View Full Version : Rolling commutator 'brushes'

10-06-2012, 03:44 PM
Well there is metal involved in this. I'm wondering if there's a good reason to not have a rolling contact to an armature commutator instead of brushes. I have an application where the length of the armature must be held to an absolute minimum. I'm thinking of making the comm as a thin disc rather than a lengthwise structure. Then because it's thin, I'd have a pair of rolling contact discs in contact with the edge of that. It would take up more width, but it would constrain the length. Each roller might have a brush contacting it to carry the current, but that could be done without extending the length of the motor.

This is just an idea. I know that a rolling contact disc is going to be noisier than a sliding brush on a nicely machined commutator, but that's of little concern. I think there's also going to be a little 'knock' where each comm segment gap passes the rollers, and there will be the momentary shorting between comm segments as the gap comes past the rollers. These are probably both idea-killers, but just wondering if there's ever been anything made like this-

Of course, the best answer is to build the motor inside out, with stationary coils and electronic switching, but that also means the armature gets to be too wide for my application idea.

I'm not asking for much am I - the narrow width of a typical PM motor with a flatted can housing, with the short length of a pancake-type motor.

10-06-2012, 04:27 PM
The momentary shorting is there with conventional brushes. The bigger problems I think would be debris holding the roller off the commutator (as with roller type steady rests) and perhaps intense vibration or pounding of the roller, since the armature will not stay round for long. Also, by adding a brush to provide current to the roller you add another potential problem area.

John Stevenson
10-06-2012, 04:42 PM
What about a face comm ? Can't get any shorter than that .

10-06-2012, 05:01 PM
I could use a face comm and some low-profile brushes. Now that I'm thinking about this again, I could also put the comm external to the rear bearing, with the wires coming through a hollow in the end of the shaft. The brushes would be external to the housing the motor has to fit within, which might be ok. They would have their own protective cover. I've done that before, and it works. Makes it easy to clean the comm and re-machine it if needed.

There's also the post mounted armature which I'd forgotten about. The post is fixed, and the bearings are inside the armature, while the comm is in the usual position except that no shaft extends out beyond it. Hm, I think this is my best option so far if I want to stay with a comm.

Alan Douglas
10-06-2012, 10:30 PM
I have an idea that the Ford model T used rollers in its low-voltage "timer" (distributor) but I can't think of anything else that did.

My understanding is there is no rubbing contact in a commutator; the brushes float just above the surface and the current is carried by a plasma film between brush and commutator. Another brush contacting a roller would only move this plasma somewhere else, and there probably won't be enough surface area, either at the roller-to-commutator or the roller-to-brush point. Also the roller-to-commutator contact would be copper to copper. Copper brushes went out of use in the 1890s and I expect for good reason.

john hobdeclipe
10-06-2012, 10:47 PM
What would you use for bearings on the rolling contacts? It seems that anything sturdy enough to hold them steady and not wear out quickly would have to take up a lot of space. And just a wee little bit of unbalance or out-of-roundness would translate into a lot of vibration and shock.

J Tiers
10-06-2012, 11:43 PM
Copper brushes went out of use in the 1890s and I expect for good reason.

Extremely good reason... they flat didn't last long enough to be commercially usable for any tough usages (or many others), streetcars being one such... IIRC I read that Charles van DePoele was the one who tried the carbon brush first, but I could easily be wrong, as it isn't something I *need* to keep in mind.....

It seems that "everyone knew" that carbon was way too high in resistance to work..... while in fact, the high resistance is one reason it DOES work.

10-07-2012, 02:17 AM
The idea was that the copper discs would roll, and thus would not drag along the comm like a brush. But for all the reasons mentioned, I can see why it wouldn't be very workable.

A brush, having some resistance, limits the current that would flow between comm segments when two segments are in contact with the brush. This would have the effect of reducing the idle current, and may be one good reason why a carbon brush 'works'. Of course the low rate of wear keeps them working longer. But otherwise, having resistance means they become a point where power is lost and heat develops.

I recall a time when I first learned about brush composition. There was hard, medium, and soft- there was copper impregnated, silver impregnated, and carbon impregnated silver brushes. I've also seen brushes which appear to be 'painted' with either a copper colored or silver colored layer around the outside, and sometimes this coating also seems to be the 'glue' that holds the brush to a holder. As far as I know, the carbon/silver brushes were the best, and maybe still are. But copper/carbon seem to be the most common where current is high and voltage not so high. In line voltage motors, I still see what looks like pure carbon brushes. I'm thinking that's because the voltage between comm segments is relatively high, and only a 'resistive' brush will be able to limit the losses in that area.

10-07-2012, 05:41 AM
My understanding is there is no rubbing contact in a commutator

I can't buy that. If there was a high current plasma between the brush and comm the brush would only last minutes or at best a few hours. Plasma is extremely effective at eroding almost anything.

J Tiers
10-07-2012, 12:57 PM
I can't buy that. If there was a high current plasma between the brush and comm the brush would only last minutes or at best a few hours. Plasma is extremely effective at eroding almost anything.

It's not a large area plasma, but IIRC the individual "contact" areas have an odd sort of current conduction mode. I have a fairly good sized book which is all about brushes, which I have not read for a while. I can look it up.

10-07-2012, 06:34 PM
The theoretical contact between a disc (commutator) and another disc (rolling brush) would be a line. The Theoretical contact area between a disc (commutator) and a concave brush surface would be a curved plane. No? Would a rolling circular brush give you sufficient contact area for the required transfer of current? Just wondering...

Alan Douglas
10-07-2012, 09:18 PM
I don't think it would.

As far as I know, the carbon/silver brushes were the best, and maybe still are.
My old EE prof once mentioned buying a surplus lot of tachometer generators that had silver brushes.

10-08-2012, 01:09 AM
Theoretically a line contact is quasi dimensionless. It has no area.

10-08-2012, 02:14 AM
I suppose we could compare the line contact using low resistance conductors (copper rolling disc and copper comm) with the area contact where one conductor has significant resistance ( carbon brush with copper comm). In either case we could have conductor bounce, which we might ignore for the sake of the comparison, but bring up again as regards this plasma issue.

I'm concerned at this point that our line contact is going to have a variation in the length of the line as the rolling disc wobbles, even imperceptibly. Essentially then, we would have little more than a point contact- which we could of course compare to the contacts in a switch. Now, there is going to be some area contact anyway because the materials will 'crush' together to some extent as you add some pressure, whether it's a switch or a rolling contact.

I'm not sure if the mechanical properties of a carbon brush would actually enable an area contact- if the brush was 'worn in' to the comm, you would think there would be area contact, especially when there's pressure from the brush spring. I wonder if that is ever really the case, especially when the armature is turning. It seems likely that there will always be random areas where there's gap- maybe not large gaps, but enough that were electricity to be flowing across those sections, it would indeed be through a plasma. The resistance inherent in the brush material would generate some voltage between the brush and the comm, so given thin enough gap areas, current could flow there as well as through full contact areas.

I don't know that I can say whether one system or the other would have lower loss, though I believe as some have suggested that it would take a bit of an engineering feat to make the rolling disc system work.

10-08-2012, 04:14 AM
the contact area would not be great enough to pass the current, also the brush must span more that one comm seg.

in large variable speed motors the brush could be lets say 1 inch wide with the segs. only 1/4 inch .

it also matters as to the winding , wave or lapp on the com.

stator and armature winder in the past