Before going into an electronic solution

Actually, I plan on going into several solutions at once. Since this is for an important project (whose deadline looms nearer every day), I cannot allow myself for the time that a sequential trial an error approach might consume. So I'm working on several approaches simultaneously, and intend to incorporate them in the first model. Among them there will of course be the friction brake technique. But truth being told, I have much more faith in the tensioning counter-motor, mainly because it's something that can be controlled dynamically and which can (and will) be commanded from the electronic controller I'm incorporating into the whole thing. It's also mechanically simpler, because the spool will be directly mounted (or perhaps coupled) on it, and there will be no need for extra components such as a clutch, friction device, belt and pulley, etc... . Once set up, the control loop can be closed (if needed) by either adding a load cell, or a spring whose length will be constantly monitored in order to adjust the wire's tension.

 

These are the magnetic particle brake systems I ended up using. They allow adjustment of the torque using a current source on the units I had. They afford a wide torque range of Torque range = 0-1 oz.-in. to 300 lb.-feet. There should be similar units out there from other manufacturers. Their current sources also allow placement of an external pot and the only power to the brake is the current source with no other power required. You can also fabricate your own source.

Ron

Hi,

Looks like a great idea.

Another idea for a basement engineering level project would be to use a portable power drill, the kind that run off of those sub C cells maybe. They have a limit slip chuck on them so that when the torque becomes too great it slips. Would be funny to watch
Dont know what the dynamic range is and of course it may not be in the right range for coil winding.

To the point of just how much careful torque control is needed, most of the winding i have seen done in the past just use humans as the tensioner. They hold the wire as it winds on, that's it.

There is a wire size limit here though where you cant use tension alone. For example, with 1/4 inch round wire the wire has to be beat down as the turns are added so that it bends properly around the coil form. This is especially true with square or rectangular coil forms where the radius is small near the corners. 1/4 inch square wire has to be wound more carefully anyway so that it goes on to the form squarely. A bit more difficult, but also needs to be beat down with a rubber mallet so the turns remain flat against the previous layer.

 

I am embarrassing.

Lol! ... I'm sure you meant to say that you feel flattered, but I understood your meaning

Anyway, an important question:

I plan to apply pulse width modulation (PWM) to the motor using a couple of mosfets mounted back to back, so that the complete unrectified ac wave can reach it. Will this allow me to control the torque being generated by the motor, or will it somehow interfere with the series (red) capacitor and not work as expected?

 

I plan to apply pulse width modulation (PWM) to the motor using a couple of mosfets mounted back to back, so that the complete unrectified ac wave can reach it.

I would be very suspicious of that working on a PSC motor as those appear to be.
And as you pointed out in a previous post, these motors typically used in ceiling fans often require full voltage at first before rpm control.
Max.

 

I would be very suspicious of that working on a PSC motor as those appear to be.
And as you pointed out in a previous post, these motors typically used in ceiling fans often require full voltage at first before rpm control.
Max.

Yeap... what you say makes perfect sense. Besides, I don't think I'll need to adjust the motor's torque once I set it to an acceptable level. So I think I'll keep things as simple as possible and play around with the cap's values instead, as Danko suggested.

But I intend to turn the thing on and off through the controller using a triac, though... I'm pretty sure that can't hurt.

 

Lol! ... I'm sure you meant to say that you feel flattered, but I understood your meaning ...

Exactly.

...I plan to apply pulse width modulation (PWM) to the motor using a couple of mosfets mounted back to back, so that the complete unrectified ac wave can reach it. Will this allow me to control the torque being generated by the motor, or will it somehow interfere with the series (red) capacitor and not work as expected?

It works. Here is file for PWM 20kHz:

 

Exactly.

It works. Here is file for PWM 20kHz:

Wow... that's a very interesting model. I'm gonna play with it a bit, see how much I can learn. Thank you!

 

OK. This is your circuit: View attachment 151167Black capacitor is 1.3uF. Red capacitor you will select for tension, you need.
1.3uF - it is value for full speed of motor, but in your case motor is stopped.
For this mode max torque will with value of black capacitor about 13uF. Torque is almost proportional to value of C_black.
Then you will select C_red for torque you need.

Danko, I couldn't find 13 µF caps where I am. Standard available values are 10µF and 15µF. Is it ok if I use a 15 µF cap instead? Or should I connect a 10µF and a 2µF cap in parallel?

 

Danko, I couldn't find 13 µF caps where I am. Standard available values are 10µF and 15µF. Is it ok if I use a 15 µF cap instead? Or should I connect a 10µF and a 2µF cap in parallel?

13uF is very approximate value. You will need to specify it on real motor.
It is why you need to fabricate capacitor bank. This appliance will useful right now and for all your rest life.
Did you receive your motor?

 

13uF is very approximate value. You will need to specify it on real motor.
It is why you need to fabricate capacitor bank. This appliance will useful right now and for all your rest life.
Did you receive your motor?

