Coil winders Ideas

[Ampclutz]

David,
RG,

The cam actually serves three important functions. It very effectively linearizes the traverse feed. And more importantly it allows the spindle motor to run at a constant speed. That is more important than you suspect at this time! For the spindle and traverse to remain in sync (under full computer control) the spindle has to slow down or even stop each time the traverse motor changes direction. What a waste of time that is! Also, if the stepper motor looses a few steps, and that is not so uncommon when running the motors near their highest speed, the winding will still be centered on the bobbin.

In my real world the stepper motors smooth out at real winding speeds, like anything over 100 RPM. That is slow on my machine... My machine runs smoothly at 800 RPM. If I had a servo motor for the spindle I could easily get 3000 to 4000 RPM. Probably much more but that is up in my scary zone!

Do you have a CAD drawing system available? If so cams are extremely easy to draw once a few basic concepts are understood. I must admit that I cheat! I use my CNC machines to cut the cams. It just takes a very few minutes. I guess that is why a CNC pickup winder is the easiest solution for me... This may sound a bit strange but the CNC winder is actually cheaper for me to build than a manual one because the counters and stuff like that cost so much. It's all built into the computer and its keyboard. Dumpster grade computers work just fine.

Don't fear cams. They are your best friend.

I can't do computer simulation. I don't know how... It would be much easier for me to just video my machine winding a coil. But, alas, I don't have any pickup parts right now. So, nothing to wind...

AC

[R.G.]

I have a long history with cams in coil winders. We once had a full-bore manual winder with a full set of all possible gears and cams in the power supplies lab.

I ... hated ... them.

You're right. Cams are not hard, they're trivial to set up. I had to do cam layouts as part of the class projects in my first semester in college. This was back when mechanical drawing was done with T squares and triangles. Lots of manual work but the developments were not hard.

I'm fully aware of the issues of traverse versus winding speed, and of the mechanical issues with instant reverse at speed.

Let me put it to you another way. The only reason to have a machine be FAST! is that you want that coil done right now. If you have a number of coils to wind, that gets worse.

But if the machine is autonomous, can take care of itself, and it's simple and cheap to have more than one machine, the issue of production rate goes away, as you can pipeline them. If you're not making a living winding coils, the time a coil takes doesn't really matter if you are not tied to the machine while it does its thing.

My main gripe with cams is that they are fixed. You have the same profile, all the time. I can hypothesize a situation where I wanted to wind a coil with a different degree of scatter at different depths in the coil. Or, I might want to switch to a sectional winding in the middle. (Hey, anyone ever section-wind a single coil????) These things are impossible with a cam. It's a fixed mechanical part. If there is nothing fixed, only a program feeding the motion control, the winding can be anything you can figure out how to tell it to do, and the mechanical get even simpler.

The simplest situation is one I could not find a cheap source for. But I found the deluxe commmercial ones. There is at least one company that sells voice coil actuators. These are like a power analog meter movement. They will have dynamics that far exceed the nimbleness of a stepper. They may not be able to follow a high power cam mechanism (like, say the overhead valve cams in a car) but they will certainly be able to move wire from side to side and reverse on a dime.

But my objectives when I started this included the idea that a winder could be build using (mainly at least) bought motors as the precision mechanical parts and use computer control to do the interactions that made it wind.

I still think it can get there.

[Ampclutz]

RG,

Oh! You mean like the parts inside a hard drive for positioning the heads!

Back to the cams for another moment... Stepper drives make that a whole new ball game. You do realize that you don't actually have to turn the cam a whole turn... Ever... Thus making a very simple linear actuator. So, if you take a regular stepper motor running in half step mode (400 steps per revolution), reduce that by a reduction of 5 so you get 2000 steps per revolution. Then use half of a heart shaped cam that gives a total of 1 inch travel. Therefore, each step will move the traverse exactly 0.001 inches. Stepper motors are really good at going forward and backward at the speeds you are talking about. So, you can put that wind anywhere you want to any time you want it there. The throw of the cam and the reduction ratios can be tweaked any way you want.

I think it's the word "cam" that has you put off. Try using "stepper motor controlled rotary to linear actuator" in place of "cam". Please explain what that "stepper motor controlled rotary to linear actuator" won't do that you want it to do. Then I'll show you how to make it do it. All you have to do is tell the software how big each step is (each step would be the same size with a heart shaped cam) and it will convert it into whatever linear unit you ask it to use. Examples are inches, millimeters, wire diameters, big toe joint lengths, bobbin heights. Then tell the program EXACTLY what you want.

