• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Subscribers 28 subscribers
  • Views 5819 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.

Back EMF or Hall sensor.. which would be the best

Dharmendra Sharma
Intellectual 735 points
Other Parts Discussed in Thread: DRV10983, MOTORWARE, CONTROLSUITE, DRV11873, DRV8313EVM

Hi All,

I have 3 hall sensors and I aware how to use them to drive a 3 phase BLDC motor using microcontroller. But I wantt to know is there another method named as "back Emf method" in which there is no need to use the hall sensor for sensing the mechanical angle of the motor.

Is it better method to motor drive then the hall sensor?

Do we gonna achieve same RPM and Torque using the back em method? Please suggest.

Is the back-emf based method uses the 120 degree commutation algorithm for brushless DC (BLDC) motor.

 

Regards,

Dharmendra Sharma

INDIA

  • 0
    TI__Mastermind 26065 points

    Hi Sharma

    Sensorless method can save your Hall sensors and it's related installation cost. It's disadvantage is that it has poor low speed or stall condition torque stability. So the sensorless method is normally adapted for fan, pump applications which does not have big torque variety at startup and normal running. Also sensorless method could produce little reverse running at startup.

    We could not simplely said which is better. Normally both sensored or sensorless can get your motor to the same RPM and Torque at similar speed (With the same leading angle).  

    BEMF detect method is one of the most commonly used sensorless control method. But we do have other algrithm such as the BEMF integration method in TI instaSPIN, or the current detect and calculation method in DRV10983. Normally 120 degree driving is used in sensorless control but we can find DRV10983 is 180 degree control.

    You can refer to DRV11873, DRV10983 for the integrade solution. Also you can find lots of algrithm resource in TI C2000 controlSUITE or motorware software.

    Best regards,

  • 0 in reply to Wilson Zuo
    Intellectual 735 points

    Zuo,

    Thanks for your quick and valuable reply. 

    1. I am making this 3 phase phase motor driver solution (for CD-ROM motor) ,to learn only that how it work. So i am not going to use any ready made integrated solutions or any BLDC motor driver IC.

    2. I love to make things from raw materials so I want to use the 6 MOSFET configuration (as seen below ) to drive the motor . For commutation I am gonna use any suitable MSP430.

    3. The doubts are 

     i) The feedback which I am getting from hall sensor about the position of motor, is be the same for back emf              method. 

     iv) what would be the range of back- emf if we use maximum +12v for CD-ROM BLDC MOTOR.

    Regards,

    Dharmendra Sharma

    INDIA

  • 0 in reply to Wilson Zuo
    Intellectual 735 points

    Hi Zuo,

    I also need some things to be clear.

    1. do we need to use ADC for measurement of the back emf or any compare-tors or we can usen simple I/O.

    2. if I am right then we gonna measure the back emf between the non excited coil and GND.

    Regards,

    Dharmendra Sharma

    INDIA

  • 0 in reply to Wilson Zuo
    Intellectual 735 points

    Zuo,

     

    and the last one (I forgot to mention in previous post)

    Do we need any type of starter algorithm to make the BLDC motor start.

    Regards,

    Dharmendra Shrama

    INDIA

  • 0 in reply to Dharmendra Sharma
    TI__Guru 54675 points

    Dharmendra,

    Can you please tell me a little bit about the volume of your product and production schedule? 

    If you are looking to learn only with no actual product, it is best to go to the following site where we provide training videos and you can also purchase a launchpad + boosterpack to experiment. 

    www.ti.com/motors

    I would recommend the following boosterpack:

    http://www.ti.com/tool/TIDM-LPBP-BLDCMOTORDRIVE?keyMatch=drv8301%20boosterpack%20bundle&tisearch=Search-EN

    Thank you!

     

  • 0 in reply to Ryan Kehr
    Intellectual 735 points

    Ryan, 

    I am developing my own quad-copter. So I need to drive 4 BLDC motors . I am using the CD_ROM BLDC motor (upgraded--- since I had changed the magnets and winding to achieve high torque and RPM ). 

    Now can you suggest me which method would be the best??

    Regards,

    Dharmendra Sharma

    INDIA

  • 0 in reply to Ryan Kehr
    Intellectual 735 points

    Ryan ,

    and thanks for your recommendation of booster-pack but I can't use them due to weight and size limitations.

