Automated Coil Winder (in the works)

[RedeemerbassesScott]

Hi All,

My name is Scott, I build bass guitars. I've been working on this coil winder for about a month and thought some of you might find the information about it usefull.

For the motor I use a 24volt DC 2speed motor. I have my spindle direct mounted to a shaft that I fabricated. No belts no pulleys. I use a pulse width modulation speed control to control the speed of the motor. It can make the motor barely crawl, go full speed (3200 RPM) or anywhere in between.

For counts I use a Veeder Root preset counter. I can program in the number of windings I want and when the desired number is reached the counter will kill the motor power via an internal relay. Its got a big lcd screen that's really nice and easy to see.

The counts are triggered off of the spindle shaft by a Honeywell Bi-polar hall effect sensor. The spindle shaft has 2 neo magnets mounted on it, one with the south pole up the other with the north up. The great thing about this sensor setup is that it requires both poles to pass the sensor before it will send a pulse, so there is no interference from a loaded bobbin since the bobbin magnets will only have a single pole faceing the sensors.

The most complicated part of this winder is the automatic traverse. I use a linear stepper motor to move the wire guide left and right. The stepper motor is controlled by an AVR microcontroller and EZ-Driver board. The EZ-driver is what actually powers the motor with its running voltage and supplys that voltage to make it either step left or right, whatever numbers of steps.

The microcontroller is an AVR Butterfly, its a low priced promotional package type set up. It has a Avr atmega 169 chip onboard along with a number of in's and out's, a 5 position joy stick and a little lcd screen. You program it via serial port using soft ware that AVR provides. You can program the microcontroller to move the stepper however you need. The actual program I'm using now is quite complicated and was not written by me, but by another hobbiest that came up with this setup. Although his is a bit different than mine he helped me a lot in getting everything setup, and he really knows his stuff about programming. (Lord knows I don't! but hey I'm learning..) The new program's got all kinds of different settings so you can adjust from the butterflys joystick including the traverse speed, max left ,max right, initial delay, decel accel, built in counter and rpm guage, ect.

The traverse setup uses a second Honeywell hall effect sensor to tell the AVR each revolution of the spindle. This way the AVR knows how fast the spindle is turning and can sync the travese to match the left and right speed. In other words when the spindle is moving slow or fast the traverse does the same.

To power the setup I use a converted pc power supply. I have tons of old pc parts so this was a cheap way out.

I'm still in the building process of this winder, I've got a video posted here that you can watch if you like. This was when I first got it going and before the new software. With the software shown you had to reflash the Avr with the different settings needed for the particular bobbin being wound at the time. It still works good this way just not as user freindly. Also all of the guts are showing and the wiring is a mess, I can assure you it's all tidy now. Here's the link http://www.youtube.com/watch?v=M1qnbqcpR_c Its a little different now and I plan on posting a new video of it completed soon.

All and all the parts for this winder cost me less than $150.00 Check ebay for the parts you would be surprised how cheap you can get them, it just takes a little searching. Here is a rough breakdown of cost of the main components.

Airpax linear stepper motor 15.00 Ebay
Avr Butterfly 25.00
Ez Driver Board 20.00
Speed control 30.00
Veeder Root preset counter 10.00 Ebay (lucky find!)
2 Honey Well hall effect sensors 10.00 Ebay
Dc motor free from a dc fan I had
I fabricated a few parts and also had parts like knobs, power supply, wire, switches already.

Tell me what you think?

[David Schwab]

That, my friend, is amazing!

really nice work.

I want one!

[RedeemerbassesScott]

That, my friend, is amazing!

really nice work.

I want one!

Thanks David, I updated some more recent video last night. This forum is great, I wish I would have found it sooner!

[R.G.]

Nice work. My approach was with rotary steppers for both the bobbin and the traverse, the traverse being run by a leadscrew on the shaft of the traverse motor instead of a linear stepper. Both approaches will work fine.

The mechanics are the easy part. What remains is a A Simple Matter Of Programming (ASMOP). 8-)

[R.G.]

