I have an 3 phase brushless dc motor with no neutral wire. I measured the resistance from phase to phase and found them to be on average 0.255 ohms(between A and B, A and C, etc...all the values are pretty close).

These 3 phases are in a star formation. You can see the motor here: http://www.cda-intercorp.com/Products.cfm?cid=1&Data=Design#ScrollTop

I am running the motor on 28V. How do I know what current are running through each phase(A,B,C). To add to this, if I stall the motor, only two of the phases would have current right? What would the current be then?

  • \$\begingroup\$ Normally, the motor drive circuitry performs pulse width modulation, meaning that the current will not just be 28 V / 0.255 Ohms. What will be the stalled rotor current? It depends on the drive circuitry. The highest it could possibly be would be 28V/0.255 Ohms (Assuming you did not make a mistake measuring winding resistance). \$\endgroup\$ – mkeith May 4 '16 at 3:20
  • \$\begingroup\$ How did you measure the winding resistance? \$\endgroup\$ – mkeith May 4 '16 at 3:20
  • \$\begingroup\$ @mkeith it was measured previously by someone else so i'm not sure. But the motor specs show similar numbers so those are correct. At the highest current 28/0.255ohms, would the current just be split evenly between two of the phases(55A)? Still the number is so much higher than even the peak momentary current limit(7A). This doesn't make too much sense to me.. \$\endgroup\$ – mugetsu May 4 '16 at 4:25
  • \$\begingroup\$ to add to that, does this mean that it will just draw increasing amounts of current until the motor hits the stall protection limit? \$\endgroup\$ – mugetsu May 4 '16 at 4:34
  • \$\begingroup\$ It is just Ohm's law (V=IR). But the drive circuitry would not normally apply max voltage continuously. It is just going to depend on what the drive circuitry does. Three phase brush-less motors cannot be driven directly from DC, even though they are called "DC." \$\endgroup\$ – mkeith May 4 '16 at 5:56

The absolute best way to measure the current in a line is to use a current probe such as the TCP0030A. They are a bit pricey, but totally worth it if you are doing motor control.

Of course, if you are doing this in an application, you can use three ACS711 (pdf warning) chips (or similar). These are good enough to use for current control loops or similar, but not necessarily good enough for lab-grade measurements.

I have used both successfully.

What happens in a stall condition depends entirely on the drive architecture. Most drives have some sort of stall protection built-in so that they don't burn up the motor.


First, you said it was a star wound motor with no neutral.

That means the 0.26 ohm measurement you made is already across two (0.13 ohm) windings in series.

That also means the stall current at 28V is somewhere over 100A.

As you suggest in a comment that you have a peak current limit of 7A (what imposes this limit???) that means your motor driver must soft-start the motor, i.e. start it at a much lower voltage (or with PWM at a much lower duty cycle (about 7%).

It's not clear which motor from that page you are using, but the typical "performance" data on that page is unusual - it shows a "stall torque" that is only slightly higher than the torque at max power - and a much lower power dissipation at stall. That confirms that the permissible drive voltage must be reduced by the motor controller at stall, and probably also at startup.

Either their recommended controller will do this for you, or you'll have to actively measure the current and control the PWM duty cycle to stay within the motor's limits.


if I stall the motor, only two of the phases would have current right? What would the current be then?

Probably not: -

enter image description here

The current waveforms will be phase shifted in each winding but there are only transient ponts when the current in one winding is zero.

How do I know what current are running through each phase(A,B,C).

You can't know without measuring it unless your drive circuit has some feature to tell you.

  • \$\begingroup\$ cough those are BLAC & all 3 will have current depending on rotor position and the controller. BLDC, as per the question will only be two phases UNLESS a boundary condition and it jumps between commutation point, even then current only in two phases \$\endgroup\$ – JonRB May 4 '16 at 14:08
  • 1
    \$\begingroup\$ @JonRB I hear what you say and the OP says DC but it is a permanent magnet rotor and he has access to the stator windings and this makes it an AC motor. maybe I'm being stooooopid? \$\endgroup\$ – Andy aka May 4 '16 at 14:42
  • \$\begingroup\$ Well all motors require an AC flux (even brushed DC ;) ) He mentioned he is "running the motor from 28V" which would imply some form of ESC & ASSUMING it is a BLDC ESC (and not an AC controller... the motor wouldn't care ... I drive "BLDC optimised machines from a DQ SVM controller all the time), only two phases will be sourced. Question is a bit ambiguous in all fairness \$\endgroup\$ – JonRB May 4 '16 at 14:49
  • \$\begingroup\$ There is nothing special about the terms BLAC or BLDC ... the only difference is the shape of the back-emf. Either one can be driven by a 6 step drive or a sinusoidal drive and unless we know what control he has, it is impossible to answer that part of the question. \$\endgroup\$ – Eric May 4 '16 at 15:59

If you want to know the current in the phases of a BLDC machine you really need a current sensor.

Since ONLY two phases are energised at any one time you could infer via having a single DC-side (pre inverter, post capacitor) current sensor. you can then infer to a reasonable degree of accuracy what current is in what phase via the present commutation state (NOTE: zone change will result in some inaccuracy as the current commutates from one phase to another).

If the motor is stall however... depending on the complexity of the ECS & whether there is a maximum PWM duty you could infer via two series phase resistance and the percentage of the DClink applied via PWM.


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.