Is there a benefit in smoothness of rotation by increasing the outer rotor magnetic poles? Is there a rule of thumb for a 24 slot core vs number of magnetic poles?
3
-
\$\begingroup\$ If you are running your motor from line frequency (50 or 60 Hz), then Marco has your answer below. But if you're running from a variable frequency drive (which may be true, based on you mentioning "at low speeds"), then the answer will be more complex. Which is it? AC line or VFD? \$\endgroup\$– MarkCommented May 25, 2016 at 9:22
-
\$\begingroup\$ I assume that this is a continuation of your previous question and that the information provided in that question and the comments apply to this question. \$\endgroup\$– user80875Commented May 25, 2016 at 14:15
-
\$\begingroup\$ Of course it's VFD...as the speeds will be 33.3 RPM, 45 RPM, 78 RPM (vinyl turntable direct drive). The speed will be monitored and corrected on the fly by a closed loop system taking account of the speed of rotation against a crystal clock. \$\endgroup\$– J.K.Commented May 25, 2016 at 16:56
Add a comment
|
3 Answers
\$\begingroup\$
\$\endgroup\$
6
Now that I know that your building a turntable, I have some tips, although not necessarily an answer. Instead, I will expand on Charles answer above.
To minimize wow and flutter (turntable terms), you need to minimize torque ripple (a motor term). Besides the motor, the drive electronics and platter play a big roll.
What you want is a motor that is magnetically optimized for a sine-wave drive. Many (maybe even most) brushless motors are optimized for a trapazoidal drive, and they will likely have too much torque ripple.
Then, your electronics need to provide a sine-wave drive. That means that you can't depend on hall sensors for feedback, but need an encoder for high enough resolution. But I assume you have the encoder already, since you need it to phase-lock to.
Another way to produce the sine-wave drive instead of the encoder is to do what Sansui did in their high-end direct-drive turntables (look at the SR-929 and SR-838). They added windings to the motor to sense the rotor position in an analog fashion, and using the sine-wave sensed from those windings to drive the motor windings. It had very simple drive electronics, but the motor was custom.
Increasing the number of poles raises the frequency of any torque ripple that you may have left, and that makes the rotational inertia of your platter more effective at damping any speed variations caused by that ripple. As you know, the heavier the platter, the smoother the ride.
-
1\$\begingroup\$ As I understand it, J.K. is trying to build (or modify) a permanent magnet synchronous motor. A 3-phase PMSM has essentially the same construction as a BLDC motor, but it should be possible to operate it open loop with tight frequency control of the VFD. \$\endgroup\$– user80875Commented May 26, 2016 at 1:52
-
\$\begingroup\$ @Charles Cowie - Yes, You're right, you could drive it open-loop, similar to microstepping a stepper motor. Not having done that, I don't know the caveats, but I don't see why it wouldn't work if the sine wave is smooth enough. \$\endgroup\$– MarkCommented May 26, 2016 at 6:24
-
\$\begingroup\$ Thinking some more about it, when you drive a PMSM open loop, the phase delay (the lag of the mechanical position to the electrical position) is dependent on the friction and the motor torque. But motor torque will vary with current, which will vary with the electrical angle. So unless you have very low friction, you could experience torque ripple. That could be mitigated by a platter with more rotational inertia, but a heavier platter would increase friction (that's why high-end platters have all their weight along the outside edge: maximize rotational inertial while minimizing weight). \$\endgroup\$– MarkCommented May 26, 2016 at 6:42
-
\$\begingroup\$ A clarification: When I refer to a PMSM (permanent magnet synchronous motor) I'm talking about a motor with one primary winding typically driven from the AC line, and a secondary winding that's usually driven through a capacitor to create phase lag. The above torque-ripple issue APPLIES to this motor. A BLDC (brushless DC) motor, on the other hand, has three equal windings, 120 degrees apart. When a BLDC motor is driven by three sine-waves, 120 degrees apart, then torque is constant, and the above torque-ripple issue DOES NOT APPLY. Since J.K mentioned "3 phase", that would mean a BLDC motor. \$\endgroup\$– MarkCommented May 26, 2016 at 8:17
-
\$\begingroup\$ I took "3-phase" to mean 3-phase sine-wave power as produced by all power station generators and nearly all variable frequency drive (VFD) control units and utilized by most ac motors larger than about 3 Hp (2.2 kW). A BLDC motor driven by three sine-waves is essentially identical to a 3-phase permanent-magnet synchronous motor powered by a VFD or mains power. \$\endgroup\$– user80875Commented May 26, 2016 at 13:33
\$\begingroup\$
\$\endgroup\$
3
Synchronous speed of induction machine: \$N[rpm]=\dfrac{60[s]\cdot f[Hz]}{p}\quad p..\text{number of pole pairs} \$
Increasing the poles number, the nominal speed is decreased.
answered May 25, 2016 at 9:00
-
\$\begingroup\$ The number of pole pairs is p/2 \$\endgroup\$– user80875Commented May 25, 2016 at 13:52
-
\$\begingroup\$ @CharlesCowie One pole pair is made of two poles North and South. 2p=1 means two pole machine at 50Hz it has a synchronous speed of 60*50/1=3000rpm. \$\endgroup\$ Commented May 25, 2016 at 14:18
-
\$\begingroup\$ @CharlesCowie I have corrected, to be more readable. \$\endgroup\$ Commented May 25, 2016 at 14:28
\$\begingroup\$
\$\endgroup\$
3
In my estimation, for smooth operation, you would like as many slots per pole as possible and slots as narrow as possible. The phase coils should be distributed rather than limited to one pair of slots per phase. The slots will cause the reluctance of the magnetic flux to vary with rotor angle. I suspect that it would be preferable to have an interior permanent magnet rotor, one in which the permanent magnets are in the interior of the rotor iron rather than affixed to the surface.
For the VFD, I suspect that some form of multilevel inverter topology would be preferable.
Is there a rule of thumb for a 24 slot core vs number of magnetic poles?
Excerpt from John H. Kuhlmann, Design of Electrical Apparatus, Second Edition John Wiley & Sons, New York, 1940
More than one coil can be wound through a given slot. I believe that coils can overlap to some extent, but I don't have an explanation of that.
-
\$\begingroup\$ So if my math is correct the most poles i can have is 8. \$\endgroup\$– J.K.Commented May 27, 2016 at 8:08
-
\$\begingroup\$ I am sticking with external rotor motor as it has more torque than the opposite...I am testing an optical sensor system that scans directly from specially engraved under platter indentations. Come to think of it...a 36 slot 12 pole motor looks on paper to be even better (but much more expensive!). The platter i am using ranges from 6Kg to 12Kg in different models so i am using the flywheel effect to smooth-en out the movement. \$\endgroup\$– J.K.Commented May 27, 2016 at 8:35
-
\$\begingroup\$ For the matter of friction...i have designed an "oil bath" cartridge for the axle to spin in (on a ball-thrust plate scheme). The ball is relieved (up to a 90%) from the platter/rotor weight with the use of opposing ring magnets. So minimum friction is achieved. \$\endgroup\$– J.K.Commented May 27, 2016 at 8:42