I want to drive something like these ac synchronous motors:

Small Mige:


or Kollmorgen: (let's say 52k, doesn't mattter ultimately)


So is driving these possible with VESC? I downloaded vesc software and there is a FOC mode which uses sine waves. But this is still a BLDC drive, not BLAC (PMSM)?

Oh yea since i'm confused about driving BLDC and PMSM. Can you drive PMSM with trapezodial (theoretical square wave)? BLDC can be driven with ac so I'm curious, and there are no straight answers on the internet.

Edit: My plan is to use vesc torque control. The motor would not spin above, let's say 500 RPM. The torque is the most important to me.

Position feedback would be done with incremental encoder.

I will feed the data to vesc using another microcontroller. Haven't decided how to do it yet.

  • \$\begingroup\$ What is do you mean by VESC exactly? The only place I have ever heard the term ESC being used is in hobby radio-controlled vehicles because, as far as I can tell, there weren't any electrical engineers amongst them to tell them what they were actually called in the beginning, so it is by no means an engineering term with a specific definition. \$\endgroup\$
    – DKNguyen
    Jul 30, 2019 at 20:26
  • \$\begingroup\$ That is Vedder Electronic Speed Controller, or the Open Source ESC. Sorry for not mentioning anything more specific. Here are some links: vedder.se/2015/01/vesc-open-source-esc vesc-project.com \$\endgroup\$
    – Dwarf Yorf
    Jul 31, 2019 at 13:44
  • \$\begingroup\$ FOC for BLDC should work fine on a permanent magnet AC synchronous motor, at least in theory. I believe some synchronous motors use slip rings and have windings on the rotor. I don't think Vedder can control those. \$\endgroup\$
    – user57037
    Jul 31, 2019 at 16:12
  • \$\begingroup\$ As far as I know, those AC servo motors are not really different in any way from BLDC's. I have seen a servomotor assembly line, and the construction is virtually identical to a BLDC except that at the end of the line they install an optical shaft position encoder in a clean room. I know that theoretically the back EMF waveform is different, but honestly I haven't seen a BLDC with a trapezoidal back EMF. Maybe I just haven't looked at enough BLDC backe EMF waveforms yet. \$\endgroup\$
    – user57037
    Jul 31, 2019 at 16:17
  • 1
    \$\begingroup\$ @mkeith Thx for your help, now i can only hope that it will drive correctly :) \$\endgroup\$
    – Dwarf Yorf
    Aug 1, 2019 at 14:42

3 Answers 3


Both {BLDC, PSMS} have PM on rotor, but the difference is the distribution of stator windings. enter image description here

BLDC's are cheaper, better slightly more transient torque, when you do not need any precise position control unless you have a position encoder feedback then other specs determine which is better like EMI, THD, current crest factor, but generally any Sine wave drive gives better servo control for smooth torque, and position control such as using a low cost (<0.3$/W) VFD with some kV/RPM and computer control.

There are MANY types of position feedback encoders with different cost and resolutions. Cheapest and least accurate are Hall Sensors but adequate for commutation in most cases.

Both can be 2 phase quadrature or 3 phase Y ( or more phases ). ( You are considering 3 phase.)

Both tend to be 82% of no load speed at max rated power and 66% of no-load speed at max torque.

BLDC tend to be DC power with DC to Neutral or full bridge while PSMS are full bridge such as simple velocity+ acceleration controlled or AC input power.

Both use full-bridge PWM control. Feedback options include: current , and back EMF while BLDC tend to use HALL Sensors and PSMS tend to use more precise position feedback ( 3 types). ( Although I once use a micro-sized brush motor with 1000:1 gear reduction to turn a sin-cos Bourns pot for PLL vector rotation of a TEK XY vector display of an eddy current signal with 100 ppm resolution on impedance in XY planes at 100k,200kHz in late '70's) PSMS may offer active brake after coil power is off as a safety feature, while BLDC needs more active coil power for static torque.

Both can use 6 step Scalar control (quasi-sine) or FOC linear Vector true sine.

Both can have a variety of sensor feedback or used in sensorless control, depending on position accuracy and torque-phase jitter of control needed.

