Coil winders Ideas

[R.G.]

Thanks for the link, that looks about right at first blush. I'll dig more into it this weekend.

As far as writing the code, we use at work a development system called TestPoint; it's basically a cheap far less functional version of LabView. In this system it is easy to develop code to control such a board, and makes it even easier to get a nice GUI to access it all. If this controller is like the Pontech one, I can see how it would be easy to write a program that would allow automation of the entire setup, allow speed control of the winder and traverse, direction control, acceleration/deceleration of the winder, and even count # of turns of the winder. Plus TestPoint allows unlimited distribution of executables. My only limit would be the time available to do this (lotsa irons in the fire). But what a fun project!!!! I'm very interested...

KEWL! I was only kidding about you being drafted to write the drivers, but I'm much too experienced at what to do when I find a gold nugget to turn your offer down. TestPoint does sound like a great place to start.

I turned up some new info on steppers. Looks like my idea about using a dead DC motor for the dummy motor on the belt traverse was better than I knew and replacing the rotor with a bare shaft a worse one. Steppers seem to need a load inertia about equal to their own rotational inertia to do a good job. A DC motor of the same frame size gives about the right load inertia for stability, and if not dead would give a DC voltage which is proportional to traverse speed for feedback if that were ever needed.

It could also be shorted to provide instant electronic braking on the traverse, so you could use the dummy DC motor to assist the stepper in stopping and reversing at the end of a traverse.

[Achiles]

How to cauculat CAM sizes?

[R.G.]

How to cauculat CAM sizes?

There are no cams in this kind of coil winder. Both the coil motor and traverse motor are driven by stepper motors under computer control. That puts the entire operation of the machine into the software.

In fact, the whole point of doing it this way is to avoid having to have any machine parts at all if possible - no cams, gears, belts, pulleys, bearings, adjustments, fitting, machining special parts, none of that. I was trying to make a coil winder with no special machining or parts.

[Achiles]

I mean the mecanical heart cam,for a 19mm hight bobin.

[R.G.]

Off topic for the thread. Try asking this in a separate thread. It'll never be noticed here.

The short answer is that the "radius" at the minimum has to be 19mm less than the radius at the maximum point, and in between it changes linearly with the angle, but there is a huge amount more to designing a good cam than knowing that. Open your own thread.

[erikbojerik]

Been away from this forum for awhile, and just found this thread. This has been a great read!

Have you guys thought any more about the interface and a choice of driver that might use USB or RS232 (no parallel ports here...)? The Linistepper concept seems inexpensive but perhaps a bit too DIY for my tastes....yet some of the off-the-shelf serial controllers don't seem to be available for under $150 or so.

[R.G.]

I'm off building my version of this. It's gotten simpler to build. I'll post stuff when it works.

The simple way to do this is to use a USB to Parallel adapter. They cost as little as $13.

For the drivers, look at Unistepper, as you've found; if that's too DIY, check out Hobby CNC (still kits) or www.ohmikron.com. You can get single-axis controllers at Ohmikron already assembled and tested. In particular, the USD0906 is $41 per channel, so $82 gets you all the controllers you need for this, preassembled. It's available as a kit at $23 (!) per channel. That's what I'm doing. If you want FAST motion, you can get the BSD0906 for bipolar motors at high currents for $69 per channel, assembled.

You may want the SDS0707, which is a parallel port interface card (mates to the end of that USB->parallel adapter) plus two channels of motor controller, power supply and all mounted and prewired on an aluminum base.

I am not associated with ohmikron, have never bought their stuff (yet), but it's what I found in looking on the web.

I think TurboCNC or EMC2 is going to be the software if I don't write my own. That's still up in the air.

[LeftyStrat]

Hi everyone, I just stumbled upon this thread. I've been collecting parts for various CNC projects, and just got the idea yesterday that a CNC pickup winder was a project I'd like to tackle.

I had originally started collecting parts for a three-axis mill using a dremel tool mainly for engraving text into die cast aluminum, until I found the dremel wasn't so good for accuracy (too much play in the spindle).

Anyway, being the cheapskate I am, I have been cannibalizing old printers and scanners, and almost everything you would need to build this is in these.

In fact, I just purchased one of those "all-in-one", copier/scanner/printer units at the local Goodwill for $7. It has multiple belts, gears, precision guide rods, etc. and even some of the hardware can be adapted for mounts.