I hope this helps,

AC

[R.G.]

Oh! You mean like the parts inside a hard drive for positioning the heads!

Yes. There is a company that makes that mechanism as a bolt-in kind of actuator. Alas, it's quite expensive, something like a kBuck.

Back to the cams for another moment... Stepper drives make that a whole new ball game. You do realize that you don't actually have to turn the cam a whole turn... Ever... Thus making a very simple linear actuator.

I do realize that, yes. I'm an EE, not an ME, but I worked alongside the MEs on a lot of our manufacturing equipment. Mechanics is not a mystery to me, nor is mechanical design. Nor is computer controlled mechanics, as most of our manufacturing equipment was one-off computer controlled designs if not outright industrial robots. I've aware of if not actively playing the stepper motor ball game for a long time.

I'm not missing the simple issue of a cam as a linear actuator and a linearizer of rotary motion in the face of rotational mechanics. I'm past that, down the road of how fast you can reverse the traverse at the end of travel, and what that does to the linkage - mechanical or computational - to the rotating coil.

There is an issue with simple mechanisms like a crank as a rotary-to-linear translator, that being the sine wave velocity profile, which both you and I have mentioned. What's needed is a linear velocity profile and an instant reverse.

A cam can do that if properly designed because it avoids the issue of the motor's inertia in the reverse, and so you can use a big enough motor to simply overpower the traverse and force the reverse to be effectively instantaneous in a mechanical sense.

The issue with a cam is that once you build one, you've built it. If you want to change it, you build another one. You mention having CNC machinery at your disposal to build cams. That's cool. But having advanced capability at your disposal limits you just like not having a capability. Consider the task of getting a row of metal fence posts to the same height in rocky ground. The guy who has only a posthole digger will view that differently from the guy who has a plasma torch.

My point is that it is probably possible to make a winder out of hardware store bought stuff if you buy the precision mechanics in motors and do as little construction of mechanisms as possible. This makes a winder available to the guy who has a drill press, hacksaw, and file as well as to someone with a lathe and milling machine. Think of it as doing the task with one hand tied behind you. Because if you can, lots of people who are not as well equipped as you can also make winders.

So, if you take a regular stepper motor running in half step mode (400 steps per revolution), reduce that by a reduction of 5 so you get 2000 steps per revolution. Then use half of a heart shaped cam that gives a total of 1 inch travel. Therefore, each step will move the traverse exactly 0.001 inches. Stepper motors are really good at going forward and backward at the speeds you are talking about.

Yep, I know that. And if they're that good at reversing, then let's compute the fastest speed that they can reverse, taking into account motor inertia an mechanical load inertia, and use that to figure the upper rotational speed of the coil drive motor so we can ensure no-step-left-behind linkage between the two for a coordinated wind.

What you haven't seen is my dinking with the math on motor inertia and how fast a forward-step-to-reverse-step can be followed by a stepper. There is a stepping speed where you lose a step or two doing that. The faster that can be done, the faster you can spin your coil and be sure that you don't miss steps on the edges of the traverse.

As I said, a cam can sidestep that by letting the motor's inertia not reverse. But in doing that you lose the advantage of being able to computationally decide where the reverses are and the "gain" of how much traverse a step is. They're fixed in the mechanics. So you have to change the cam, or change the mechanical gain of the cam. You absolutely can use a cam as a linearizer and never have it turn a full turn. But in doing so, you're right back at the motor inertia problem, only now you have to reverse the cam as well as the motor inertia at the edge of a traverse. That lowers the stepping speed at which you can reverse the stepper.

So, you can put that wind anywhere you want to any time you want it there. The throw of the cam and the reduction ratios can be tweaked any way you want.

As I said, yes, that's possible. But you have to either change cams or reduction ratios. How would you do that so the computer changes all that - no mechanical dinking involved? Think of this as a design challenge. How would you let the computer do any reduction ratio you like, on the fly?

I think it's the word "cam" that has you put off. Try using "stepper motor controlled rotary to linear actuator" in place of "cam". Please explain what that "stepper motor controlled rotary to linear actuator" won't do that you want it to do. Then I'll show you how to make it do it. All you have to do is tell the software how big each step is (each step would be the same size with a heart shaped cam) and it will convert it into whatever linear unit you ask it to use. Examples are inches, millimeters, wire diameters, big toe joint lengths, bobbin heights. Then tell the program EXACTLY what you want.