Forgot:
This approach is breaking out all over. Pretty soon you, or me or someone else is going to finish a "recording" version of this and unwind a few hand would or vintage pickups, then be able to replicate them more or less exactly.

[JGundry]

Am I correct that you are stuck with the rotary motion with any push rod you would attach to a rotary stepper?

[R.G.]

I'm not sure what you mean.

Any push rod you connect to a rotating disc or eccentric will have a nearly-sine-wave motion profile ( the math is fairly complicated, but it's aproximately sine). But linear steppers step ... linearly, and any rotating motor connected to a screw produces linear motion.

[JGundry]

I'm talking about the linear screw type stepper motor used here. I have never seen one of these screw linear stepper motors in action so I may be wrong about how they work. Does the shaft on the screw linear stepper motor rotate as it moves? If it does then you have to have a wire guide that works with a rotating shaft. Or maybe the shaft does not rotate at all?

[RedeemerbassesScott]

I'm talking about the linear screw type stepper motor used here. I have never seen one of these screw linear stepper motors in action so I may be wrong about how they work. Does the shaft on the screw linear stepper motor rotate as it moves? If it does then you have to have a wire guide that works with a rotating shaft. Or maybe the shaft does not rotate at all?

No the shaft stays still, the inner threading inside the motor turn and since the shaft is anchored to where it can't spin the motor draws in or pushes the shaft left or right depending. It's kind of like when you tighten a nut on a bolt, if you keep the head of the bolt held secure when you tighten the nut it will draw in the bolt in, same principal.

[chevalij]

I'm trying to figure out how the AVR can control the stepper driver. I know you can load a program onto the AVR, but does it simulate a printer port or similar? I guess it's like you said, the programming is the tricky part.

[RedeemerbassesScott]

I'm trying to figure out how the AVR can control the stepper driver. I know you can load a program onto the AVR, but does it simulate a printer port or similar? I guess it's like you said, the programming is the tricky part.

The Easy Driver actually controls the stepper motors voltage, steps, direction ect. The AVR sends pulses to the Easydriver telling it what to do to the stepper.

[chevalij]

The Easy Driver actually controls the stepper motors voltage, steps, direction ect. The AVR sends pulses to the Easydriver telling it what to do to the stepper.

Yeah, I've got that part. I've got to do more research on the butterfly. I have no problem controlling a stepper using a printer port capable driver board and mach 3. It's transfering it to the AVR and being able to modifiy the limits with the joystick while on the fly etc.

[NB70]

Stepper winder

I have been also thinking about a simple stepper based coil winder, just like the one R.G. is suggesting. I chose this design because I have recently made a CNC router so I already have the stepper motor drivers, power supplies and control software. I just needed to buy a couple of extra stepper motors.

My router uses the same principle of stepper motor driven all-thread leadscrews with Delrin antibacklash nuts. The winder's rotating axes will be secured using double nuts onto skateboard bearings. I have added pulleys to allow more than one pickup to be wound at a time, and to speed up the rotation (direct driving using a microstepping motor would be rather limited in speed).

I haven't drawn the wire guides, bobbins or tailstocks, but using the aluminium sections means that you can swap out such components and adjust their positions easily. I will also mount the spools of wire on a dereelers.

The obvious way to control this type of machine is using G-code. I use the free and opensource EMC2 to drive my router. It is an excellent program, used for toy machines and industrial 5 axis monsters alike. In EMC2 you can specify linear and rotary axes, the wire traverse is the X-axis and the rotating bobbin is the A-axis. You can then specify a long series of helical moves in G-code, to wind at whatever rotation speed and wire spacing you want.

I am planning to run the machine in open-loop mode (no rotation counters or positional feedback) relying on the accuracy of the steppers. This works with my router - but I am not sure whether it will work with a winder.

I am in the middle of moving house at the moment, so no workshop space. But I will keep everyone posted with photos, plans and G-codes, when I make some progress.

[David King]

Scott,
is the linear actuator for the traverse one of these?
http://www.mpja.com/prodinfo.asp?number=17284+MS

[R.G.]

