OT Electric " train"

[harryagain]

You'll like this!

[harryagain]

Well, you should also like this as well. The only true DC motor.

[Another John]

You are right Harry: I like that :-)

J.

[Colin Stamp]

This one will quite happily run on DC without commutation.

And it runs on AC too!

Cheers,

Colin.

[newshound]

I liked both of them! I do recall seeing a similar homopolar motor made at school using a bowl of mercury (pre-neodymium magnets, of course).

[harryagain]

I think that one's a fiddle. Look at the "bearing" on the LH side.

[harryagain]

Ah.

[Syd Rumpo]

It's a Marinov motor. The current goes through one ball race - flowing through the balls - along the shaft then out through the other ball race. There are no magnetics involved, the balls get hot as the current flows, and hottest at the contact points. Thermal expansion deforms the balls into something like ovoids which press against the ball race housing. This unstable condition will 'fall over' and repeat as the next part of the ball expands. The process is continuous, the balls constantly 'falling' sideways as the thermal expansion moves around them. They usually need to be started one way or t'other, efficiency is risible and they get very hot very quickly.

Cheers

[Syd Rumpo]

That's it. The cited electromagnetic explanation is bunkum though, it's driven purely by thermal expansion.

Cheers

[Chris Hogg]

Quite a lot of examples of Marinov motors on YouTube. This, for example

. I'm still slightly sceptical about the proposed mechanism, but I don't have an alternative. Has it been proven, or is it just 'we can't think of anything else so that must be it'? The explanation would be supported to some extent if the effect was specific to ball race bearings, and didn't work with simpler sleeve bearings, for example. Does it work with roller bearings, I wonder.

[Colin Stamp]

That video is from the same bloke as the one I posted. His channel is well worth checking out if you're into electrical curiosities.

I'm not quite 100% convinced about Marinov's theory of operation either. I wouldn't mind reading more about the supposed electromagnetic theory mentioned in the Wikipedia article, but it seems I'd need to pay. I'm guessing the Marinov explanation is more plausible anyway...

Cheers,

Colin.

[Syd Rumpo]

I think Occam's razor applies here. The thermal explanation works perfectly well; the balls expand most at the contact point because they are the points of smallest area thus highest electrical resistance so most heat per unit volume. It doesn't work with, for example, sleeve bearings, only balls and presumably rollers.

It works on DC - either polarity - or AC, and can go in either direction, which rules out magnetics as far as I can see. Also, any magnetic fields are very small, there are no windings, just a single turn from the PSU through the device and back. It works without a flywheel or with a flywheel made of anything. It's certainly not a homopolar motor as suggested by some who don't understand what a homopolar motor is - it won't generate current if rotated.

It's essentially useless, but something similar may have applications in MEMS technology if the 'balls' were perhaps piezoelectric discs.

Cheers

[Chris Hogg]

Hmm. Several things I don't get. The outer and inner races are both thick metal (usually) and a good electrical conductor, so any voltage applied to the outer race will be the same all around the race, and current will flow radially from the outer race, through the balls at their points of contact with the race, to the inner race. Even if there is local heating at the contact points, and local expansion at those points, I don't see where the driving force for rotation comes from. The system would seem to be circularly symmetrical. And I'm not sure that the balls would expand in any way other than uniformly through the whole sphere, rather than in some oval fashion. They are going to be pretty good conductors of heat, after all. Lastly, the balls are (or should be) a pretty snug fit between the inner and outer race, without much room for expansion unless they deform either or both races elastically, and as I said earlier, I would expect it all to be symmetrical around the bearing and I don't see where the turning force comes from.

It would be interesting to see a diagram of how the theory is supposed to work, preferably an animated diagram such as those sometimes see in Wikipedia.

>

[Chris Hogg]

OK, I'm prepared to go along with localised expansion et the points of contact between the balls and the races, and I'm stretching my imagination to see how this produces a rotating force, even allowing for the expansion being slightly displaced from along the radius of the ball race by the initial rotation. Perhaps!

As I said, a diagram would be nice, preferably an automated one (I'm not asking you to produce one, BTW). I'll dig around on Google until I find something, or get bored!

PS: I see you've been discussing this topic before :-)

[Syd Rumpo]

Push against a sliding door at right angles to its intended motion. Nothing much will happen. Now do the same on an already moving sliding door. Your arm will move left or right with the door, and some of your push will now be sideways, helping the door move.

With the Marinov motor, it's a continuous process and the pushing is caused by thermal expansion due to ohmic heating. As soon as the contact point (your arm) touches the ball race (the door) it expands a little - as soon as it loses contact, it cools and retracts a little.

I'd completely forgotten that. It's an interesting device.

Cheers

[Tim+]

Try this link instead. Same video but without the bazillion ads.

[Martin Brown]

Seems quite reasonable to me. Rather hard on the bearings though! Efficiency will be risible - mostly ending up as waste heat.

A less violent version using a bicycle wheel suspended from its hub by a large number of identical rubber bands with one side give or take heated by an anglepoise does a similar though much slower motion version of an expansion based mechanical heat engine.

Although in the case of rubber bands heating makes them contract raising the centre of gravity and causing the wheel to turn. The direction of rotation can be altered by moving the lamp. eg.

[Syd Rumpo]

This one works in reverse, too. Build your own fridge!

If you take a thickish rubber band, stretch it, hold it stretched for a few seconds then release it gently while holding it against your philtrum, you can feel a definite cooling.

I'm guessing trapped gas bubbles being compressed when the band is stretched, but that's only a guess.

Cheers

[Martin Brown]

You guess wrong. It is an entropy thing. Rubber is a tangle of long polymer chains and stretching them introduces order. Allowing it to relax permits disorder again and as a result a small amount of cooling.

Rubber is one of a handful of materials that do not expand when heated.

A similar trick is used in cryogenics by first aligning magnetic spin and then relaxing the field to chill the isolated system.

[Syd Rumpo]

You're right of course, I looked it up. One thing I found was this...

...so bang goes my Nobel prize for inventing the rubber-band fridge.

Cheers