I'm just wanting to confirm that this circuit would be safe/reliable to implement. I'm hoping to run two servos - one 6V one 7.4V - in parallel with an ESP32 dev board with built-in linear regulator, from an already regulated 7.4V source. Each regulator is a store-bought module with all surrounding circuitry.

The first switching regulator is pretty tolerant to current draw so I'm not too concerned about voltage sag under load (also servos using maximum stall current will be unlikely). Am also not too worried about voltage ripple to each servo from the switching regs. I've heard that sharing current can be an issue between linear voltage regs in parallel, so am hoping a switching reg doesn't have the same problem and have used it in place.

I'm pretty sure I should also have Schottkey diodes on each separate voltage line but wanted to confirm if any other problems may arise. Appreciate any feedback, thank you.


  • \$\begingroup\$ I hope that C1 is not the only capacitor in this circuit. \$\endgroup\$
    – Reinderien
    May 5, 2021 at 15:27
  • \$\begingroup\$ Why chain the second and first switching regulators? The first regulator will need to take the current burden of both loads. Can you not connect the second regulator directly to the LiPo? \$\endgroup\$
    – Reinderien
    May 5, 2021 at 15:30
  • \$\begingroup\$ I had 100nF for noise and 10uF for dropout across the ESP32, but 7.2V is so much higher than the 3.3V required that a brownout seemed unlikely, suppose it wouldn't hurt to add it back; and yes I could connect directly to lipo, if switching regs don't have the same issue as parallel linear regs \$\endgroup\$
    – trijoco
    May 5, 2021 at 15:52
  • \$\begingroup\$ Without capacitors in their input and output voltage regulators are unstable and may oscillate. \$\endgroup\$
    – Miss Mulan
    May 5, 2021 at 22:42
  • 1
    \$\begingroup\$ Sorry I should clarify that the switching voltage regulators are on pre-fabbed boards with 220uF caps across input and output, + inductor, diode, resistors, etc \$\endgroup\$
    – trijoco
    May 5, 2021 at 22:47

1 Answer 1


I would run the 6V reg directly from the 11.1 volt supply too, then run the LDO from 6V instead. This would isolate the higher power servo from the others and also reduce the LDO losses.

You might even consider using a switcher to make 3.3V direct from the battery too. TI and others make triple DCDC devices that would meet your spec and reduce your chip count.

Your power sequence depends more on keeping the servos quiet until the ESP32 comes up, so you should design a reliable ‘mute’ signal to disable them.

  • \$\begingroup\$ Thanks I appreciate the response, will look into those triple DCDC devices. Do you think I need any additional diodes or capacitance? Am thinking the servos might have their own flybacks for protection and that they won't be too sensitive to voltage fluctuation so don't need decoupling caps but am not sure. \$\endgroup\$
    – trijoco
    May 5, 2021 at 21:49
  • \$\begingroup\$ Decoupling is recommended - the dI/dt for servos can be quite high. As far as flyback, that should be part of the drive circuit already. \$\endgroup\$ May 5, 2021 at 23:37
  • \$\begingroup\$ Hi, have been looking around for recommended values of decoupling capacitors for servos, are there any general values to abide by? Have seen 470uF thrown around a bit but can't find any definitive answers on servo inductance and how long it takes for the inductive field to collapse etc. Thanks again for your time. \$\endgroup\$
    – trijoco
    May 8, 2021 at 7:58

Your Answer

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

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