Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

  • State Resolved
  • Locked Locked
  • Replies 9 replies
  • Subscribers 83 subscribers
  • Views 25388 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

BLDC and PMSM. Let's point out the difference.

Lucifer Angel
Intellectual 995 points
Other Parts Discussed in Thread: CONTROLSUITE

... Just in order to clarify.

As far as I get it, in TI terminology BLDC motor has trapezoidal back-emf (when in generator-mode, so to say) while PMSM has sinusoidal one, that connected with stator windings distribution.

What else?

P.S.: Wonderful doc to look into: "Motor Control 1-day Workshop Presentation" from controlSUITE.

  • 0
    Prodigy 130 points

    i can`t agree with you more,  i add a point:  form the control methods, PMSM`s algorithm much more complicated than BLDC

  • 0 in reply to George Li
    Intellectual 995 points

    Yes, it is - algorithms are different. It is a consequence. Let's figure out reasons first.

  • 0 in reply to Lucifer Angel
    Genius 5885 points

    Hi all,

    I'm just adding some thoughts on the issue:

    The differences come from construction. BLDC's have concentrated stator windings, while PMSM's have distributed stator windings. The latter are more expensive for manufacture. Before anybody involved in machine design point out, I should say that with careful design of rotor and stator iron of BLDC you can come near sinusoidal BEMF. Ther reason why sinusoidal BEMF profile is desired will be discussed latter.

    If you look from energy/power transfer view, it is obvious that you have highest power factor if the phase current has the same form as phase voltage. Thus the different control strategy.

    With so called 120° commutation used in BLDC (there only two branches of the three phase inverter active) you can get current waveform that is similar to BEMF waveform. Additional advantages of this commutation strategy is that you only need one current sensor in order to control current/torque and this sensor can be simple shunt resistor on negative rail of the inverter. This reduces system costs significantly. As the 120°commutation only needs that you know rotor position with resolution of 60°electrical, you can use three hall effect sensors for position information, which again translates into low cost solution. Also from monitoring the phase voltage of unexcited phase you can derive rotor position information which you can use for sensorless operation. Some of the disadvantages of BLDC's are high cogging torque (air gap between rotor and stator is not constant), and high torque ripple resulting from simple commutation principle. This makes BLDC's hard to use in position control applications and results in higher noise. If the BEMF waveform does not have flat top you also get higher harmonics in torque response -> more noise. The simple control algorithm is a good thing in most aplications, but it does not support operation of BLDC in flux weakening mode (thus you can not achieve higher speed than that where phase to phase peak voltage equals DC-link voltage)

    With PMSM's you have sinusoidal BEMF thus you have to generate sinusoidal phase current. This requires the use of space vector modulation (which some might call complex, but with today's computing power it's piece of cake) and at least the well know field oriented control algorithm (again, some call this computationally demanding algorithm, but you basicaly need couple of multiplies and adds and two reads from sine/cosine table - not really state of the art by my measures). What you really need is two current sensors for monitoring phase voltage and an absolute position sensor with resolution better than 3°-4° electrical. This increases the cost of the system significantly. What you get in return is low noise, low cogging torque, low torque ripple drive, that is very suitable for position control. Also the control algorithm can be easily adapted for operation in flux weakening region.

     

    Best regards, Mitja

  • 0 in reply to Mitja Nemec
    Prodigy 120 points

    Hi to all

    Mr Mitja you have give a good explatation!

    and you look to know very well about motors so i have some questions for you if is not problem

    1) for edication porpuse can you teel to us any side with picture to see the diference between BLDC and PMSM windings

    2) when i have a BLDC motor with a encoder 2000 resolution is less accuracy (position control use) from a PMSM with the same encoder?

    3) what is better to use for position control a BLDC or brush DC?

    4) what is the difference between space vector modulation and field oriented control?

