modifying the speed of a brushless AC motor

Discussion in 'General Electronics Chat' started by staratsx, Sep 30, 2011.

  1. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    I am trying to slow the speed of a 120V brushless AC motor. I don't want to buy a VFD. How can i use a 555 IC to accomplish this?

    If someone could provide me with some reading material on how this is done that would be nice. my knowledge of circuits is limited to that which i learned in physics II years ago.
     
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,729
    758
    Welcome to the Forum

    Ever think of learning μC's such as PIC or ATMEL
     
  3. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    a microcontroller? no I haven't. that sounds like it would be complicated though.
     
  4. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    let me clarify what i'm trying to do. I'm trying to take this motor http://www.spectraservices.com/Merchant2/pdf/Drucker_642E_Rev2.3.pdf), which puts out 1581 G's, and modify it to give around 72 G's. As I understand it, given that it is a brushless AC motor, the only way to slow it down to a calculated speed is to rapidly turn it off/on with use of a 555 IC. I'm sure there are other ways though. What would be the easiest route for someone in my position?
     
  5. iONic

    AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    I'm surprised that it doesn't include a speed adjust feature.
    I know PWM with a 555 is possible for brushless DC motors, but never thought of it's use for AC motors.

    Did you consider a 120V incandescent light dimmer, or one made for ceiling fans?
     
  6. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If it's rated for 1,600 G's at 3380 RPM, then it has a radius of 12.526cm, and you need the rotor to be turning ~717 RPM in order to get ~72 G's.

    The relationship between RPM and RCF (relative centrifugal force) is as follows:
    g = (1.118 × 10-5) R S^2
    where:
    g is the relative centrifugal force,
    R is the radius of the rotor in centimeters,
    and S is the RPM.

    No, you can't use a 555 timer, nor a light dimmer to control the motor speed. The electronics are inside the motor itself. If you try to use some external means to control the speed, you'll likely fry the electronics in the motor.
     
  7. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    Thanks wookie. My rotor is 12.5cm. I've worked out the math already. The factory calibration sticker on my unit reads 3,365RPM, a tad lower than the standard. It still comes out to about ~717RPM as you said.

    And doesn't a light dimmer work by just varying the voltage? if so that wouldn't work....

    So you're saying there's no means by which to lower the speed? What are my options here, if any?

    Thanks!!
     
  8. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I misread the datasheet on the motor. It's a brushless AC motor, not a brushless DC motor - so that means it probably doesn't have electronics in the motor itself. It may have start and/or run capacitors however.

    A regular "squirrel cage" capacitive start AC motor will overheat and shut down if you try to run it at low speeds like that. I am not absolutely certain what will happen to your motor if you try running it on a lower voltage, or using a fan speed control on it - but I'm fairly certain that you will overheat and damage the motor.

    I think it is also a high probability that you will have very coarse control over the spindle speed. After all, you want to operate this motor at about 21.3% of it's calibrated speed; it's not even into it's torque band at that point.
     
  9. iONic

    AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    I knew the The fan controller method would not be a precision control, to say the least, but did not know that you could cause damage. My thoughts were to intercept the power directly at the motor as apposed to from the plug. All electronics at that point would be affected by the reduction from the can speed control. Didn't think about how low the torque would be compared to it's rated torque.

    staratsx, ya tell me they don't make variable units? I assume it's one you already have and want to modify.
     
  10. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    Yes it is one that I currently own. And there are plenty of centrifuges with varying speeds but single speed units are more common. Mostly because higher G's are best for seperations and it's not really sensible to go lower.

    And I don't think varying the voltage will work as I am under the impression that these motors are locked to a specific frequency and it just wouldn't function if I changed only the voltage. If that's incorrect let me know.

    So what then....should I just break down and buy a VFD? lol

    even with the VFD there is still the issue of the torque.
     
