Shop OBEX P1 Docs P2 Docs Learn Events
Neat BLDCs with controller and encoder built in — Parallax Forums

Neat BLDCs with controller and encoder built in

Mark_TMark_T Posts: 1,981
edited 2015-03-22 09:07 in Robotics
I saw these on eBay a while back and bought one:
http://www.ebay.co.uk/itm/DC-24V-20W-High-grade-brushless-motor-brushless-motor-with-encoder-PWM-speed-/261713199819

Basically a small BLDC with built in controller and 100 line optical encoder, very easy to interface to, just needs 24V supply and 5V for the encoder.

I had a play with turning it into a servomotor (on Arduino since its 5V, I think level shifting would be needed for the Prop). It takes
PWM, brake and direction inputs. The encoder is totem pole output and claims to need 5V (haven't actually tested with 3.3V though).

About 6400 rpm at 24V no load, very smooth and quiet, handy for many projects I would think. All it needs is a selection of cheap reduction gear units! Impressive as a servomotor, too fast to see the thing rotate!

Here's it filmed at 480 frames/s:
https://www.youtube.com/watch?v=4xQ7u5cleAo

Comments

  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-03-13 21:14
    Any idea of the torque?

    If you find a good gearbox for it, I hope you let us know.

    The ebay page says "about 100" lines for he optical encoder. Do you know if this is exact?

    I hope you share any additional information you learn about these.

    Thanks for the info so far.
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-14 08:37
    100 lines, 400 counts/rev... Must remember to see if it works at 3.3V too, it claims to be 5V only, but claims on eBay
    need generous pinches of salt.

    The shoulder is 22mm, same as NEMA17's and the two M4 mounting holes are 52mm apart.

    They originally stocked a 10W part but it went out of stock, not sure of the details for that now.

    [stop press: encoder not reliable at 3.3V - level shifting required. Also determined that the 5V
    supply is only for the encoder, the control circuitry derives its power from 24V supply ]
  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-03-14 09:37
    Mark_T wrote: »
    encoder not reliable at 3.3V - level shifting required.]

    I'd think this would be rather simple. Just provide 5V to the encoders and use 10K (or so) resistors between encoders and Propeller. Right?

    I'm still really curious about the torque. Could these motors power a BOE-Bot sized robot without a gearbox?
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-14 12:05
    You can figure out the torque, its 6400 rpm no-load and rated at 20W...

    20 * 60 / (2 * pi * 6400) = 30 mNm
  • ercoerco Posts: 20,255
    edited 2015-03-14 16:14
    Duane Degn wrote: »
    Could these motors power a BOE-Bot sized robot without a gearbox?

    IMO, Mark answered that question with:
    Mark_T wrote: »
    About 6400 rpm at 24V no load,

    Looks like the motor is quite torquey and cogs at 90-degree increments. I don't have much experience with BLDC motors, but without any gearing, I doubt it's suitable for any sort of fine motor control. Great for raw power output though.

    There, now go and prove me wrong Duane. My stomach is howlin' for a slice of humble pie.

    (Obscure Little Rascals' reference when Stimey tells his friend that "divorce" is something to eat. "Mmm, mmm, mmm, my stomach is howlin' for some di-vorce.")
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-14 17:45
    I laughed!

    No its not cogging, I'm driving it with a PID loop and its turning 270 degrees each time under encoder
    feedback. It has very slight cogging 9 times per revolution and runs extremely smoothly at all speeds
    if not used as a servo-motor.

    [ And to show it under gentle ramp-up-ramp-down cycle: https://www.youtube.com/watch?v=xpaIxgPDGxI&feature=youtu.be ]

    The 24V consumption varies from about 7mA stationary to 130mA at full speed no load (6400rpm). I'm using
    32kHz PWM of something like that.
  • Duane DegnDuane Degn Posts: 10,588
    edited 2015-03-14 19:06
    Mark_T wrote: »
    You can figure out the torque, its 6400 rpm no-load and rated at 20W...

    20 * 60 / (2 * pi * 6400) = 30 mNm

    Would one really use the 20W figure in the calculation?

    Shouldn't one use the power used rather than the power rating?

    The same equation using 3.12W (.130A * 24V) gives a torque of 4.7mNm.

    I'm thinking erco's right about it not being a good BOE-Bot motor without some gearing.
  • ercoerco Posts: 20,255
    edited 2015-03-14 21:11
    Mark_T wrote: »
    No its not cogging, I'm driving it with a PID loop and its turning 270 degrees each time under encoder
    feedback. It has very slight cogging 9 times per revolution and runs extremely smoothly at all speeds
    if not used as a servo-motor.

    Great work Mark! So are you actively powering the motor in order to get to stop at a non-cog location?
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-15 11:22
    Its a servo loop, the motor is driven by a PID loop that compares actual to set point positions.
    PID loop running at around 1 to 2kHz, PWM is 30kHz, so latency is pretty good. The coarse
    granularity of encoder means it jiggles a bit around the set point but as its a small diameter
    rotor the MoI is very low meaning it can do several full rotations faster than the eye can see
    between stops!

    Shall keep an eye out for the 10W version since that is probably a better size for most
    projects.
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-15 11:26
    Duane Degn wrote: »
    Would one really use the 20W figure in the calculation?
    power = torque x angular-velocity

    Shouldn't one use the power used rather than the power rating?

    The same equation using 3.12W (.130A * 24V) gives a torque of 4.7mNm.
    The 130mA includes switching losses, copper losses, iron losses and
    friction losses (true torque) - so about 4W is wasted at full output, 20W useful
    mechanical, from 1A.
    I'm thinking erco's right about it not being a good BOE-Bot motor without some gearing.

    Yup, that's right, 0.03 Nm is not big. But all DC motors without gearing display a rough relationship
    between torque and rotor volume, owing to the properties of steel and copper...
  • Mark_TMark_T Posts: 1,981
    edited 2015-03-22 09:07
    I guess what this motor needs to be a stepper-replacement for CNC/3D is an adaptor that
    takes its non-standard mounting holes to NEMA17 mounting holes and includes a 3:1
    planetary reduction gear.

    Yes, NEMA17 steppers tend to be 0.4Nm torque, but that's hold-in torque, dynamic torque
    is a lot less whereas the BLDC has approx constant torque across the speed range. I reckon
    0.1Nm will be enough (especially as that's continuous rating), and 2000rpm top speed is
    nice. Lack of misstepping and stall detection are both great to have.
Sign In or Register to comment.