Printed-Circuit Motor Theory

Thanks! That makes sense... If I understand that correctly, The motor can only stop when a set of armature conductors is centered directly over the centers of the magnets (this assuming the magnets are equally spaced around the motor). Since there are a large number of conductors, There will be a small angular distance between possible stopping positions, but there will be a discrete number of such positions.

But wait a minute! I also understand the motor uses slip rings rather than a commutator. This is important, since the motor must operate in a vacuum, where sparking at a commutator could be disastrous! Now If the motor is stopped with the conductor exactly centered over the magnet, what is done to create starting torque?

Reversing the current will of course cause repulsion rather than attraction, but what if I want to continue in the same direction as before it stopped?

I suspect that commutaion is then handled in your controller with some form of position feedback.

There is definitely feedback. There are two cables leading to the motor assembly. One is a multi-conductor cable, which I presume is the encoder connector, while the other is a two-conductor cable, logically and presumably carrying the armature current.

I've been told that disconnecting the presumed encoder cable does increase the speed of the motor significantly.

Everything is consistent with Nof60's motor descriptions. The driver board is very similar to the 0-90V Minarik™ drivers that control the speed of standard PM DC motors in several other machines around our plant, but none of these other machines have feedback of any kind (Other than back EMF).

I'm away from the plant this week and part of next. When I return, I plan to connect a 'scope to learn more...

After re-reading your post several times, it seems like I understand it less...

"Thus the holding torque will be highest because the stator and rotor fields will be closest to each other when in a holding (zero velocity) position."

This will be true if there is no offset between the angular position of the active rotor conductor and the corresponding brush contact. But if there is no offset, then more current would simply hold it more firmly in the stopped position. Reversing the current would provide repulsion, but with zero offset, there would be no way to predict which direction it might then rotate. The motor does very reliably rotate in either direction selected, so there must be an offset between conductor and brush. It's conceivable that there could be a mechanism for shifting the brushes, but that sounds like a likely failure point...

I'm back to square one on understanding holding torque at zero velocity, unless I accept Andy's rapid switching between cw and ccw. I have difficulty accepting that scenario, since I'm reasonably sure that switching would be audible, and I don't hear it. As soon as I get back to the plant, in about a week, I'll be connecting a 'scope. I'll post what I find...

The pancake motors that I worked with had zero holding ability at all. Once the rotor/disc/armature was no longer powered, they would free wheel to a stop.

About the only way it could have any holding ability was if, like all permanent magnet field DC motors, they had a resistor connected across the armature to oppose generated potential when the motor had rotational torque applied to it.

Yes, the holding torque is generated only when the drive imposes the desired current in the rotor. A disconnected rotor from the current source will spin freely.

As far as I am aware, the holding torque is generated when the driver drives the motor forward and backward at (relatively) high speed....net result is it stays in one position...

Due to the negligible inductance, that can be easily achieved....also very little mass....

Unless you use a current probe with your scope the waveform you see will be even more confusion for you. Remember the B field produced by an inductor will be proportional to the current not the voltage. I realize that the rotor will have a low inductance value since there is no ferromagnetic core material but it will still be the inductance that generates the magnetic field. The rotor inductor will also be generating back EMF spikes when currents try to switch ON and OFF . This will be particularly annoying if the drive is using a PWM technique to modulate the current.

Here's another printed circuit motor web site for you. Here's one that discusses multiple motors including the PCB motor.

Two really excellent links that shows that these motors are generally "Flat".....and are all DC driven.....thanks.

The pancake motors I have worked on (checkweigher weigh table drive motors in the food industry) used permanent magnet stators. They were round and were glued into the cover on the motor. The rotor was fed straight DC via brushes that were 180 degs apart and were fed pure DC from a little DC drive, at a max of 90 volts.

I have no idea what the "poles" were like on the circular stator magnet, however, the magnetic aspects of the stator ring were constant around the full 360 degs of the part.

There had to be numerous pole pairs however as there was no apparent "cogging" like you would expect if there was a limited number of poles. Cogging would add vibration to a checkweigher one would think, which would not be a good thing.

While the physical arrangement of the brushes would suggest "slip rings" they did in fact commutate the DC by the way the traces on the rotor were laid out.

(see my reply #5 to RedFred)

As you say, there is absolutely no cogging of this motor. Likewise, I've only seen photographs of the armature of similar motors, and I don't plan to take this one apart!

This motor has something like a 10-1 speed reduction gear set between it and the table, and there may be additional gearing inside the motor. It must turn at a very steady slow rate; currently the table must take close to 5 minutes to make one rotation. if the speed is just a bit too fast, we fail to get adequate weld penetration; if it is just a bit too slow, we burn holes in the parts.

You have a good point about the commutation. Since it is DC, there must be some form of commutation. I get the impression that there must be well over a hundred radial armature "wires", so the brushes may be in contact with several at once. If the spacing of the magnets and armature conductors is done right, I can imagine a situation where the conductors losing contact with the brushes would be in between pairs of magnet poles, so would have little if any inductive current, and therefore little if any sparking.

I'm still open for further education!

The ones I worked on many years ago, were in large heavy printers, to move and stop the paper very accurately.