  11. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I was looking at what other options they might have, and seems that the lowest that a variable speed model 642VES:
    http://www.spectraservices.com/642VES.html
    ...would go is 1,000 RPM, which would be nearly 140 G's.
    Manual: http://www.druckerco.com/literature/manuals/documents/7711072_642VES_RevA.pdf

    The Drucker Co. has a contact page for support and customization of their products:
    http://www.druckerco.com/contact/customization.asp

    I suggest that attempting to modify the unit yourself without guidance from the factory would likely result in an expensive repair.

    If ~140 G's would work for you, you might be able to simply upgrade the control board. It costs you nothing to ask them, and no one will know your machine better than the manufacturer.

    I'd kicked around the idea of building a whole new one, but the potential kinetic energy of such a machine would present a large safety hazard if not constructed properly.
     
  12. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    Good idea. I sent them an email and asked. Why not??


    I'm not going near that one with a 10 foot pole. I don't have the knowledge or experience to undertake a task like that, and I wouldn't want to be the one standing around a faulty assembled centrifuge as it's whirling around at max RPM. That **** could take my head off. It could probably go through two doors, a window, and still take my head off.

    Thanks for the interest in the thread and all the helpful comments ppl - I'll keep you posted on how this turns out. What I'll probably end up doing is buying a VFD from radioshack and test it out - if it works I'll buy one off craigslist :D
     
  13. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    Question....

    I contacted someone who is selling a VFD for a resonable price. it is listed as:

    "Model KBVF-13 (P/N 9957N)
    Inputs 115VAC @ 50/60Hz single phase.
    Max AC line current 9.6 amps.
    Output 0-230 VAC 0-50,60,100,120 Hz.
    Rated 2.4 amps/phase."


    Here is a manual I found. It is the KBVF-13 Model.

    http://www.kbelectronics.com/manuals/kbvf_2g_manual.pdf


    Would this do anything for me?
     
  14. shortbus

    AAC Fanatic!

    Sep 30, 2009
    4,000
    1,512
    The output of most VFD's and this one is 230V THREE PHASE. Your centrifuge is 115V SINGLE PHASE. You will need to find a VFD with single phase output.

    I don't think that a VFD will work on a single phase motor. The magnetic fields in a single phase motor work different than three phase.
     
    Last edited: Oct 2, 2011
    staratsx and SgtWookie like this.
  15. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    If your centrifuge runs at 3,365 RPM on 60Hz, then it should spin at roughly 718 RPM at @12.8Hz. It will most likely spin a tad faster, as friction losses due to air will be less at lower RPM.

    HOWEVER - I strongly suggest that you consult with the factory before using this VFD with your centrifuge. It's not just the motor that's involved; there is a control board that is expecting to see 120v@60Hz as its' power supply. I have no clue as to what that board even looks like. If the supply uses a traditional transformer designed for a frequency range of 47Hz to 63Hz, trying to operate it on 12.8 Hz will result in a burned-up transformer and a centrifuge that is unsafe; as the lid will most likely no longer lock down when the centrifuge is rotating if the control board is non-functional.

    I still do not know if that motor can be operated at such a low speed; the company's own variable speed version of your centrifuge has its' lowest speed @ 1,000 RPM. I don't know if that is a limitation of the motor used, the circuit design, or what. As I cautioned earlier, the motor may not be capable of operating at low speed without overheating. You need to consult with the manufacturer to find out. All I can do is guess; if I guess wrong, you will be mad at me and out a good chunk of money. I am trying to avoid either of those situations by referring you to the experts on your particular machine.

    Also, you will not have a means of checking the rotational speed of the centrifuge unless you build some sort of frequency counter and interface to the unit; as the standard method of calibration will no longer work, and the unit does not come with an RPM indication as the VES model does.