Been there, done that, got the T shirt. My issue - make the stepper motor make the heart shaped cam not needed. I can make cams, have made cams. Don't like it, because in applications which change, you typically don't make only one cam. By being clever with how you run the stepper, you should be able to make the stepper speed up and slow down so the computer linearizes the movement, not the cam. In effect, you construct the cam inside the computer for each wind. That's the same computational load as setting up the CNC code to tell a machine how to cut the cam.

It's not that (a) I don't know what a cam is (b) don't know what cams can do (c) can't visualize how to apply a cam (d) can't make a cam. Cams are good, cams are clever, cams can do things that other mechanisms can't. But what I was trying to do was dumb down the mechanics so an average Joe with a hacksaw, file, screwdriver and some wiring could make a winder out of it, not make a winder that can finish a coil in 30 seconds.

I hope this helps,

Not yet, but I think we might get there. Keep em' coming. Try to think of stuff you can buy from a hardware store. OK, OK, Small Parts Inc for under $50. How's that for a challenge?

How about a rack-and-pinion? Off the shelf, $30. Spur on the stepper, traverse on the rack.

[Ampclutz]

RG,

Cool.

How much resolution do you need to meet your goals?

Although I have never seen them, I have heard there are linear actuators used in some of the newer automobiles. Somewhere in the emission controls I think... Might just be worth looking into. If so there are probably junkyards full of them.

OK. Forget all about the cam. The rack and pinion is too expensive and has too much inertia. That leaves the cable drive as a workable option. Think of old radio tuning devices. Linear, very low inertia, easy to understand, may be able to just use the motor shaft itself... Can you get any cheaper than a piece of string?

Maybe this one is a keeper?

AC

[R.G.]

It's almost a keeper. I've done string and pulley mechanisms before. Too much stretch. The reason I mentioned "timing belt" back when I was describing how to build a rolling cart to hold the traverse was because I have bad experiences with strings and steel cable drives.

But the timing belt... no stretch, relatively self damping, long lasting, as well as low inertia, non slipping and cheap.

What I had reference to in that previous post was here: Small Parts Inc timing belt pulleys
You can get 1/4" bore timing belt pulleys for $2.30 to $3.90 depending on diameter. Cheaper if you cannibalize them from a printer, probably. You need two same-diameter pulleys. Mount to a plate or board one stepper motor and one dummy shaft. The dummy shaft is there to allow you to tension the belt.

Now you have a belt drive that runs the belt left/right 1/200 the circumference of the timing belt pulley per step (on a 200 step motor, ignoring the fancies you can do with microstepping for the moment).

What I was doing with the pressed-in nylon pulleys was making a rolling plate that the belt drove.

However, now that I think of it, we can just clamp an eye for the wire to pass through onto the top side of the belt. Now the wire goes through the eye, the belt moves the eye side to side. The clamp-on-eye is something that can be made with stone axes, as can mounting a motor and a dummy shaft.

And we have a CNC winder with the following contents:

1 stepper motor for turning the coil
1 fitting for holding the coil on the previous stepper's shaft
1 stepper for moving the belt
2 timing belt pulleys
1 timing belt
1 idler shaft to tension the belt.
2 boards to put them on.

And thanks - you made me think of just putting an eye onto the belt itself. Under proper tension, it will hold wire accurately enough that a secondary holder/roller mechanism is not needed.

This fits what I was looking for: simple, precision can be bought in small chunks instead of having to be machined, and suitable for a home hacker.

I think I'll go make some drawings of this one.

[David Schwab]

Oh this is getting good! I'm going to make a pot of coffee and get some pop corn!


I like where both of you guys are taking this. I'm not building anything yet... I want to see where this is heading.

[R.G.]

OK, next step.
You need two live stepper motors and a dead one, all NEMA 23 frame size.

One live one is the coil spinner.

The other live one spins the belt drive pulley. The *dead* one is taken apart and its rotor and shaft thrown away, replaced by a length of 1/4" diameter drill rod. It serves as the bearings and mechanics to hold up the shaft for the idler pulley for the other end of the belt. You want to toss the old rotor to keep the inertia down.

Now you can mount both traverse motors, the live one and the dummy on a piece of flat plywood with a single length of aluminum angle which serves to hold the motors to the plywood and let you bolt to the faces of the motors.

