• State TI Thinks Resolved
  • Locked Locked
  • Replies 12 replies
  • Answers 3 answers
  • Subscribers 33 subscribers
  • Views 3847 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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 rotor position detection at standstill based on motor inductances

Bas Putter
Prodigy 45 points
Other Parts Discussed in Thread: DRV8301-69M-KIT, INSTASPIN-BLDC

Just wanted to share my success building a system for sensorless detection of the rotor position of a bldc / pmsm motor. (I hope this is allowed). This with the rotor standing still and from power-up, so with no history present in the microcontroller.

I get the same 0-359 degrees electrical rotor position info as you would with a resolver. While the rotor position measurement is running it is possible to apply phase current for generating torque.

I posted two videos on youtube, channel bmp72 . In both videos the motor is only connected via the 3 phase wires....

  • 0
    TI__Guru 71526 points

    Hello Bas,

    Wondering if you could share what TI components you used in this design and the reason for choosing them?  Perhaps even a block diagram and description so others could benefit from your design expertise. And advantages over this method of position detection over other methods.  

    Thanks,

    ~Leonard  

  • 0 in reply to LeonardEllis
    Prodigy 45 points

    Hi Leonard,

    Thank you for your reply. There is not much TI content, it is more that I would like to know how this compares to the Instaspin. I've heard on Endless-sphere that Instaspin can spin a motor from standstill without using hall sensors or a resolver ? I have tried to build something similar (starting from scratch):

    Basically what I built is a processor feeding into gate drivers feeding into a classic output stage. From there the power signals go through current sensors into the motor. The signals from the current sensors go to the processor in two ways, straight and via some analog processing. There is no motor voltage feedback.

    With the above schematic (and the appropriate code) the system can detect the electrical rotor phase of the BLDC. The first video (see previous post) shows that the system can generate similar output as a resolver, 0-359 degree info (and not just 6-stepped). It does this by measuring the motor inductances. The second video shows that it can do this while also supplying phase current (to generate torque).

    The advantages ? No need for a resolver or hall sensors to detect the rotor phase, but still capable to provide torque at standstill and low rpm (even from a cold start, so powered on from completely off with no memory about any previous rotor movements)...

    Just showing off I suppose, sorry if this is not the right place.

  • 0 in reply to Bas Putter
    Prodigy 195 points

    This is exactly the performance we are looking to achieve for our application. We are evaluating InstaSPIN using the DRV8301-69M-KIT and are having some trouble figuring out how to do this with a sensorless system. Is zero-speed, sensorless position control achievable with InstaSPIN?

  • 0 in reply to Elliott Partridge
    Prodigy 45 points
    I am not familiar with TI's Instaspin at standstill and from a cold start, this here is purely my own private development.

    On youtube video szgVUfyX8JM has information about TI instaspin, I hope I am allowed to copy TI's own comment here:
    " InstaSPIN-BLDC can start up from zero speed with nearly full torque, and it can transition through zero speed, but it is not a zero speed solution. InstaSPIN-FOC and the FAST estimator provides stability at zero speed, full load start-up, and <1 Hz angle lock and continuous low speed operation."

    So based on this comment, have you looked at InstaSPIN-FOC and the FAST estimator ?
  • 0 in reply to Bas Putter
    Prodigy 195 points

    @Bas - I am using the InstaSPIN-FOC and -MOTION labs.  It appears that position control is only demonstrated for sensored solutions. I am wondering if anyone at TI supports zero-speed position tracking & control using InstaSPIN-FOC or -MOTION.

    I just found lab21, which looks like it might address low/zero speed operation better than the only using the FAST controller.  However, I'm still unsure as to whether position control is going to be feasible with a sensorless system.

  • 0 in reply to LeonardEllis
    Prodigy 195 points
    Leonard,
    Can you comment on whether what Bas has demonstrated is possible with InstaSPIN-MOTION, with the additional requirement of position control/tracking at zero speed? Can HFI be integrated with InstaSPIN-MOTION to improve performance?
    Thanks,
    Elliott
  • 0 in reply to Elliott Partridge
    TI__Guru 71526 points

    Hi Elliott,

     I cannot comment, I don't have InstaSpin expertise, but I have alerted someone in that Product Group to join the conversation.  Are you designing a product and want to base it on this technology?  

    Regards,

    ~Leonard

  • 0 in reply to LeonardEllis
    Prodigy 195 points
    Leonard,

    Thanks for forwarding the question. Yes, I am evaluating InstaSPIN for a product redesign. The previous product uses hall sensors for position feedback, but we need to move towards a sensorless solution.

    I can start a new thread if appropriate.

    Thanks,
    Elliott
  • 0 in reply to Elliott Partridge
    TI__Guru 71526 points

    Hi Elliott,

     Yes, I think starting your design in a new thread would be well-advised.  Thank you!

    ~Leonard 

  • 0 in reply to LeonardEllis
    Prodigy 195 points
    I've started a new thread, here:
    e2e.ti.com/.../712987
  • 0 in reply to Bas Putter
    Prodigy 10 points
    Hallo Bas,
    I am really impressed with your development!
    I have seen in your videos that you use only outrunners. Do these motors have this periodical variation of reluctance like the inrunners with buried magnets?
    What kind of commutation do you apply? Is it possible with six-step commutation?

    In advance, thank you alot for your answer and of course, for sharing with us your achievements.

    Kind regards,
    Stratos
  • 0 in reply to Efstratios Petrou71
    Prodigy 45 points

    Hi Stratos,

    Thank you very much for your kind words. Yes, the motors I use have the periodical variation in reluctance.

    Unfortunately, as the whole algorithm is private and unprotected I cannot say anything about it...

    Best regards,

    Bas