They had a glass disk inside on the same shaft, that had on the markings (graticule) for 6 and 8 Lines Per Inch. Which was the only two possible LPI in those days.

The electronics that drove the motor would drive it "effectively" backwards and forwards at a certain (I forget) frequency. The motors would sometimes "squeal" like a pig!!

To hold still, the motor was driven backwards and forwards at that frequency, switching fast enough that it actually didn't move. Increasing the signal drive to one direction against the other would allow the motor to move in one direction or the other.

The greater the difference in signal drive to either the forward or the backward direction, the greater the speed. Speed was held down to a certain maximum by the electronics, 35 feet a second if I remember correctly, just for long paper skips.

There was other electronics to read the graticule on the glass disk to check the distance moved, and to control acceleration, deceleration and maximum speed....also to move in units of 6 or 8 LPI.

The motor would receive quite a large quantity of cool, filtered air to get rid of any heat build up.

The motors did not have a particularly long life as the brushes were a weak point in the design....a year was around the longest such a motor would work. We were not allowed to replace brushes, they had to go back to the factory for cleaning, brush and bearing replacement....

The ones I worked on used 48 volts DC at around 36 amps if I remember correctly.

There was another sensor to check airflow, if the air supply failed, the whole printer was stopped.

Even while just "holding" position, there was a lot of heat caused. To test correct operation, we would try and move the motor by had when "held", motor would "Squeal" but not move as far as the eye could seen. It was quite a strong torque...a human hand alone could not move it.....do you now understand why? Simply being driven rapidly backwards and forwards....

The ones we had, had no physical brakes, only electronic braking was used.

Printer designs that I actually worked on was some time in the 70's. Several different companies. All American. All much the same. We called them PCB motor.

All the above is from memory, I haven't had anything to do with such printer since 1990....

"Printer" maintenance Guys were the real clever people in the computer industry as they had to understand and fix a device full of mechanics and electronics, mechatronics we would say today.....as were the Tape Guys for similar reasons....

I thought of that back and forth possibility, and was going to mention it in my original post, until I also thought that the driver board uses SCRs. I believe (but I'm far from certain) that the supply voltage to the SCRs is 60Hz 3Ø, so any such switching would be clearly audible. This motor, at least to my aging ears, is totally silent.

Judging by the size (TO220) of the SCRs in the driver, this motor can't use more than a few Amps. The motor is around 250mm in outside diameter, with a smaller section at the exposed end, where the 2-conductor cable connects. Only around 80mm of length is visible; I rather suspect there is some planetary gearing in the other end of the unit, since the maximum speed for the table is only a few (3-5?) RPM. The motor housing appears to be quite massive, probably because there can not be any forced cooling, since this unit always does its job in a vacuum.

As I said in an earlier post, there is clearly feedback of some kind, and it works!

Thanks for the info!

If its AC, there will not usually be any permanent magnets inside....or better said, not for long!!! See if there are any inside, that might be a good indicator?

The PCB motors that I mean are technically DC driven, but have the polarity switched rapidly by electronics to provide directional control...its quite an old, but relatively simple concept. Even I understand it!!

Unfortunately, The motor is in the San Diego area, and I'm at home in northern CA this week and part of next. I'm trying to learn as much as possible before I return.

Here's a photo of the unit: The upper-right, lighter color, cylinder is the brake (Held ON by permanent magnets, released by an opposing current). The lower-left cylinder is the motor; it's shaft (s?) drive a small spur gear, which connects to the large gear. The two bronze gears drive auxiliary encoders, which may or may not be in use.

As you can see, there are no openings in the motor, so I can't see inside. I understand that the permanent magnets in these motors are magnetized after final assembly, so I'm certainly not going to attempt to disassemble it! To the right of the motor are visible the two cables that connect to the far side of the motor. The 2-conductor power cable connects to the smaller diameter section of the motor, while the encoder cable connects to the larger diameter section.

The frame for the unit is 25mm thick welded stainless steel, so it is pretty robust.

Since SCRs can't be turned off until pretty close to zero crossing, 3Ø-60Hz can provide a maximum of 180 pulses per second for a given direction. I presume that for the +/- driving you refer to, you would not want to turn on the opposite direction of current until the previous direction of current had dropped significantly towards zero. I definitely agree that it is fundamentally a DC motor.

Thanks again...

From old CEM and Infranor DC pancake servomotors I remember that disassembling is not allowed because of the permanent magnets (if disassembled there is a partial loss of magnetization).

There are several manufacturers of modern permanent magnet synchronous pancake motors (e.g. Heinzmann).

It does not have the appearance of a PCB motor from the Photo.....at least, not like any that I have seen....

I am trying to remember the electronics better and I do believe it was an H - Switch configuration, but I am not 100% certain anymore....rapidly switched as I mentioned before.

PCB motors are generally (in my experience) fairly flat, and can also be called a "Pancake" or even "Pizza" motor......

The motor in the supplied picture is the round flat object on the lower right.

Correct. Right half of the photo, yet left of and below the brake.

Indeed the flat rotor is usually not a real PCB, see for example here:

[url]http://newt.phys.unsw.edu.au/hsc/hsc/electric_motors5.html[/url]

I've see some very small flat motors with a real PCB rotor though.