For high-performance Servo's with PSMS are best for resolution, the linearity of current and THD of power and thus lower eddy current losses.

Depending on cost, performance compromises, BLDC can be better such as using inexpensive ESC controllers.

It all depends on your error tolerance for every spec. {Position, velocity, acceleration, EMI, power distortion, dynamic control specs {step response, overshoot, load regulation)

Kollmorgenoffers best in class Servo solutions with at least 3 different servo feedback solutions, Sin/Cos resolver, RS485 digital incremental encoder with 20 bit resolution or


  • Trapezoidal back emf
  • Square wave current
  • Concentrating winding

More efficient , less servo position accuracy, more harmonic distortion


  • Sinusoidal back emf
  • Sinusoidal current
  • Distributed winding

If I made any errors in generalizations ( please suggest corrections) TYVM.

  • \$\begingroup\$ Thank you for the definitive answer that both methods work for both types of motors. As far as I could see and from my limited knowledge, everything seems correct. \$\endgroup\$
    – Dwarf Yorf
    Aug 1, 2019 at 14:45
  • \$\begingroup\$ What do you mean by "distributed winding?" You mean overlapping windings like in induction motor stators? \$\endgroup\$
    – user57037
    Aug 1, 2019 at 18:21
  • \$\begingroup\$ I visited a servomotor production line in China where they assemble servomotors for machine tools. The windings looked just like inrunner BLDC windings. No overlap. Not sure if this can be extrapolated to ALL AC servomotors. \$\endgroup\$
    – user57037
    Aug 1, 2019 at 18:22
  • 1
    \$\begingroup\$ No. THe winding distributions are different. PMSM are half sinusoidal to give sinusoidal BEMF. BLDC are not. researchgate.net/profile/Guillaume_Verez/post/… \$\endgroup\$ Aug 1, 2019 at 18:37
  • 1
    \$\begingroup\$ So what I am saying is that at least some motors specifically sold as AC servomotors do NOT have distributed (overlapping) windings. Perhaps we could say that those AC servomotors are not PMSM, but are actually BLDC. The second thing I am saying is that concentrated windings don't actually produce a trapezoidal back EMF. At least not in my experience so far. Hard drive spindle motors are much more sinusoidal, and the BLDC's I have seen in my job are pretty sinusoidal. Further, it is very common now to use sinusoidal drive for BLDC's instead of 6-step commutation. So I think there is gray area. \$\endgroup\$
    – user57037
    Aug 1, 2019 at 18:47

PMSM and BLDC are almost identical, except the one outputs sine wave back EMF while the other trapezoidal back EMF, respectively.

VESC is a more broad abbrevation for a moto speed control. A BLDC usally uses six-step commutation technique with the use of hall feedback sensors, or it can be sensorless with the help of back EMF sensing. At each time only two coils are energized, the third one is floating and used as antenna for BEMF sensing.

The FOC is used for control of PMSM, the injected current is sinewave, At the same time all three phases are energized. However with few tricks it is still possible to measure the BEMF if needed. A FOC needs at least sensing of two phases currents and absolute position of the rotor, last one can be estimated by use of some sensorless methods: BEMF sensing, PLL observer, Cordic, high frequency injection.

The FOC and PMSM is suitable for control speed, position and torque from zero speed up to max speed. While BLDC and VESC is usually limited to speed control which can be unstable at low speed if not using hall feedback, the same is also for the FOC - a sensorless control without an encoder feedback is not reliable at low speeds.

P.S. : the Kollmorgen motor from link is a PMSM with sinewave BEMF suitable for FOC. Some manufacturers use the incremental encoder, thus the absolute rotor position is unknown. For this purpose, the rotor initial position is estimated during startup an then counted with incremental encoder, which also has a zero mark (or many of them) used to set the true zero position.

