Looking for the right DC motor

[chicagopianotuner]

Hello forum.

I'm looking for a DC motor for a little project. I don't have a lot of experience with different types of motors, though I'm fairly confident I don't want a servo motor or a stepper motor. I have a disk of non-uniform weight distribution. I want to mount the disk vertically (motor axle parallel to the floor), use the motor to slowly (~10 seconds?) spin the heavy side of the disk to the top, and then cut the motor and allow the disk to oscillate like a physical pendulum for as long as possible.

The purpose of this project is to make a watch winder sort of like this one. I can control the whole thing with an arduino, but I'm not sure whether or not a https://www.amazon.com/dp/B006W2HLSY/?tag=pfamazon01-20 will have too much "friction" for the pendulum aspect to work well. Does anyone have any experience with this sort of thing?

Thanks in advance!

 

[NascentOxygen]

What radius is the disc? What weight is it? You don't want to use gearing to get the slow rotation?

Wouldn't you want to disengage the motor to allow the pendulum to oscillate freely by itself?

 

[chicagopianotuner]

Thanks for your reply. The radius of the disc is about 2". The mass of the disc is about 500g. I could use gearing, but I figured it would be easier if the gearing was built into the motor, which, I believe, is how these high torque motors work.

What do you mean when you say "disengage the motor?" If you mean kill the power to the motor, then yes. If you mean remove the disc from the motor axel, then I hadn't planned on it, but I could. I'm not sure how I'd go about doing that.

 

[CWatters]

I think I would use a very slow motor ad gearbox that takes around 120 seconds to make a revolution AND use a coarse ratchet/freewheel to turn the output. The motor will rotate until the pendulum is vertical them the pendulum will fall over the top and oscillate until the motor picks up the drive again.

 

[chicagopianotuner]

I think I would use a very slow motor ad gearbox that takes around 120 seconds to make a revolution AND use a coarse ratchet/freewheel to turn the output. The motor will rotate until the pendulum is vertical them the pendulum will fall over the top and oscillate until the motor picks up the drive again.

Thanks for your post. I'm familiar with freewheels as you've mentioned (I've built a few bicycles) but I've never seen anything small enough to work with a tiny DC motor, and I can't seem to find anything via google. Can you link to a product that you're thinking of?

For what it's worth, I don't need the motor to be in constant operation. The whole process, which will last about 1 minute for the half-revolution and then for the pendulum action to stop, will only need to happen about once per hour. I figure I can control this with the arduino.

 

[JBA]

All you need to do what CWatters suggests is a pin in the pendulum and a rotating arm on the motor drive shaft that will engage that pin to rotate the pendulum to the top before the motor shuts down.

 

[chicagopianotuner]

All you need to do what CWatters suggests is a pin in the pendulum and a rotating arm on the motor drive shaft that will engage that pin to rotate the pendulum to the top before the motor shuts down.

Ah, I see now. If I want this thing to last a long time, should I look for something more robust to allow the pendulum to rotate? For example, some sort of bearing set that attaches to the motor shaft and that I can inset in the disk? My gut says that if I just drill a hole in the acrylic disc and allow the disc to oscillate simply on that, then eventually it'll wear away as it rubs against the motor shaft.

 

[JBA]

You are correct, simply drilling the disc is definitely not the way to go. Quality watch armatures and auto winding pendulums use pointed shafts with jewel end bearings.

 

[chicagopianotuner]

You are correct, simply drilling the disc is definitely not the way to go. Quality watch armatures and auto winding pendulums use pointed shafts with jewel end bearings.

I see. Let me preface this next bit by saying that I have access to a laser cutter, so I can cut the acrylic pretty accurately.

Let's say that I cut a hole/fitting into the acrylic disc and inset http://www.vxb.com/626RS-Bearing-6x19x6-Sealed-Miniature-p/kit7465_1.htm. How do I attach the bearings to the motor shaft? Again, this is my first project involving these little DC motors so maybe I'm missing something super obvious. Thanks again for all of your help so far!

 

[JBA]

The main issue with the acrylic is that it is not really a hard and durable very low surface friction material.
Just as a thought, what about placing a hard metal double pointed shaft thru the disc center and then placing on the end of the motor shaft a small collar with a jeweled bearing inserted in its opposite end, and a second such bearing on a supporting stand on the other side of the pendulum. The collar on the shaft end might also be the mounting for the lever arm to rotate the pendulum.

 

[Rx7man]

Can you draw a little sketch of the proposed pendulum?.. How far off-center from the shaft is the center of gravity of the pendulum? How much swing angle are you going for? full turn or a couple degrees?

If you're only going for a handful of degrees, I'd use a far oversized motor and forget about the gears as they are typically rather inefficient...

Another thought is to not use a motor at all, but an electromagnet/solenoid to 'kick' the pendulum once in a while if you don't need to swing it too far.. this would eliminate gearing and a lot of friction too.
Just some of my thoughts

 

[chicagopianotuner]

Can you draw a little sketch of the proposed pendulum?.. How far off-center from the shaft is the center of gravity of the pendulum? How much swing angle are you going for? full turn or a couple degrees?