I'll be getting it next week. I'm glad you've told me about the need for a capacitor bank. I'll be buying several values and experiment with them.
I'll start using a 10uF with a 1uF and then tweak it a little until I'm satisfied. The standard available values here are 1, 1.3, 2, 5, 6, 10, 15, and up in increments of 5uF

I just want to make sure I don't burn the thing the first time I set it up.

Question, can this setup be also controlled using a triac with phase angle control instead of a pwm'd mosfet? I've built that sort of thing already in the past (along with a zero-crossing detector) and it would be easy for me to incorporate it in my design.

 

How to select capacitor (black one):
1. Mechanically stop shaft of motor.
2. Connect capacitor bank instead of black capacitor.
3. In series with every wind connect ammeter.
4. Power motor in series with red capacitor 0.5uF.
5. Change value of capacitor bank for equal or almost equal readings of ammeters.

 

Question, can this setup be also controlled using a triac with phase angle control instead of a pwm'd mosfet? I've built that sort of thing already in the past (along with a zero-crossing detector) and it would be easy for me to incorporate it in my design.

I will think about it.
EDIT: And you can not burn motor with red capacitor in series. And check how warm will your motor in experiment.

 

I'm at my wit's end. I'm currently designer a miniature coil winder, and I need a way to maintain tension on the wire while it is doing its work. The first thing that came to mind was to buy an ordinary BLDC cooling fan and adapt it to the machine, mounting it on the spool and then make it gently counteract the motion generated by the step motor that I placed on the winder.

Then things got complicated. After breaking the fan apart, I found that it's using a deeply integrated special purpose control chip (U1) that uses an external hall sensor (U2) to monitor position (as it should) and that cuts power to the wingdings if it senses that no motion is taking place, which defeats the whole plan that I had in mind. And I can't think of a way to bypass that condition.

View attachment 151049​

So I ordered a dc motor from sparkfun, and it should be delivered in the next few days. But I doubt that it's going to work the way I want it to, mainly because its reduction gearbox will maybe add too much torque resistance for what I need. I tried to order a motor with a 1:10 gearbox, but ended buying one with a 1:18 because they ran out of stock of the former. The other motors listed in their site have output shafts of 3mm dia and look to flimsy for my application. The motor I ordered has a 6mm dia shaft, which feels about right for my purpose. My main concerns are that maybe the motor will produce too much torque (even though I plan to PWM it) due to its gearbox, or that it might overheat because it will be slowly rotating backwards while I apply power for it to run forward.

Plan B would be to use a passive system by simply adding a friction clutch that could be adjusted with a spring and a screw, but I don't like it because it would leave too many variables out of my control; such as losing tension when the winder stops, or wide swings in tension when it changes speed.

The coils will be 10mm id x 20mm od x 5mm thick using 0.010" (I think it's #30 gauge) magnet wire. And I'll be using a couple of small step motors to control the winder and the wire guide. I've already designed and tested the main mechanism and the motion controller and it's working as it should be. The only thing left for me to do is find a way to control the tension in the wire.

I really need to make this thing work ASAP, or I'll smash my face into a deadline ... any suggestions?

If you want to get really tricky, you could use a torque transducer in a feedback loop, but they don't come too cheaply.

 

Question, can this setup be also controlled using a triac with phase angle control instead of a pwm'd mosfet?

This simulation demonstrates impossibility of control PSC motor by changing conduction angle.
It works properly first 10ms only.

 

This simulation demonstrates impossibility of control PSC motor by changing conduction angle.
It works properly first 10ms only.
View attachment 151336

Excellent demonstration. You have just saved me a whole lot of work and possible headaches. Thank you.

I'm going to stick now to the capacitor bank. And if I need to refine things further, then I'll dip into the PWM mosfet technique.

 

The goodies are here!

​

Here's a pic of the small induction motor's plate, it seems that its very low torque (1.8 oz-in) is just what I need, since the step motor that will be doing the actual winding is much stronger (45 N-cm).

​

 

Congratulations! I am in waiting of next steps

 

If you want to get really tricky, you could use a torque transducer in a feedback loop, but they don't come too cheaply.

Thanks for the suggestion, but I'm not trying to dig in too deeply here. I just want this thing to be practical and reliable. But if things get complicated then I'll give it serious consideration.

 

Danko, I found the motor's data page, and it says that it's "Impedance protected to prevent current overload (motor can be stalled indefinitely without overheating)"

Question, is the extra capacitor still needed in the circuit to obtain the rated torque?

 

I've just tested the motor as it is, with the recommended capacitor value of 1.3uF, and it worked just fine. I stalled it with my fingers (it gives pretty good torque, just about what I was expecting) and nothing overheated, or made sparks, or spewed smoke, or blew up in my face ... thankfully