For that matter, a shorter piece of aluminum (or steel, aluminum's just easier to cut) angle mounts the coil spinner on the plywood. You probably have to either mount the coil driver an inch or two above the plywood base or cut a slot in the plywood base to keep the coil form from hitting the base.

OK, you don't get this stuff for free, David - what's the longest pickup coil you'll ever wind, from end to end? We need to know that to figure how high to mount the coil spinner.

[R.G.]

I should have mentioned that "typical" NEMA 23 steppers can do enough steps per second to get up to 1200-1800 rpm. My impression is that that's fast enough to spin the coil - somebody check me on that, OK?

[David Schwab]

OK, you don't get this stuff for free, David - what's the longest pickup coil you'll ever wind, from end to end? We need to know that to figure how high to mount the coil spinner.

You mean the dimension of the bobbin? I can't imagine doing one wider than 5-6 inches (for one of these crazy 9 string bass things you see). Right now the 5 string bass pickups are 3.75". I only have plans to go to about 4.5" or so.

I should have mentioned that "typical" NEMA 23 steppers can do enough steps per second to get up to 1200-1800 rpm. My impression is that that's fast enough to spin the coil - somebody check me on that, OK?

The winder I use right now only goes to 725 RPM, and I never wind that fast by hand. Maybe half that. Of course if it was all automated, I think 1200-1800 rpm sounds good.

[Ampclutz]

RG,

The clearance above the base on my winder is 4 inches. So, I could wind an 8 inch bobbin. The wire guide is about 2.5 inches from the center of the spindle. I can adjust that dimension. I use a nylon tube for the wire guide. I have more than one guide on the traverse to make it easier to get to different size and shape bobbins. I use a ceramic fishing pole guide to collect the wire from the spool. It's smooth and slick and designed not to snag and is very durable. I could see using a fishing pole guide as a "screw eye" on a timing belt for the guide. If you need more precision than a fishing pole eye can provide, run the wire through two of them and offset them to control the position better. Those fishermen are fanatics about saving weight!

My motors are the 34 size and they start getting fussy at about 925 RPM with the bipolar chopper drives I have. I run them at about 850 RPM and they do just fine. That feels slow to me... I think the 34 size motor is way overkill on the traverse but I have them.

I mounted my spindle motor on a gusseted plywood l bracket. It makes it sturdy enough to resist the tail stock I use. I mounted the spindle motor to the L bracket through a motor body sized hole so I could mount the L bracket to the back of the motor flange. It's a little more trouble but I get an extra 3/4 inches plus the thickness of the motor flange of room at the spindle's business end. It's a big help for fat fingers!

David... RG is right! You don't get this stuff for free! So, What kind of a coil do you want to wind? Be specific. Give details. We need a real world test. Pass the popcorn.

AC

[R.G.]

Hmmm... OK, tensioners.

I've seen the notes here and it looks like the majority opinion of pros is that a felt pressure pad is the thing to do. Is there any obvious update for that?

[Ampclutz]

Hmmm... OK, tensioners.

I've seen the notes here and it looks like the majority opinion of pros is that a felt pressure pad is the thing to do. Is there any obvious update for that?

Yes there is an update... No time now but I'll be back.

AC

[R.G.]

Here's a couple of sketches. As per the first ideas, there's not much there but the motors, a belt and a couple of pulleys.

http://geofex.com/FX_images/coilwinder.pdf

The things that have to be fabbed are the bobbin holder, eyelet on the belt and the tensioner.

Of course, now the operation of the mess is A Simple Matter Of Programming.

[David Schwab]

David... RG is right! You don't get this stuff for free! So, What kind of a coil do you want to wind? Be specific. Give details. We need a real world test. Pass the popcorn

Well.. I don't think there's anything too different going on. I've been winding medium impedance humbuckers. So we are talking maybe 1500 turns on each coil. I make the bobbins. They have a steel core, and phenolic PC board flatwork. I just got some styrene I'm going to try out for the flatwork.

I wind with a back and forth motion... not too fast, not too slow. Obviously the wire is not laying down right next to last wind. I'm not sure at what point someone considers what they are hand winding as "scatter" wound, as I think it's all scattered if you are hand feeding.

I also have no idea what a "precision" wound coil would sound like in this context! I gather most people don't care for the sound of very neat windings.

I'm still experimenting with wire gauges and tune counts. I'm going to try some fairly heavy wire (36 awg) and also some "regular" high impedance pickups.

I've also did one for a 9 string guitar, wound on the hot side (in a bass pickup case) and some pretty standard PAF type humbuckers.