     Regards,

    Dharmendra Sharma

    INDIA

  • 0 in reply to Dharmendra Sharma
    TI__Guru 54675 points

    Dharmendra,

    I think Wilson provided a good list of the tradeoffs between halls and sensorless. 

    http://www.ti.com/tool/drv8313evm

    Please check out the DRV8313EVM.  It uses an MSP430 for InstaSPIN BLDC sensorless control.  I recommend you investigate this.  Information below.  We also have a nice 45 minute video to walk you through the algorithm and hardware requirements.  Please find it on www.ti.com/motors.

    http://www.ti.com/ww/en/mcu/instaspin/instaspin-bldc.shtml?keyMatch=instaspin%20bldc&tisearch=Search-EN

    Thank you!

     

  • 0 in reply to Ryan Kehr
    Intellectual 735 points

    Hey Ryan,

    The motor driver is not a problem at all. I will develop that one using Some mosfet's and a microcontroller. I learnt that how to drive a bldc motor with hall sensor. BUT after I found another way of driving "BACK EMF METHOD" , which is cost effective one . I read about it but have some doubts.

     

    As I mentioned above my doubts are,

    1. The feedback which I am getting from hall sensor about the position of motor, is be the same for back emf              method.

    2. what would be the range of back- emf if we use maximum +12v (the voltage which I am using to drive the motor) for CD-ROM BLDC MOTOR.

    3. do we need to use ADC for measurement of the back emf or any compare-tors or we can usen simple I/O.

    4.  if I am right then we gonna measure the back emf between the non excited coil and GND.

    5. There is a minimum speed that must be attained for sufficient BEMF to be measured.

     

    Regards,

    Dharmendra Sharma

    INDIA

  • 0 in reply to Dharmendra Sharma
    TI__Mastermind 26065 points

    Hi Sharma

    Sensorless method is much more trick technology than Hall sensored BLDC control method. There is no detailed and standard implementation for this sensorless method. Merits and faults are always come together. And most good experiense of sensorless control are protected IP in different companies. If you want to learn, start from TI instaSPIN should be a good choice. And also you can "google" lots of related materials.

    Some comments to your questions:

    1. The signals feedback from Hall sensor and BEMF are totally different and need different dealing as well.

    2. BEMF is proportional to the motor's speed. So at startup or low speed, you will get not enough BEMF to do the commutation control. This is the most tricky point of sensorless control. To your question, normally the motor's BEMF range will below VM in the normal running speed range. You can understand this through the basic model of the DC motor. Please google the "DC motor model".

    3. Comparators are used very common. But if you need complex filter or integration of BEMF, ADC will be needed such as in TI InstaSPIN. GPIO may not be enough.

    4. Usually it is compared between the non excited phase node and the center tap. (It is easy to understand in star connection while could be artificial one in delta winding motor). If sampled by ADC, it is measured refer to GND.

    5. Yes. Before you can get enough BEMF, you don't know the rotor's position and just control in open loop or apply some observer. 

    I think you will get so many questions during your development, it is recommend you choose one method first and get a deep diving. Also I have some doubt on your modified CD-ROM motor, it may not able to output enough power to lift your aircraft. There is too much different between a CD-ROM motor and a high power density aircraft motor. It is really critical to your success fly. Driver or driver method doesn't help a lot about this.

    Best regards,

  • 0 in reply to Wilson Zuo
    Intellectual 735 points

    Wilson,

    Lots of thanks for useful suggestions. Now I got It.  

    I have to use tricks at the start-up of the motor using commutation with PWM to generate sufficient BEMF.

    I did lot of research on CD-ROM DC motor in last few days. Before choosing CD-ROM  motor I was also thinking about the success of quad-copter since it have less torque because the Kv (RPM per volt factor) of a standard CD-ROM motor is usually too low and the prop might not turn fast enough to fly a model satisfactorily. That means that there are too many turns (usually 30+) of too thin wire on the stator. Thus I made rewinding with less turns of thicker wire will result in a higher Kv and therefore the motor will turn faster while drawing more amps thus more torque. The Kv of a motor is inversely proportional to the number of turns on the stator. 

    And also In most cases (if not all), a CD-ROM motor will perform better if you install Neodymium magnets. So I had Installed N50 Neodymium magnets replacing the fer-rite magnets therefore not only will the motor have more power, but it will also run cooler. I have read that starting problems also go away after installing magnets.

    And the motor weights only 25g and the whole flying QUADCOPTER (4 motor). So the whole structure will have Enough torque/weight ratio to fly.

    Thanks Again

    Dharmendra Sharma

    INDIA