Can a DC motor work without either a commutator or a controller?

[k.udhay]

TL;DR Summary

I watched two YouTube videos. The first one showed how a simple DC motor works and the need for a commutator. But in the second video which is DIY of a DC motor, commutator is not used. Can someone help if I am missing something important?

I am a mechanical engineer and my experience with electrical systems is almost nil.

The concept of a simple DC motor explained here was quite interesting, especially the need of a commutator part:


And then I checked this DIY simple DC motor here and was confused because there was no commutator used:


I am confused because with the armature rotating for every 180°, won't there be a change in polarity thereby not letting it to rotate continuously?

 

[DaveE]

Yes, that does seem odd. Here's a quick hypothesis:
The coil (rotor) seems to be unbalanced, eccentric. So when it spins 180^o, it also shifts horizontally. The magnetic field strength also varies in the horizontal plane. So the rotational force in one half may be stronger than the other half because the wire is more centered over the magnet. I do think your question based on symmetry concerns is valid, there MUST be some asymmetry somewhere. IDK, just a quick guess...

PS: Why not build one and test it (you, not me, LOL). I bet if you move the magnet around you can find a null location where it doesn't spin well, or reverses direction. Make sure to let us know if you do that.

PPS: Note that there is also a 1/2 turn asymmetry in the two sides of the coil. To get the wires to come out at opposite sides there must be an extra half turn.

 

[anorlunda]

The secret was hidden in the open comments by the video's author. It is a homopolar motor which is explained at Wikipedia.

https://en.wikipedia.org/wiki/Homopolar_motor

Principle of operation​

The homopolar motor is driven by the Lorentz force. A conductor with a current flowing through it when placed in a magnetic field which is perpendicular to the current feels a force in the direction perpendicular to both the magnetic field and the current. This force provides a torque around the axis of rotation.[9] Because the axis of rotation is parallel to the magnetic field, and the opposing magnetic fields do not change polarity, no commutation is required for the conductor to keep turning. This simplicity is most readily achieved with single turn designs, which makes homopolar motors unsuitable for most practical applications.

 

[DaveE]

The secret was hidden in the open comments by the video's author. It is a homopolar motor which is explained at Wikipedia.

https://en.wikipedia.org/wiki/Homopolar_motorView attachment 302448

Are you sure? Honestly the geometry of motors always makes my brain hurt...
The axis of rotation of the OP's coil is orthogonal to the PM B-field and wrt that fixed B-field it changes polarity with each rotation right? The homopolar motors have the H-field from the current sweeping radially, always orthogonal to the B-field (rotationally symmetric). Wouldn't the PM need to be aligned with the axis of rotation for it to be a homopolar version?



OTOH, my asymmetry hypothesis doesn't explain why the OPs motor will start up on it's own. I would expect it to just park somewhere like a step motor.

 

[anorlunda]

Are you sure?

Not 100% sure, but look at the videos at the bottom of the Wikipedia Homopolar page. One of them looks very similar to the one in the OP.

 

[Baluncore]

The secret was hidden in the open comments by the video's author. It is a homopolar motor which is explained at Wikipedia.

I expect the deliberate imbalance causes the pintle wires to lift off from the support wires for part of the cycle. That asymmetry results in a half-wave current through the coil, synchronised with rotation.

It will start because the coil hangs with one edge down. I expect that if the pintle wires were bent, so the other edge was down, it would run in reverse, unless you end-for-end reversed the armature in the supports.

 

[DaveE]

I expect the deliberate imbalance causes the pintle wires to lift off from the support wires for part of the cycle. That asymmetry results in a half-wave current through the coil, synchronised with rotation.

It will start because the coil hangs with one edge down. I expect that if the pintle wires were bent, so the other edge was down, it would run in reverse, unless you end-for-end reversed the armature in the supports.

So your vote is for a "conventional" DC motor with a subtle commutator, not a homopolar motor?

So it might start if the gravitational equilibrium (before start) is different from the electrical equilibrium. Then the rotational momentum generated by the starting torque keeps it moving through both of the equilibrium positions?

