Coil winder Ideas and traverse devices

[Woodenspoke]

I do understand the theory behind your math unfortunately my math is not up to the challenge. A DXF file of a proper cam would be my ideal answer. From that I could make several dozen different sizes using a CNC mill and some 1/4" aluminum plate. I will assume the cam only changes in size not shape for a given distance. This also many not be true.I think there is one factor though missing from your math and that is the thickness of the magnet wire including the coating. Probably why John H is still holding that cam in his hands and hand winding. Now I am assuming this has to be added to the calculations since a thicker wire will require a smaller cam. This is because the wire will lay down faster at larger sizes given everything is a constant and the reverse will be true when going from lets say a 42 to a 43 wire. Again this is all assumptions I am making from posts on this thread.

But you did explain why the v shape at the top will work I think the correct figure is .5:1 for each wind layer. I think the 1:2 which is also correct had thrown me off. Sometimes I have to be hit on the head a few times before it sinks in. Cam makes 1/2 turn per coil rotation but I kept seeing 1 full turn per rotation of the coil.

I guess if I could devise a winder that rotated the cam 1/2 a turn for every turn of the bobbin all I would need is a different cam for each pickup and wire size. A mechanical reduction should be easily, the cam design for me is the hard part.

I have thought about using my small CNC mill as the bobbing winder and cam and programing the z axis to change direction at the proper time. But the CNC does not have auto motor speed control so this would be much harder to do using a manual speed knob. The the ideal situation would be to kill the motor at a certain count again another thing which is missing.

I have ordered some basic parts to start building a winder since the CNC winder is a good idea but has too many variables for now and I am not ready to start putting in motor and rotational count controls since they may exceed the cost of building a winder from scratch. I also like the idea it can me small enough to be put away until needed.

Well again thanks for putting the cam into perspective for me I finally have the operation down just my math skills are not up to the task. If there is a formula handy that may help me.

[Wrnchbndr]

I think you are thinking too hard and you don't need to really calculate every degree. Lets say your desired stroke is 0.5". There are 180 degrees in half a circle. 0.5"180=0.00277... . To cut out a heart shaped cam you could probably get away with just plotting out ever ten degrees around your circle and then connect the dots. 10x.0027=about .03 so for every ten degrees around the circle you'd decrease the radius by 0.030 or 30 thousanths of an inch until you get to the opposite side and then you reverse the process.

[Woodenspoke]

Sounds easier now. But I go back to my original question about the thickness of the magnet wire. Should you deduct the full thickness from your calculations given you are looking to end the wind 1/2 the thickness of the wire gage away from the end of the bobbin. I am using the word bobbin to include flatwork.

Just off the cuff I believe AWG 42 is about .0056 or.006 inches wide with a coating but if this is wrong this does not matter just the principle about deducting this dimension from your calculations. EX: .5 - .006 = 0.494/180=.00274 then 10 x .00274= .274. OK so I can do some basic math.

I should be able to do this arch in Solidworks down to .027 without rounding it off and export a dxf file. Yes you can duplicate one half mirror it and combine the two ends. From what you are showing with the math I can just resize the finished cam to accommodate wider bobbins. Make one 101% and another 102% and so on untill you get tired of making duplicates.

I am surprised no one has posted any dxf files for this cam if the math is so simple. Somehow it seems too easy to be true or is it just the big secret no one will reveal. I am going to wait and see what other people have to add to this discussion before I start my work.

[John_H]

I do understand the theory behind your math unfortunately my math is not up to the challenge. A DXF file of a proper cam would be my ideal answer. From that I could make several dozen different sizes using a CNC mill and some 1/4" aluminum plate. I will assume the cam only changes in size not shape for a given distance. This also many not be true.I think there is one factor though missing from your math and that is the thickness of the magnet wire including the coating. Probably why John H is still holding that cam in his hands and hand winding. Now I am assuming this has to be added to the calculations since a thicker wire will require a smaller cam. This is because the wire will lay down faster at larger sizes given everything is a constant and the reverse will be true when going from lets say a 42 to a 43 wire. Again this is all assumptions I am making from posts on this thread.

