Variable reluctance motor drive?

Hi There are two types of steppers. One type has a permanent magnet the other does not. This means that one will hold with no power applied while the other will not. For the type that has no magnet, turning the power off to a winding before it is safely centered on that pole will cause it to start up in an erratic direction. Does this sound like the problem your having? The other thing is that you have to ramp the stepping rate and avoid staying in the resonant region for that system when changing step rates. Dwight

(I'm the OP)

I'm now having my doubts I am using a VRM. The motor has definite cog stops which I believe would indicate it's *not* a VRM.

The motor is clearly an OEM model. It has 36 cog stops. It has three windings with a common point verified with an ohm-meter. I can see nine winding "lobes" through holes in the rotor. The windings are stationary and are mounted on a PCB. Each lobe's winding axis is radially oriented from the rotation axle and distributed evenly (360/9 = 40 deg separation). The rotor is cup shaped hiding a clear view of the internal parts. It appears to have some sort of black material ringing the inside periphery of the cup sides. I assume that is a permanent (or a group of permanent) magnets.

I wonder if this motor is a special type of three winding stepper motor that would require microstepping of some kind to spin it up and down smoothly. Its former function (IIRC) was a direct drive capstan motor for a PC tape drive. Without microstepping, there's only two ways to drive it: turn on one winding at a time in succession or or two windings at a time in succession.

JJS

or both: 1,1+2,2,2+3,3,3+1

how fast are you trying to run it?

-Lasse

You're right. That's what I meant to say but failed. Both ways still cause erratic stepping.

I'm running it very slowly, about 4 steps per second. I could be fighting some sort of resonance problem. I've noticed that stepper motors like a certain amount of torque load which might be part of the problem. There's so little I know about this motor. :(

JJS

Sounds like it could be a Brushless DC Motor (BLDC) and the common is the star common point of a 3 phase BLDC, number of poles would have to be determined from the analysis of steps and steps required for one revolution.

BLDC are quite common in disk and tape drives, this could be a sensorless motor that relies on driving two coils and measuring the third in various ways to determine position/speed/torque/load.

......

Still sopunds like a BLDC.

Sounds like a BLDC, look at Maxxon motors data sheets for a comparison.

This layout is typical, the nine lobes are the three windings split to give an 'interleaved' windings on a PCB so the motive force is generated at three points around the axis, possibly giving smoother and drive and less vibration.

Have you checked the impedance of each winding and compared impedances? If the impedance is doubled when measuring across two windings, this MIGHT confirm the star configuration.

Unit is likely to be 12V drive (could be 5V), and can be driven with 3 phase sinusoid or trapezoid waveforms, varying frequency changes speed of rotation, changing voltage drive gives changes in torque. By using PWM it is possible to simulate the average drive voltage level to give varying torque.

Basically you drive one winding to VCC (or PWM modulated) and another winding to GND (or -ve rail), then step through a sequence to get rotation. The star point is often used to measure the effects of third winding being undriven becoming a generator.

Reversing the sequence gives you the reverse rotation.

This is documentated in lots of places and lots of website tutorials exist.

Capstan motors are/were often BLDC motors for simplicity of driving, cheap to make, efficiency and controllability.

If you are getting any rotation currently by only driving one winding at a time it is inefficient.

I have driven BLDC from a simple controller and FET drives from 0-1000's RPM. Get it working in open-loop mode first then start adding feedback to get closed loop control.

Really sounds like a sensorless brushless DC motor, as mentioned by another poster.

Easiest way to drive these is by using e specialized driver IC, the TDA5140A is one example of such a chip.

I think you're right about it being a BLDC motor. An ohm-meter does confirm the star configuration (1.9 ohms at DC across one winding and double that across two, very easy to confirm).

I think that can be described as bipolar drive for a three phase Y motor.

Paul (and others who responded via usenet and email), thanks for the excellent advice and pointers!

In summary: It's most likely a BLDC and certainly not a VRM as I originally thought. My homemade stepper driver is built for one way current switching (unipolar) so major changes would be needed to achieve bipolar current switching (IOW, some sort of H bridge driver).

JJS

I agree - all evidence support that observation.

As a newbie, I don't understand by it's called a brushless DC motor. The "brushless" part is obvious. But the "DC" designation seems wrong. It clearly needs an AC drive applied with just the right timing to work properly.

According to Wikipedia, a BLDC motor is the broad class of motors that includes VRMs and stepper motors. In conclusion, I believe my motor can also be called a three phase Y bipolar drive stepper motor which is in the BLDC class.

Thanks for the help,

JJS

the

It's DC because the current is only traveling in one direction. That is from the driver through the coils and to the return. In AC it would be traveling in both directions, first one then in the opposite, making it alternating.

The current travels in both directions as the phase rotates (remember each phase on the Y is driven at different points as +, - and floating). Each phase sinks or sources current depending on what portion of the drive cycle it is in. The only difference between a BLDC and a PM AC motor is the name. Some make a distinction on the drive waveform but then those who drive BLDC with sine waves couls claim that they were converting them into PM AC motors.

BLDC came to be called DC mainly as a result of them being considered DC motors turned inside out and having the commutation done electronically* rather than via brushes as near as I can tell. Also they conventionally are fed from a DC bus where AC motors are conventionally fed from the AC line. Of course with modern drives that has changed but we haven't started calling AC motors DC simply because we feed them from an invertor.

Switched reluctance motors are a different beast. They are generally driven in a single direction through the coil (which direction doesn't matter). I still don't really consider them a DC motor myself but they really don't fit the term AC either.

Robert

• and thus sometimes they are refered to as ECM (electronically/externally commutated motors).