If you're only going for a handful of degrees, I'd use a far oversized motor and forget about the gears as they are typically rather inefficient...

Another thought is to not use a motor at all, but an electromagnet/solenoid to 'kick' the pendulum once in a while if you don't need to swing it too far.. this would eliminate gearing and a lot of friction too.
Just some of my thoughts

The "disc" is actually going to be a piece of PVC pipe cut to about 4" long. I'll insert a piece of acrylic cut to fit the ID of the pipe and then glue the acrylic inside the the back of the pipe. The other end of the pipe is open, and I will fill it with a piece of high density foam with a watch clasped on.

The bearings will attach (somehow??) to the acrylic, and then will fit onto the shaft of a stepper motor (I now think a stepper motor is probably best.) The arm can just attach to the motor shaft as well. I'm not sure how to fit it all on. I've attached a simple sketch of the disc that fits inside the PVC pipe. I would like the whole thing to swing as much as possible.

I'm open to suggestions.

 

[JBA]

How about trying to accomplish the same thing by taking an approximately 2 inch long piece of 3 inch nominal diameter PVC pipe and installing a disc in one end that will connect to the shaft of a continuously running slow rotating gear motor and then simply strap the watch around the pipe using the watch band; then, the rotating motion carrying the watch around one side to the other will activate the watch pendulum twice for each motor rotation. No timing, stopping and starting or tripping mechanisms required.
You might also want to put a counterbalance weight on the opposite side of the pipe from the watch to balance the watch weight during rotations

 

[chicagopianotuner]

How about trying to accomplish the same thing by taking an approximately 2 inch long piece of 3 inch nominal diameter PVC pipe and installing a disc in one end that will connect to the shaft of a continuously running slow rotating gear motor and then simply strap the watch around the pipe using the watch band; then, the rotating motion carrying the watch around one side to the other will activate the watch pendulum twice for each motor rotation. No timing, stopping and starting or tripping mechanisms required.
You might also want to put a counterbalance weight on the opposite side of the pipe from the watch to balance the watch weight during rotations

Yes, that would also work, but I like the idea of the watch oscillating back and forth - it only requires the motor to make one full revolution in order to get a lot of watch movement. With a stepper motor there really wouldn't be much difficulty in programming the movement.

The main issue I'm having now is how I attach things to the shaft of the stepper motor. The shaft isn't very long, so I need some sort of coupling, right? Is there a specific kind of coupling to use with a perfectly round shaft vs. a "flat" shaft vs. a shaft that's flat on one side but round on the other? How do the bearings fit onto the shaft/coupling? Is it just friction (because the bearings and shaft/couplings are machined precisely enough to provide a very snug fit)? Or do I need some sort of hex coupling/bearing?

Thanks again for your suggestions.

 

[Rx7man]

I'm good at seeing problems where there might not be any, but if the pendulum is acted on by gravity, and the watch is on it, wouldn't the weight in the watch accelerate at the same rate as the watch, thus not there'd be no net force on the watch's weight and no winding action.
I think just rotating watch around the axis of the watch's weights would work, or a sort of mild shaker mechanism.

How about wearing the watch? :P

 

[chicagopianotuner]

I'm good at seeing problems where there might not be any, but if the pendulum is acted on by gravity, and the watch is on it, wouldn't the weight in the watch accelerate at the same rate as the watch, thus not there'd be no net force on the watch's weight and no winding action.
I think just rotating watch around the axis of the watch's weights would work, or a sort of mild shaker mechanism.

How about wearing the watch? :P

I think you're seeing a problem where there isn't one. The argument you made could also be made no matter how the watch moves. The watch will wind if moved in the way I want to move it. All the evidence I need is that one of the top companies who makes watch winders uses this sort of mechanism, and I'd like to copy that. Yes, a shaker would work (though it might damage the watch), as would simply making the motor rotate slowly, but I'd like to make the watch oscillate. It's supposed to mimic a slightly more natural winding process.

Wearing the watch is a good solution, but sometimes it looks goofy wearing 2+ watches at once!

 

[Rx7man]

You're right about wearing 2 watches :)

When I said a shaker, well.. there's shaking, and there's SHAKING... For a mechanism, I was thinking of something like a windshield wiper assembly.. you could put your PVC pipe where the wiper arm goes, and it would rock back and forth

 

[Rx7man]

So I've been practicing autocad.. I'm getting better... and I drew up something I think would be simple and work well... You mentioned an arduino to control the motor, that could be implemented, or just a timer. The motor I have in mind for this would be a windshield wiper motor, though it's much bigger than necessary, it returns to the same point every time, which is nice.. a stepper motor would work too... I'm thinking you'd want perhaps a hall effect sensor for sensing when to give it another 'kick' located at some angle off vertical, so that if the sensor isn't getting a reading, it's time to nudge it again. The thing *I* like about this idea is that the motor doesn't have to keep moving for the pendulum to swing, reducing friction.

I left the front of the tube open, you can put weight in it if necessary, but it will need a little pivot point for the shaft and a retainer.. I don't think you need any bearings, using a small shaft 3-4mm... 3/16" will pivot plenty well enough.

How does it look? Motor mounts on the backside