But you did explain why the v shape at the top will work I think the correct figure is .5:1 for each wind layer. I think the 1:2 which is also correct had thrown me off. Sometimes I have to be hit on the head a few times before it sinks in. Cam makes 1/2 turn per coil rotation but I kept seeing 1 full turn per rotation of the coil.

I guess if I could devise a winder that rotated the cam 1/2 a turn for every turn of the bobbin all I would need is a different cam for each pickup and wire size. A mechanical reduction should be easily, the cam design for me is the hard part.

I have thought about using my small CNC mill as the bobbing winder and cam and programing the z axis to change direction at the proper time. But the CNC does not have auto motor speed control so this would be much harder to do using a manual speed knob. The the ideal situation would be to kill the motor at a certain count again another thing which is missing.

I have ordered some basic parts to start building a winder since the CNC winder is a good idea but has too many variables for now and I am not ready to start putting in motor and rotational count controls since they may exceed the cost of building a winder from scratch. I also like the idea it can me small enough to be put away until needed.

Well again thanks for putting the cam into perspective for me I finally have the operation down just my math skills are not up to the task. If there is a formula handy that may help me.

I'm just a dumb old carpenter, but I think you're making things too difficult.

I re-read my original post. Although it was hypothetical situation that would make for a poor strat pickup the math works.

I was trying to illustrate the ratio of the cam speed to the bobbin speed. A full rotation of the cam will guide the wire back and forth (two layers). The ratio was 336:1 from the bobbin to the cam. 180 degrees cam rotation = 168 turns = 1 layer of wire. If you wanted 100TPL The ratio would be 200:1 .

I could easily enough build another winder with a mechanically driven auto traverse, and maybe someday will. It's just not very high on the list.

Good luck with your winder. Post some pictures when you get something together.




wrench; That was a great description on the cam.

[Wrnchbndr]

My spin on the magnetwire is to just do a number of physical measurement of the wire itself with a micrometer. Remember that there is an actual thickness to the insulation and the average thickness will change by manuacturer because of the different processes they may employ to apply the lacquer or enamel. I'd take the largest of these measurements. You'll have a bobbin width measurement to use based on your bobbin design or existing pickup. If your wire is for example .0028 (bare clean 43 gauge is .0024) and your intended bobbin is 0.55", You divide .55 by .0028 and get 196. This means that your transverse device needs to do one stroke in one direction for every 196 rotations of the bobbin. If you're using this heart-shaped cam, it will yield a a full back on forth movement in one complete revolution which is actually two layers of wire -- one layer to the right and another to the left as it returns to the starting point. Two layers is 2x196 widths of wire so the ratio is 392 revolutions of the bobbing to 1 complete revolution of the transverse cam. Its the max and minimum radius of your heart shaped cam that needs to match the width of your bobbin. What is unfortunate is that this whole concept limits you to doing the same bobbin width if you're concerned about laying the wire neatly side by side. Not only do you need a cam of the proper size but you'll need to change the ration to other bobbins of different width. This is something that I did not understand until I though about your question. For another size pickup, if you maintain the ratio and only change the size of the cam the wire will not lay properly.

[Woodenspoke]

Sounds like I should have just kept to my original idea to hand wind. First I was thinking of only using one wire size AWG 42 and single poly coating and doing three different pickups. All of which will be a different width. So if a coated wire is for argument sake is .006 then we divide by the distance between the width to get an even coil wrap lets say 1/2 " so .5 /.006 = 83 wraps per layer.

OK I got that part and getting the shape of the arch math; now where do you combine the two???

Then we (royal we) have spoken of an L shaped wire guide which is adjustable. It will also change the distance and speed at which the arm moves and thus the wrap count depending on where the follower for the cam pivot's on this wire guide. Assuming this guide has a fixed or adjustable pivot point. Just theorizing here since we all by now I have no clue.

My quick story is I have been a woodworker for 30 years and over the past three years added metal working tools to my shop. When I was young even worked with a few top builders but decided it was better as a hobby than as a career of you like eating (this was in NYC in the 70's). So now I am selling parts for guitars on eBay as well as doing some manufacturing for one of the two major guitar parts suppliers trying to slowly build up a business. So I have been around the block way too many times to just give up on automation just yet.