I have been also thinking about a simple stepper based coil winder, just like the one R.G. is suggesting. I chose this design because I have recently made a CNC router so I already have the stepper motor drivers, power supplies and control software. I just needed to buy a couple of extra stepper motors.

My router uses the same principle of stepper motor driven all-thread leadscrews with Delrin antibacklash nuts. The winder's rotating axes will be secured using double nuts onto skateboard bearings. I have added pulleys to allow more than one pickup to be wound at a time, and to speed up the rotation (direct driving using a microstepping motor would be rather limited in speed).

Cool setup. I just lucked into a bunch of 80-20 section cutoffs, too, and that's a good way to use them.

I was actually going to put the bobbin pulley right on the stepper motor shaft with a shaft coupler. For me, that's the single place in this style of design where some calculation is needed. The tension on the wire is transmitted directly to the bobbin, and hence the motor shaft. So the max wire tension times half the width of the bobbin appears on the motor shaft as load torque.

It's that load torque which determines what the motor torque has to be to avoid missing steps at whatever speed, as the stepper must supply the wire-load torque at the max winding speed. That will either specify the motor size or the max speed, or the number of bobbins that can be wound at the same time.

Interestingly, the torque is almost a half-sine wave, so you could probably run two bobbins at 90 degrees out of phase with about the SAME max motor torque as one bobbin, and it would be smoother on the motor, as it would approximate a constant torque. Not perfect, but better.

I haven't drawn the wire guides, bobbins or tailstocks, but using the aluminium sections means that you can swap out such components and adjust their positions easily. I will also mount the spools of wire on a dereelers.

The pulleys let you modify the torque/speed for the motors. I was on the other tack: figure the torque needs, and just use the motor. It makes setting up the bearings for the bobbin a non-task, as the motor bearings already provide the precision, ready bought.

I am planning to run the machine in open-loop mode (no rotation counters or positional feedback) relying on the accuracy of the steppers. This works with my router - but I am not sure whether it will work with a winder.

A stepper will work fine on a winder, just like on the CNC router as long as you keep the motor torque and resonance-bounce down to where the motor does not lose a step. The same concepts apply - step size, oscillation and resonances, torque, etc.

I've done a lot of the mechanical hacking, impeded a bit by my getting into a seat of the welding courses at my local community college. I was planning to tie down the control software later.

Frankly, I'm not sure that G-codes are needed. The task will always be the same - spin the bobbin in a single direction at a coordinated position/speed to the traverse. Move the traverse in coordination with the bobbin position on its turning.

One of my objectives has been to build in the ability to "record" a bobbin being either wound or unwound into a long series of rotational/traverse position pairs.

If you think about it, every bobbin being wound can be described as a series of number pairs. The first number is the rotation of the bobbin, from zero degrees to maybe 10,000 times 360 degrees. The second number is the lateral position of the traverse for that angular position, with a fineness which allows you to place the wire within one half-diameter of the position you want it.

The rotational position is a maximum of 3.6E6 degrees if you take one-degree step. Stepper motors don't take one-degree steps usually, they're 200 or 400 steps per rotation, so expressed in steps, the rotation is from 0 to 400x10,000 steps or 4M steps. A modern computer will handle that nicely.

The lateral position needs to be a half to one inch of travel in steps smaller than about 0.002, so 0.001" is convenient. For 1/4"-20 leadscrew and a 200step/rotation motor, the step size is 0.05"/200 = 0.00025". That's far too small to be believed, probably better than you can arrange the accuracy of the bearings and leadscrew nut. So your increment of lateral position is something like four to eight steps per position change; plenty of resolution. It will be a challenge to keep backlash down to tolerable limits on a setup like that, but that's always the challenge for precision. It may be that one has time at the end of every traverse to stop the bobbin (for a few milliseconds 8) and de-backlash the traverse for the other direction.

Ballscrews are a possibility, especially for a 1" traverse, but it might be that you could get a microscope stage for that small a traverse and not have to deal with backlash on a homemade leadscrew.