Most of the motors I have scavenged are bipolar, but I found these plans for a simple diy controller:

http://hobbyrobotics.blogspot.com/20...ontroller.html

I plan on using an arduino board (http://www.arduino.cc/) as my controller driver, since my unit has usb, lot's of IO, and is easily programmed in C. I'm hoping to put enough smarts on the controller so that I don't even have to dedicate a computer to running it. Simply connect a laptop and run some software to download parameters, disconnect and leave it running.

Lastly, I thought of using a hall effect sensor off of an old bicycle speedometer as my counter to handle missing steps.

Anyway, it was real nice to come here, find this thread, and see R.G.'s drawing, which is pretty much exactly what I had in mind (guess I'm not crazy after all). Of course, my idea was based on the parts I had on hand. But if you get a chance, open up a scanner or all-in-one and you'll find everything you need.

It also helps to go with older model scanners, these tend to have heftier steppers and guides. If you can find an old SCSI scanner, these are the best.

[R.G.]

Hi, Lefty! Good stuff there. I didn't know about the Arduino, but first glance looks very good indeed. I'll dig further.

I think I have about junked the timing belt approach as too difficult for the average joe to make move the wire. It's not hard, but I'm shooting for something where everything can be assembled easily, and I'm afraid that the coupling to the belt would be difficult.

What I came back to was a 1/4-20 leadscrew on one stepper running a bit of bent sheet metal screwed to a drawer slide. This gets you finer resolution and more stable linear motion, I think. The parts are easy to get.

Thanks for the tip on the all-in-one and scanners as well. I have an old HP4C SCSI that might give it's life for the cause.

But I want to push on not needing the mechanical parts at all. If I can get it to two steppers, a drawer slide, and a few parts ordered from McMaster-Carr, I think there will be a lot of them built. It's even easier than destroying a sewing machine.

As for missing step sensing; I intend to print alternating light-dark sectors on paper and use reflective sensors for keeping track. The sensors cost about $1.50 each these days and it's easy to position them for quadrature sensing. Picked that up from the DIY robot guys.

[LeftyStrat]

As for missing step sensing; I intend to print alternating light-dark sectors on paper and use reflective sensors for keeping track. The sensors cost about $1.50 each these days and it's easy to position them for quadrature sensing. Picked that up from the DIY robot guys.

Thanks for the tip, I haven't heard of this, much better than hall effect, no interpolation between a full revolution. Gives me something to experiment with.

Some CNC DIY links to get the creative juices flowing:

http://www.engadget.com/2006/06/29/h...achine-part-1/

http://www.instructables.com/tag/key...?sort=COMMENTS

Twenty years ago, I programmed multimillion dollar 5-axis milling machines for Michelin Tire Corp. What is exciting for me is now the technology is so cheap, we can not only build general purpose CNC machines for pennies, we can build single task CNC machines.

I'm still dreaming about a "pick and solder" machine that could assemble a "handmade turretboard"

[R.G.]

Gee- thanks... If I had any more creative juices flowing I'd drown... 8-)

This did kick off an idea in my head. The base of a CNC machine is always tricky. You really need flat and parallel to very tight tolerances, and most small CNCs don't do this. The black-iron pipe thing is cute, but seriously non-planar and non-square, if very cheap.

If you sign up with enco they send you monthly free-UPS-shipping codes. Any order over $50 and under 125lb comes shipped free. Their import-grade 12x18 granite surface plate costs about $28 and weighs 85lb. Toss in enough other stuff to get to $50 and you get a CNC base that's planar and flat to less than 0.001" for under $30. Good base for high accuracy stuff.

[ken]

How do you attach your tool to the granite? Drilling a hole in the stuff and putting in anchors to hold things down is a royal pain.

Ken

[Joe Gwinn]

How do you attach your tool to the granite? Drilling a hole in the stuff and putting in anchors to hold things down is a royal pain.

Granite is easy to drill. Use diamond glass-cutting drills and plenty of tap water.

[erikbojerik]

bump... any updates of note?

[R.G.]

I collected a batch of parts for the last version I concocted, then real life intruded on my fun again.

I'll get back to it as soon as possible. Right now the mechanics are a bobbin holder on the end of a stepper motor, a stepper driving a leadscrew to move the tension/guide assembly left/right on a ball-bearing drawer slide, and a roller/brake tensioner on the moving part of the slide.

I was suprised - I got an accuride drawer slide, and the straightness (which matters not a farthing for this!) was under 0.001" in four inches. Play was zero, indicating a residual preload on the balls as received. Just remarkably good.