Now I will build a winder and hand wind for my first go round but I find I pickup things easily so I expect to wind less garbage than one may expect from a newbie. But adding any automation has been the most difficult concept. I figured someone had a preset formula which takes in all these variables, but it does not look that way. If there is something out there its a closely guarded secret. Or using a cam is really a stop gap before spending the 10K plus on a dedicated CNC machine. I have also seen the homemade winder pages as has everyone else has but no secrets are provided. Besides it seems they are more interested in the winder build than the pickup, but I am sure I am wrong as I have been so far.

Honestly it was easier to retrofit a small mill for CNC than to design this seemingly simple device. But I will keep going maybe someone who has lived this retrofit will chime in here and let everyone know how these issues were overcome. Math and design. I could go the experimental route but I really don't have that kind of time frame to tinker.

[John_H]

I'd take the largest of these measurements.

I'd take the smaller. Depending on how tight you wind the wire does get thinner under tension.

[John_H]

Then we (royal we) have spoken of an L shaped wire guide which is adjustable. It will also change the distance and speed at which the arm moves and thus the wrap count depending on where the follower for the cam pivot's on this wire guide. Assuming this guide has a fixed or adjustable pivot point. Just theorizing here since we all by now I have no clue.

I'm not sure of what you mean by "(royal we)", but I'll take it in stride. If I came off as condescending in an earlier post it was unintentional, and I'm sorry. I was trying to help. That being said maybe this will help.

Changing the length of the rocker arm that guides the wire will change the distance, and speed that the wire travels. The only way to change the wrap count is to change the ratio from the bobbin to the cam. I tried to illustrate that in an earlier post. The example I gave you was for a level even wind which is probably not where you'll find your best tone.

After some experimentation you'll figure out how many turns per layer sounds best to you on a 42g 8k humbucker. When you factor in all of the variables such as magnet type, wire insulation type/build, wire tension, baseplate, keeper bar, stagger, slug material, and pole screw material the possibilities are exponential in number. That is what differentiates one builders sound from another That's only one pickup design.

Some of the guys that contribute on this board who by the way are among the best in the world have built countless prototypes, and they know what works for them be it vintage replicas, or contemporary designs. That's info that won't be thrown freely about. Most of the good info you'll find will be in the form of a subtle hint, or someone will tell you right out when something wont work, and why.

I like yourself am pretty new to all of this. I started winding only six months ago. The six months previous to that I was reading all that I could absorb on the subject. It was only after I started winding that many things actually became clear to me.

Now I will build a winder and hand wind for my first go round but I find I pickup things easily so I expect to wind less garbage than one may expect from a newbie. But adding any automation has been the most difficult concept. I figured someone had a preset formula which takes in all these variables, but it does not look that way. If there is something out there its a closely guarded secret.

When I first started I built a couple from SM kits. After that I asked friends if they had junk pickups that I could have. Those that were worth rewinding usually ended up back in the hands of the donors rebuilt for feedback. It doesn't take too long to get the hang of it.

No secrets here. I'd be happy to share my ideas with you. It seemed you were pointed in the direction of a CNC solution so I didn't expand.

[Woodenspoke]

Royal we explanation: I rebuild a room top to bottom with no help from anyone and my wife comes in with a friend and says look at what we did. "The royal we" I call it. Doesn't translate well. Your idea not mine is what it means.

Again the CNC is limited without additional modifications to run cutter head motor speed and shutoff. The program can count winds and kill the motor and adjust the speed if I had the modifications for the mill, even scatter wind a particular pattern. I think right now its cheaper to spend $100 and build a manual winder than modify my CNC machine. Besides I do use it to manufacture items so I rather not screw it up just yet.

If I do get it right as we all hope we will; the additional money right now would be better spent prototyping than buying equipment. I have come to realize the endless variations available with materials and winding techniques as you stated. It can be pretty daunting when you start searching the web for metals which are better gauss conductors. I will just start off with something basic and see where it goes. I expect the same learning curve on pickups but compressed into a much shorter time frame than the 20 or so years it took to become a good woodworker.

