit even more? Are there bistable relays on 3 volts? I... [Read more]
Battery testing will happen, although I don't know when yet. Well baaa, 10 pcs instead of one. :) I suspect I would have to pay quite a bit more for the relay than the inverter cost. Other than... [Read more]
Seemingly a device like many (from majfriends), yet awe-inspiring, hats off to you for the beautiful, complex workmanship. :) I am impressed with the ease with which you use Chinese, not necessarily popular... [Read more]
The low price is tempting to try. I remember when I had to pay almost £10 for a similar 5 V-->3.3 V inverter at our place, and I could only dream about inductors in 0603/0805. Today at the Chinese ...... [Read more]
I tested the power supply from a LiIon cell. A 14500 cell plucked from a Chinese torch due to insufficient capacity. I charged it to full, plugged it in. After 26h, the voltage dropped below 3.3V and... [Read more]
It's cool that in the age of aliexpress and this someone wants to do such projects. [Read more]
Shouldn't diode D2 in the diagram be the other way around ? [Read more]
No. Look up the datasheet BP2525D for yourself. [Read more]
Hmmm, and why did you make the relay control so complicated? [Read more]
Bistable relay. It is energised only during a state change (for 0.01...0.1s). [Read more]
Aaaaa, ok. Mea culpa. [Read more]