  • \$\begingroup\$ I updated initial post. Now my main concern is will VESC be able to drive BLAC motors at all. I didn't come across anyone posting a successful project of the above mentioned. \$\endgroup\$
    – Dwarf Yorf
    Jul 31, 2019 at 14:00
  • \$\begingroup\$ Dwarf Yorf, most brushless AC motors run at higher voltages. Can the Vedder ESC operate with higher voltage DC link? For example Ali Baba link is to a 220V motor, and Kollmorgen is 320V max (so presumably at least a couple hundred volts...). \$\endgroup\$
    – user57037
    Aug 3, 2019 at 18:42
  • \$\begingroup\$ @mkeith mkeith, I'll need to clarify exactly how motors will be used. First : motors for haptic (force) feedback are needed. I'm building a steering wheel with only one axis for force feedback. DC motors would be ideal for this, but I avoided them for two reasons: low holding torque, and consequently more cost and building complexity. Second reason is that a lot of DC drivers don't support current monitoring, thus they feed too much current and power into the motor. That's why i chose PMSM. It will be direct drive, thus eliminating gears and pulleys. \$\endgroup\$
    – Dwarf Yorf
    Aug 5, 2019 at 12:11
  • \$\begingroup\$ @mkeith mkeith, Now an answer to your question. Motor will not need hundreds of volts. It will be "mostly" stationary, in the end it will be used as a steering wheel. From my knowledge you need 200-400 volts to extract the rated speed of PMSM. But people drove those motors with 48 V without any problems, hell i think that even 12 volts would suffice. I need to check internal resistance to know that. In the end holding torque is the MOST important factor in my project. Honestly, now when i think about it logically, adjusting current is the way to go. BEMF will happen only when wheel is turned. \$\endgroup\$
    – Dwarf Yorf
    Aug 5, 2019 at 12:17
  • \$\begingroup\$ Yes, you only need enough voltage to overcome the back emf at your maximum speed. Running with a constant current will probably work, but wastes power. I would think you would want a servo controller in position control mode for this. It will use feedback to keep the shaft position at the setpoint and only apply as much power as is needed to do the job. \$\endgroup\$
    – user57037
    Aug 5, 2019 at 17:21

As near as I can tell the lack of straight answers is because the terminology is not completely settled.

On the plus side, the worst that'll happen is that your motor will be inefficient, and run warm. If you don't start out by driving the thing at high speed under load, your learning experience should only be expensive in the sense of having unwanted parts, not slagged parts.

But, also as near as I can tell:

  • Permanent magnet synchronous machine (PMSM) is a general term for BLDC or BLAC. It's probably mostly used to be synonymous with BLAC.
  • Brushless DC is a PMSM with trapezoidal (or nearly so) back-EMF. So it's easier to drive.
  • Brushless AC is a PMSM with sinusoidal (or nearly so) back-EMF.

Nobody is going to have perfectly trapezoidal back-EMF. Nobody is going to have perfectly sinusoidal back-EMF. If you really want to know for sure, get a datasheet that shows the back-emf, or that gives specifications for it.

  • \$\begingroup\$ I updated my initial post. Now i remembered something interesting. The company Granite devices makes motor controllers ( Argon, Ioni ). These ones can drive BLAC (mentioned above), BLDC, DC, Steppers. But one interesting thing in their settings is that you put 3 phase AC or BLDC in the motor type. So does this mean that they are probably using FOC so they can support both blac and bldc? Cus people in skateboard community like FOC for bldc drive (using VESC). And one last question, I promise. Has anyone tried to drive BLAC with DC ( trapezodial method) ? \$\endgroup\$
    – Dwarf Yorf
    Jul 31, 2019 at 14:10
  • \$\begingroup\$ 6 step commutation (so-called trapezoidal method) will work to some extent with any AC motor. However, it is usually used in conjunction with Hall sensors that time the commutation for you. Efficiency and acoustic noise may be an issue. The bottom line with all of this is that these motors all want something approximating an AC waveform. You just need to manage the voltage amplitude of the waveform, the frequency must be exactly the same as back EMF and the phase angle between applied Voltage and back EMF must be controlled properly for what you are trying to do. \$\endgroup\$
    – user57037
    Aug 3, 2019 at 18:46
  • \$\begingroup\$ Any motor which has a high resolution shaft encoder should be driven with sinusoidal drive. I mean, why wouldn't you? The encoder adds expense to the motor. If you don't need that information, get a cheaper motor. Even if you don't use vector control (FOC, whatever) you might as well at least use sinusoidal drive. \$\endgroup\$
    – user57037
    Aug 3, 2019 at 18:48

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.