I'm creating a motor driver for a BLDC motor. I was wondering how we can calculate or choose what capacitance we need for our purposes.

We're using the motor driver for a chain drive /transmission system. We pull around 20A and supplying our driver with 36V. Im not sure what other info is needed, but thanks!

  • \$\begingroup\$ Just go with 220uF or 470uF and start from there. It's doesn't need to be an exact thing. \$\endgroup\$ – DKNguyen May 7 '19 at 20:03
  • \$\begingroup\$ We went with a 100uF before and that exploded, so we were wondering if there's a good value. We've seen other people/universities use up to ~3000uF for the same purposes \$\endgroup\$ – Huy Nguyen May 7 '19 at 20:17
  • 1
    \$\begingroup\$ That kind of sounds like there are design factors you're overlooking such as transient suppression, snubbers, algorithms, overly long cables, bad layouts, excessively fast switching times, or any other number of things because BLDC motor drives for model aircraft of similar ratings use capacitances that can be reasonably obtained with just a couple of capacitors. You might actually need that much capacitance, but you could probably reduce it if you look at other aspects of your design more carefully. \$\endgroup\$ – DKNguyen May 7 '19 at 20:31
  • \$\begingroup\$ When did this 100uF explode? At start up? After some time the motor was running? When turning off the motor? What voltag rating did the 100 uF have? \$\endgroup\$ – Huisman May 7 '19 at 22:40
  • \$\begingroup\$ The capacitor exploded after running the motor unloaded (100% duty cycle) for an extended duration (didn't time, but let's say a minute). We currently have a 1000uF cap in parallel with a 3300uF (unfortunately the 3300uf cap is connected through wires since we didn't design the PCB we currently have for a large cap). They haven't exploded or anything as of yet with loaded or unloaded test runs. \$\endgroup\$ – Huy Nguyen May 7 '19 at 23:08

The bulk capacitor makes sure the supply voltage and/or motor voltage doesn't drop too much in case of current spikes/surges. The capacitors will not contribute for the DC current drawn by the motors.

You should find out what the inrush current / surge current is. For DC motors the inrush current equals the stall current.

Decide what voltage drop you want and approach the capacitor size with $$ C = \frac{ i_{inrush} \cdot t_\text{duration on inrush current} }{ V_{drop} }$$.


Low ESR caps are rated for high Ripple current in Amps.

Batteries are far greater capacitance and often very low ESR such as a single 18650 cell ≈ 10kF and 50 mohms.

The ripple current you choose depends on the ESR and motor surge current rise time. Choose parts with low dissipation factor.

My rule of thumb is choose Cap Array & Battery and Bridge ESR's much lower than your motor DCR to minimize losses such that it spans a wide spectrum of frequencies generated by the motor pulse load.

This cost of Caps rises with capacitance for values of τ = ESR * C << 1us but they do make them for WPT and large motor supplies.

  • \$\begingroup\$ how do you determine the ripple currents produced by the motor commutations --or do you have to measure these? \$\endgroup\$ – jrive Mar 31 '20 at 23:33
  • \$\begingroup\$ You can model it but depends on the number of phases greatly. \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Apr 1 '20 at 4:44
  • \$\begingroup\$ can you give me an example, say 3 phases? how would I model it? \$\endgroup\$ – jrive Apr 1 '20 at 19:39
  • \$\begingroup\$ Try searching for the answer \$\endgroup\$ – Tony Stewart Sunnyskyguy EE75 Apr 2 '20 at 0:39
  • \$\begingroup\$ I have, posting here is what I did after not being to find a satisfactory answer.....it's ok if you don't know.... \$\endgroup\$ – jrive Apr 3 '20 at 3:40

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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