Outer rotor (outrunner) motor control circuit 1

[Vinylforever]

I am an electronic ignoramus. I wish to use a motor such as this

http://www.hobby-lobby.com/brushless-axi2204.htm

for a hifi turntable. These motors are primarily aimed at the radio control aircraft hobby market, so the only controllers I can find are for use with R/C systems. Can anyone suggest a source of motor controllers for this sort of motor which allow accurate speed control, running from the AC mains rather than batteries?

 

[Skogsgurra]

Is this for s product you are developing?

Gunnar Englund

http://www.gke.org

 

[Vinylforever]

No, it's just a hobby at the moment. However, i'm currently unemployed, so it may develop into something more if my experiments turn out succesfully!

 

[benta]

Not a good choice in my opinion. Sensorless brushless DC motors (BLDC) exhibit relatively high torque ripple, which would probably be transmitted to your turntable even if you use belt drive (which I suppose you do).

Benta.

 

[Vinylforever]

What sort of motor would you reommend? Conventionally, a small very low power 12-pole synchronous motor is used, running at around 300 rpm and driving through a rubber belt. I want to go back to the type wich was popular in the 50s, usng a more powerful (100 watts or so) faster revving motor driving through a rubber idler wheel. The attractions of the outrunner motors are small size, very high torque for that size and also the fact that they look sexy! If i'm going to use an unconventional technique, I want to be able to show it off, although this must obviously come second to sonic performance.

 

[Skogsgurra]

I asked if it was for a product only to make sure it wasn't a hobby project. There is a policy about that in the forum rules.

Gunnar Englund

http://www.gke.org

 

[Vinylforever]

Ah. Apologies. Should I bow out and seek information elsewhere? I wouldn't want to break your rules.

 

[nbucska]

I made a control circuit for a sensorless BLDC motor
which at 3000 to 7000 RPM exhibited less than 2 microsec
variation from rev. to rev.

It wasn't difficult...

 

[Skogsgurra]

I think that your question could lead to an interesting exchange of ideas. It is a new (at least for me) combination of technologies.

Gunnar Englund

http://www.gke.org

 

[Vinylforever]

Well Skogsgurra, I certainly hope so! Any chance of further details, Nbucska? What wasn't difficult for you could be very difficult for me: we each have our own expertise.

 

[smorein]

I have also been thinking about using those motors (the brushless DC motors for RC aircraft) in different (noncomericial) applications.

They are far less expensive than alternative brushless motors.
($200 as I recall for a 1HP capable motor)

One big issue is the need for cooling. They are intended to be mounted on an RC aircraft with signficant airflow for cooling, you are going to need to add a fan and other details...

I was thinking that I would add a something, whether a magnet&hall sensor or quadrature optical for comutation rather than run sensorless.

My highest level of curiosity is about the durability. Unlike the brushed RC aircraft motors, the only wear item should be the bearings.
(assuming you do not let the magnets get hot, they are NdFeB, and as soon as the get above a relative low temperature, no magnet.

Really curious about the technology in the speed/comutation controllers that are sold to go with them.

 

[MikeHalloran]

Real breakthroughs come from people who are too ignorant to know what can't be done.



Mike Halloran
Pembroke Pines, FL, USA

 

[Skogsgurra]

PLS for you Mike!

Gunnar Englund

http://www.gke.org

 

[Vinylforever]

Mike, you have it in a nutshell. My problem now is to find out from those who know it can't be done how to do it!

 

[benta]

Vinylforever:
I understand your application a bit better now. My comment on sensorless BLDCs stands, however.
It is possible to drive BLDCs very smoothly using sinewave drive, but this is only feasible using motors with commutation sensors (Hall sensors, or whatever).
You might look to Toshiba Microelectronics, they have a range of controllers for sine-wave drive. You do need to add a sensor circuit to the motor, though.

Benta.

 

[Vinylforever]

Thanks, Benta. This is the sort of I nformation I need. Is fitting a sensor circuit likely to be practicable, do you think?

 

[MikeHalloran]

I suspect that, given a load with fixed and nontrivial inertance and without a lot of friction, you could find a starting speed and ramp rate where a brushless DC motor core would follow a polyphase driver without needing position sensors.

I.e. if substantial acceleration is not involved, one cycle should look pretty much like the next. I.e., you may not know where the rotor is in space, but you know approximately where it is in time.




Mike Halloran
Pembroke Pines, FL, USA

 

[Vinylforever]

Does this mean that if the load inertia is constant (which it is) and comparatively large compared to the motor's own inertia (which it is), then the motor will follow a slowly increasing supply frequency, and once that frequency is constant, the motor speed will be comnstant too? As I said, I am an ignoramus, so I need everything explained in layman's terms. (Always an interesting excercise for an engineer...). What happens if the load increases slightly, will the motor "slip", or overcome the load?

 

[Skogsgurra]

Yes, that is correct.

In fact, most synchronous motors work without position sensors. The trick in your case is to feed the motor a voltage that corresponds to the flux derivative as the motor turns at the desired speed.

In most larger motors, that would mean a sine waveshape. In these smaller motors it probably means that you need a distorted waveform to get a smooth running (to avoid rumble). Such a waveform is easily synthesized and needn't add much to the cost.

Gunnar Englund

http://www.gke.org

 

[nbucska]

A turntable has a narrow range of speed -- it is possible
to phase-lock to a back-EMF sensor and if necessary generate a sine-wave of any reasonably small distortion.
As everywhere, it is a compromise between performance and
complexity i.e. cost.

During acceleration the rumble doesn't matter so the PLL
may be switched in only after the nominal speed is reached.

<nbucska@pc33peripherals.com> omit 33 Use subj: ENG-TIPS
Plesae read FAQ240-1032