I have a ESP32-SOLO-1 microcontroller mounted on a moving part of a machine which moves in a circle. The idea is to power it with a capacitor and then charge the capacitor in a fixed location when the MCU/capacitor part comes around (for example via some sort of sliding contacts connected to a power brick.)

The main problem is the charging of the capacitor, as it has in the current configuration around 0.14 s to charge and it has to last about 8 s with a load of 0.04 A at 3.3V until it charges again (except the first time it goes around the MCU has to boot and it requires 0.1 A for about 1 s.)

I used so far a 0.47 F supercap with a rating of 5.5 V (here the datasheet.) I already had with a MCP1700-3302E-TO voltage regulator to have a fixed 3.3 V output. The cap has an ESR of about 16 ohm which in my understanding is far too much for charghing it in such a short time.

My idea was to use this capacitor --> 0.033F 25V and a very low ESR (0.027 ohm in the datasheet) charged with a 24V power brick and power the MCU with a DC-DC converter (this) that outputs 3.3 V.

What are your thoughts on this? Am I completly wrong or could it work? Do you have any suggestions on how to make it work?

EDIT: The MCU evaluates and sends data from a Hall sensor, which is used to mesure the distance of a magnet. So I would rather not have electromagnetic interference and therefore use "touching" parts. Also the goal is to avoid if possibly using a battery.

  • \$\begingroup\$ What about a slip ring? en.wikipedia.org/wiki/…. \$\endgroup\$
    – Parker
    Jan 6, 2021 at 15:22
  • \$\begingroup\$ The MCU unit moves in a circle but not on itself. As if it was mounted on the edge of a big wheel that spins and the MCU goes by the charging point only one time per revolution of the wheel \$\endgroup\$ Jan 6, 2021 at 15:29
  • \$\begingroup\$ Can you use a transforming action or induction. You can't mount brushes around the wheel and wire them back? (which is what a slipring would be). Also consider a battery. \$\endgroup\$
    – DKNguyen
    Jan 6, 2021 at 15:39
  • \$\begingroup\$ The MCU evaluates and sends data from a Hall Sensor, which is used to mesure the distance of a magnet. So I would rather not have electomagnetical interference and therefore use "touching" parts. \$\endgroup\$ Jan 6, 2021 at 15:47
  • 2
    \$\begingroup\$ Depending how long contact is made for, you may want significantly larger caps (>5x) on the fixed brushed which are charged up and ready so that you can immediately discharge into the super cap on your moving brush without having their own voltage dragged down by the cap on the moving brush. Otherwise your stationary supply may not react quickly enough and get bogged down, but not so big that your stationary supply can't handle all the output capacitance (or you need to have a series resistor so the stationary supply gracefully charge up and top off the caps on the fixed brush). \$\endgroup\$
    – DKNguyen
    Jan 6, 2021 at 16:46

1 Answer 1


Sounds plausible.

Maybe you could use a Qi transmitter/receiver to produce 5V and use a simple LDO regulator down to 3.3V. I'm not sure they would switch on fast enough and give enough current for long enough to work. You would need (say) 1A for less than 500msec to keep an average 40mA going, which sounds to me like it might be worth investigation.

Regardless of how you get 5V to the target, a super-capacitor such as the AVXSCMT32F755SRBA0 7.5F 5.5V with 90m\$\Omega\$ ESR might do the trick (charged to 5V and with an LDO regulator) with room to spare. If you use brushes, maybe you can inhibit communication during the noisy bits and add a current limiter to the 5V supply to keep the peak current tame.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.