 

[Baluncore]

So your vote is for a "conventional" DC motor with a subtle commutator, not a homopolar motor?

So it might start if the gravitational equilibrium (before start) is different from the electrical equilibrium. ...

Yes and yes, but I have not analysed it fully. Notice how the pintle wires are carefully bent, so it is NOT balanced, with one edge of the coil initially hanging down.

 

[DaveE]

OK, It felt inevitable around post #2 or #3... I had to make a kludge version of one. BTW winding that rotor without magnet wire and coil forms was a PIA.

This thing is mechanically finicky. Sometimes it just moves and parks like a step motor. Often it will oscillate like a pendulum, for longer than I would have expected. Sometimes it starts spinning on its own, but not usually. It works better if the magnet is displaced a bit from the center of the coil, but as it spins it will slowly move the magnet to the center and usually stop (No, I didn't tape/glue it, that would be extra work). The wire I used, based on the OP video was mechanically sketchy as an axel. I think it was AWG#22 or AWG#24. I can pretty much guarantee that that YouTube guy did, like, 27 takes and posted the best one.

@Baluncore wins, it's a "bumpy commutator". This thing acts just like you would expect from a conventional DC motor perspective.

My PS (not a battery) wasn't stiff enough, but I don't have a great desire to fix that, like with a BFC (Big F@#%$#@ Capacitor), I think I'm done here.

This is a trace of when it was spinning like a motor (mine is a crappy motor, BTW). The top trace is current at 0.1A/div the bottom trace is PS voltage at 5V/div.

I guess it's too much to ask the people who make YouTube videos to do some basic instrumentation?

 

[sophiecentaur]

@Baluncore wins, it's a "bumpy commutator".

A 'monotator'? More of a buzzer circuit than a motor.

 

[k.udhay]

Thank you all for your answers! Particularly DaveE for a detailed discussion and observations based on experimentation. I also plan to do this and will share my findings after.

 

[k.udhay]

@DaveE - In the mean time I also found this page which is very similar to that of the video in OP:

https://www.wikihow.com/Build-a-Motor
The potential different step (perhaps) is to coat one side of the wire using permanent marker:

May be the permanent marker ink acts as a thin but effective insulator? If so, the armature will rotate like a two-stroke engine - A power stroke for first 180 deg. and an idle (not the best word) stroke for the rest.

 

[artis]

@k.udhay it seems you almost found the right answer yourself before I saw this thread.

I have done a motor like that , the simple trick is to use enamel wire where you simply scrape off just one side of the wire with a knife and leave the enamel lack intact on the other side.
That jumping that @DaveE and @Baluncore talked about is more of a side effect of the hand wound imperfect rotor than the design goal or mode of operation.

In the end it is just another DC motor with a crude and simple commutator.
In fact I would say that all DC motors are commutated, because essentially there is no such thing as a DC motor/generator, their all AC machines.
Except for one machine and one machine only! And that is the @anorlunda introduced Homopolar motor/generator aka Faraday disc, only this motor talked about here is not a homopolar motor, far from it.

A homopolar motor/generator will always have two current contact, where one is located/can be located at or near the rotational axis while the other has to be located at the periphery, or furthest point from the center.
The theoretical reason for this is that a homopolar motor/generator can be modeled as a continuously expanding loop that expands through a magnetic field, thereby the number of field lines enclosed by the loop constantly increases providing a constant output voltage and current, the only true DC machine in existence. Also the very first electrical generator/motor discovered by Faraday during his experiments back in 1830's
The simplest example is a rectangle loop where one side is not stationary but allowed to move while still maintaining electrical contact.
The rotary disc is just a different geometrical form of the same loop.
The same mechanism of how the large navy's railguns work, the projectile is the moving part of that rectangular like current loop.

 

[sophiecentaur]

In fact I would say that all DC motors are commutated,

I can't really disagree with that except that what we really have here is a linear (unidirectional) motor with a form of crank (the sloppy structure) which translates the linear oscillation into circular. The 'armature' is more of a flywheel here and needs to be tiffled to work.
These "what is it really?" posts are never very satisfactory.