    [eta]
    Thanks Shortbus; didn't look to see if the VFD was single or 3-phase. No, a 3-phase VFD won't work. A single-phase VFD would be doubtful to work if you could find one, unless the manufacturer says it's OK.
     
    staratsx likes this.
  16. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    Well, it seems the further I take this the more variables arise.

    so no matter what I do to slow down the rotor, I have to worry about:
    a) overheating/damage to the motor (which poses a safety concern)

    b) torque band

    c) the fact that there is a 15-sec. warmup period before the rotor reaches max RPM

    d) with b and c combined, I would likely have to affix some form of tag to the rotor to get the timing down (which poses a safety concern), as there is no logical mathematical appraoch to calculating the speed .

    e) the 555IC and everything would have to be done externally because I wouldn't want to open this thing up and start toying with things like the expert electrician that I'm not, wind up screwing something up and bearing witness to what sort of mishaps could occur when you combine ignorance with the kinetic energy of a centrifuge.

    f) then theres the issue of the VFD, and whether or not it would even work. and if it would, there's still problems a through d


    The only way I can imagine using this centrifuge in its current state is if I reduce it's spin time at the cost of lower sample resolution...which would go something like this:

    So on my unit, if 15mins @ 717 RPM is ideal....that's:

    (717rpm) x (15min) = 10,755 rotations
    (3365rpm) X = 10,755. X =3.196

    So that's going from a spin cycle of 72 G's for 15mins to like 1581 G's for 3min 11sec

    I wonder how significantly my resolution would suffer and if this would even be worth the attempt.
     
  17. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    The variables were always there; we just hadn't thought about them yet.

    If I had to make a wild guess (not so wild, actually) I'd say that there is likely a thermal interlock in the motor housing that will interrupt power in case of excessive temperature. However, this will result in an un-powered decrease in speed of the centrifuge; as only inertia will keep it going. If the motor is cooled by an internal fan, air movement will be minimal while power dissipation will be high due to the near stall condition. More pressing is the likelihood of damage to the safety interlock control board.

    This is currently an unknown, but I suspect that the manufacturer will be able to confirm it.

    That's the acceleration of the motor. It has little torque when starting from a stop; as the motor comes up to speed it will accelerate in RPM more rapidly until it's near the calibrated operating speed.

    This actually isn't terribly difficult; a piece of reflective Mylar tape would not affect the balance of the machine, and reflections could be counted using some sort of strobe; for example an IR laser reflecting onto a photodiode or phototransistor with a counter circuit. If the machine makes an audible sound at speed which corresponds to the RPM, perhaps that sound would also reflect the speed at a lower RPM.

    We generally don't like for new hobbyists to start experimenting with projects that involve switching mains power until they have a good amount of experience under their belt. This isn't a "first project" type of thing.

    The manufacturer needs to be consulted on that.

    I have a hunch that you're now talking apples and oranges. I don't see how 15 minutes @ 717RPM/72g's would correspond to ~3.2 minutes @ ~1,581g's. Perhaps that's something to ask in the Math or Physics forums.
     
  18. SgtWookie

    Expert

    Jul 17, 2007
    22,182
    1,728
    I didn't recall what PSC2 meant that was referenced in the datasheet for the variable speed drive, so I had to do some searching around.
    PSC2 means "permanent split capacitor" motor:
    http://en.wikipedia.org/wiki/Permanent_split_capacitor#Permanent-split_capacitor_motor
    The motor in the centrifuge does have a 4uF 250v capacitor; so it may very well be a PSC motor, and the VS drive may very well work with it; if the winding connections to the motor are accessible, AND the manufacturer says that the motor won't overheat AND the control board will be OK if operating on 12.8Hz.

    But, only the manufacturer can answer all of those questions.
     
    staratsx likes this.
  19. iONic

    AAC Fanatic!

    Nov 16, 2007
    1,420
    68
    Between a rock and a hard place for sure!

    PC HDD's spin fast enough and run on DC which could be regulated with PWM down to 717 RMPs, but you would have the time of your life balancing your sample. This, of course, is a bit of humor. Do not attempt this at home!!!
     
  20. staratsx

    Thread Starter New Member

    Sep 30, 2011
    16
    1
    I had that hunch too, as I'm pretty sure I made a couple false assumptions with that math...it was more of an estimate. I would expect it to be less than 3.2 minutes realistically. Lets give it a fudge-factor of ±50% :D
     
Loading...