    5) i have a BLDC motor with 8 pole magnet rotor and use a magnet disk (8 pole) with 3 hall sensor in 60 mechanical degrees(as i see)

    between them a total arc 120 degrees.the manual say the hall is in 120 electrical deg,

    i do the simple convert the mechanical to electrical deg: 360el-deg/4=90 

    that mean 360 ele-deg of stator is 90 mechanical deg of rotor so that mean:
    with 30 mech-deg i have 120 el-deg, so the hall must to be in 30 mech-deg configuration
    where i am wrong ?because the motor work fine with my controller (is accept only hall in 120 configuration)

    thank you very much!

    best regards Yiannis

  • 0 in reply to YIANNIS PANAIS
    Genius 5885 points

    Hi Yiannis

    While I am in the teaching business I really can not go deep into details so if you need any deeper knowledge on the topic I would suggest you find a local expert an hire him.

    Anyhow:

    1. BLDC windings () are concentrated while PMSM has distributed windings. Typically.

    2. The accuracy is the same. You might have problems with BLDC if you are implementing position control as torque response around commutation angle is not smooth.

    3. Hard to say. From my point of view it comes down to torque ripple and cogging torque. And if you also need high RPM, you will have low pole pair count BLDC. Thus it will have six or twelve commutations per mechanical revolution. Brushed DC can have more, thus torque ripple can be smaller, but I can not guarantee. And brushed DC will probably have lower cogging torque.

    4. If you have to ask this you do have a lot to learn. For starters I recommend this video series

    5. The hall sensors are typically 120° electrically apart (0, 120, 240). They can also be at i*360, (120+j*360), (240+k*360) electrically. i,j and k are integer numbers. Think about it.

    Regards, Mitja

  • 0 in reply to Mitja Nemec
    Prodigy 120 points

    Hi Mr Mitja

    Thank you very much for fast response!

    You give to me very good answers! but i can not to understand the last one, can you give a example with my motor case? 

    thank you very much!.

    Best Regards Yiannis.

  • 0 in reply to YIANNIS PANAIS
    Prodigy 120 points

    Hi Mr MItja

    one last question

    I would like to ask if a PMSM or AC SERVO motor need to have hall sensor, in application with low speed and control position(like cnc).

    I ask that because i see in market some AC SERVO with out hall sensor just encoder, and i reed in some application like microchip have, in very low speed the back-EMF is not enough to estimate the rotor position, if i have understand enough! 

    thank you very much!

    best regards Yiannis

  • 0 in reply to YIANNIS PANAIS
    Genius 5885 points

    Yiannis,

    Regarding my last point I would point you towards local electrical drives expert as this is much easier explained with pen and a piece of paper.

    As it goes for encoder and hall sensors, my guess is this:

    For any kind of AC machine (PMSM, induction-ACI or BLDC) the inverter (or more specifically control electronics) needs to know rotor position or at least speed. If you don't have any kind of sensor, the information can be obtained through voltage and current measurement, but usually this does not work at low speed. This is true either for torque control, speed control or position control.

    Now if you only want to control torque you don't need any kind of sensor additionally.

    For speed control you need te get the information about actual speed somehow. You can get it indirectly through current and voltage measurement, but usually this is not really precise and does not work at low speed (good example of speed control application that does not require direct speed measurement is HVAC-heating, ventilation, and air conditioning. Air flow is practically negligible at low speed so you don't operate in that range).

    If you have position control application, you can be almost certain that you will operate at zero speed. So you will require position measurement. You will need position measurement for position control and also for drive control (in case of BLDC for commutation). For drive control you do not require high precision (60°electrical is enough for BLDC, while for PMSM or ACI cca 3°electrical is good enough). It is the requirements of position control that dictate position sensor that you need. If it is better than what is required for drive control, you could also use it for drive control. Thus some position control solutions have only one position sensor, while other (like the one mentioned) have two (one for drive control and one for position control). Obviously first solution is cheaper but second solution is modular.

    Regards, Mitja