The cam idea would be minimal cost and I always have scraps of 1/4" aluminum in my shop, so other than adding a cam drive system, my other costs are zero, Plus cutting them on a cnc mill would be what I am set up for cutting now. So it seemed at the time like it was going to be an easy thing to do since we have forums and I'll just ask how they did it??? ha

I guess I have joined the ranks of the pickup obsessed. I also ordered two kits from SM and I do have some old pickups to play with. I have an old velvet Hammer Red rodes which is in need of a rewind and an unknowm humbucker which can be canabilized. First I have to build the winder.

[R.G.]

Check this page out Coil Winder

Go to the fourth, six and seventh and pages and see a heart shaped cam.

This winder spins the wire around a stationary bobbin. I've never quite figured out how that works.

It took a minute. the wire comes into the center of the wire distribution arm, goes out to the end of the arm, and is let out of the end of the arm toward the bobbin. The end of the arm plays out wire around the bobbin. The bobbin is not fixed. It moves in and out along its axis, but does not spin.

The motion is the same as if you held a bobbin in your left hand and wound wire onto it with your right hand.

That diagram of various cams to get various degrees of scatter, the complicated machinery, and the picture of the cam being machined is one of the strongest arguments I have seen yet for doing an all-software winder.

[David Schwab]

The bobbin is not fixed. It moves in and out along its axis, but does not spin.

That part I figured out.

The motion is the same as if you held a bobbin in your left hand and wound wire onto it with your right hand.

I can't figure out how the wire doesn't wrap around the axel that the wire distribution arm is spinning on. That's the part I can't visualize.

That diagram of various cams to get various degrees of scatter, the complicated machinery, and the picture of the cam being machined is one of the strongest arguments I have seen yet for doing an all-software winder.

I agree. It's an amazing machine, but way too complex. I'd think it would be hard to change it's winding pattern.

[R.G.]

That part I figured out.
I can't figure out how the wire doesn't wrap around the axel that the wire distribution arm is spinning on. That's the part I can't visualize.

Ah. The axle is hollow. The wire goes down the hollow center of the axle for the wire arm.

[Woodenspoke]

Here is that Chinese CNC winder off ebay http://cgi.ebay.com/Fully-Automatic-...QQcmdZViewItem

Says only 40 gage minimum and looks to be of little use trying to program in any kind of pattern. But they did mention tension programing for the wind. The bobbin is stationary in a lathe style setup and the wire moves which is what you would expect, see the video. Just try returning it??????

Eureka
Light bulb just went off: I have a servo motor which I bought as a spare. I can mount the servo on the main mill head (or stationary using the rear column) with the servo axis forward toward you, throw on a bobbin attachment and you have a programmable motor for counting turns and auto stop capabilities. Forget about the main motor that was a bad idea. Then I can use the Y axis on the mill bed to move the wire. The 4th servo will drive the bobbin at no additional cost and I have an unused 4th axis on my controller card. Anyone have any gcode for mach3 LOL. I would also need some kind of stationary tension device to mount on the mill bed. I am trying but right now this seems like the solution to mechanical auto winding. Not that I have seen any indication there is a good mechanical solution.

I am still going to build that hand winder first and learn to wind bobbins but I have a backup plan now that is not too far fetched; Yippee. Still interested in seeing where this thread goes.

[David Schwab]

Ah. The axle is hollow. The wire goes down the hollow center of the axle for the wire arm.

Ah ha! That was simple...

Thanks for clearing that up.. was driving me crazy!

[Ampclutz]

Woodenspoke,

If you have a cnc machine you probably have a CAD program to do your drawings... If so, you can make almost any cam you need in just a few minutes. Don't let anyone convince you cams are difficult because they aren't difficult. Your cnc machine gives you a true advantage. You can probably make a cam from start to finish in less than an hour. That's about 10 minutes to machine it and 50 minutes to draw it! If you are quick with the drawing part make that 10 minutes to draw. That's about what it took me to make my cam.

AC