Pickup winder construction: platen design

[Jason Rodgers]

My other thread about pickup bobbin materials, construction, and design has got me thinking more about the materials, construction, and design of the winder itself. Specifically, I'm interested in the shape and function of the platen - the flat surface on the end of the shaft, onto which bobbins are attached for spooling up wire - and the pros and cons of some of the designs being used.

When it comes to the various designs out there, I see that there is the commercially available Schatten winder, and variations on a theme presented in Jason Lollar's book. The Schatten simply has a rectangular chunk of metal (aluminum?), about 4" long x 1-1/2" wide x 1" thick, to which bobbins are double-stick taped. Lollar's has a 4" disc of Plexi/Lexan with holes for screwing on bobbins through their mounting holes (bobbins can be taped on, as well, if necessary).

I've watched some of Clint Searcy's pickup winding youtube videos, and his winder uses a combination of these two designs: a sturdy disc of plastic with taped-on bobbins.

I guess I don't know what kind of forces the platen speed and wire winding tension put on the bobbin, so don't know how secure it really needs to be. (Of course, I could just trust that Lollar and Searcy know what they're doing, and pick one.)

Also, does the mass of the platen have anything to do with the efficiency of the spin? Does a heavier platen like the Schatten or Searcy's sort of act like a flywheel and give it a little inertia? Or is it better to have the lightness of Lollar's design to allow it to start and stop quickly and easily?

I'm probably overthinking this. I usually do. Heck, Markku uses a cordless drill and gets good results.

What does the gallery think?

[John Sonksen]

Jason, you're going to laugh when you see what I'm doing. Literally double stick tape on the sewing machine bobbin winder, and it works. It does start to get wonky towards the end as the mass of the coil increases. I'm planning on altering it of course, but unmodified it works fine. I would lean toward a lighter material myself, I personally like it to stop when I tell it to stop. I also don't think the sewing machines were designed to have added weight on that end so I'd guess the less strain the better. I'll probably end up milling something out of aluminum at school, I may be able to make two if you're interested.

[David King]

I never liked the rectangular bar approach because it generates some wind at speed and the air movement can pull the wire off track or simply catch it and hang an extra loop of wire off a corner if you don't have your guides set perfectly.
I use polished discs and each one has the shape and depth of the bobbin wall milled into the end. This centers the bobbin and eliminates wire hangups. It really speeds things up at the mounting/ unmounting. I use a free spinning tailstock disc as well and just flip the bobbins according to what wind direction I want to achieve (rather than reversing the motor or having duplicate discs and guides at each end of the winder spindle.
Sometimes with longer bass pickups, the discs just get in the way and I'd consider a simple screw through the bobbin into the end of the spindle shaft as the best approach. Keep it simple. The most important thing is to have a bobbin that spins true and stays centered.

If you are going to use double stick tape you might consider the 3M R100 clear "Reusable" mounting tape (Fred Meyers office section). You can wash it off with water to renew it's stickiness. I use masking tape over the ends of the bobbin to hold it in place with the benefit that the wire can't get hung up or cut on the eyelets. Eyelets can really mess you up so i now drill for and install them after the coil is wound.

[John Sonksen]

Good stuff David, i like the trick of flipping the bobbin over instead of reversing the motor. Sometimes it's easy to start looking for complex ways to solve easy problems.

Are the discs you use aluminum or something else? I'm thinking those wouldn't be too tough to mill, and I've got access to some nice mills at school.

[David King]

I've been making the discs out of Baltic birch and gluing an aluminum collar through them that holds the set screw to fasten them to the spindle shaft. I actually turned a few of them on the table saw rather than the lathe using a 5/8" hole. I turned the collars on the lathe though. You just have to glue the collar in so that the disc runs true so I do that on the lathe too.
I start with a square of ply and mill the outline of the bobbin along with the center hole and a groove for the pole magnets to fit into. I trim it round on the bandsaw and clean it up on the TS with a sharp rasp and a sanding stick. That part takes a little practice to get it perfectly round. A cnc mill could do all of that in 1-2 operations.

Ideally i think I would use screw on faceplates like on a lathe, it would just be faster and and assure that the plates all ran square and true. At that point I would just buy small ready-made aluminum lathe faceplates from Little Machine Shop or eBay

[Jason Rodgers]

For this iteration of my winder, I'm gonna go with Lollar's recommendation of a 4" Plexi (I think that's what I have) disc. Right now, I have two pieces gluing up with CA. Lollar also instructs in the fashioning of a shaft collar into a faceplate with three screws for mounting the platen. There is a fairly robust worm gear with set screws leftover from my sewing machine parts that will fit this role nicely.

[David King]

That sounds perfect. I've noticed that there are several sellers of round plastic discs in various diameters and thicknesses on eBay (perhaps leftovers from some manufacturing process). They are quite inexpensive and might be a good source if you or someone else needed more than one.
There's a fairly elaborate example of a winder based on a little variable speed wood lathe that someone posted on the MEF site a couple of weeks ago. They detailed their electronics package which included a nice programmable preset counter with relay circuit to shut the winder off at the end of the wind. They also had a tachometer so as to set the spindle speed repeatably.

[Jason Rodgers]

I'll consider that resource if I decide to use multiple platens for different pickup sizes/designs.

But here I am overthinking a spinning disc, and John is actually winding real, functioning pickups by double-sticking bobbins on the turny knob of an old sewing machine! John had me over for a couple hours this afternoon and he wound up and assembled most of a humbucker. Thanks, John! You're gettin stuff done in that garage!

[John Sonksen]

I'll consider that resource if I decide to use multiple platens for different pickup sizes/designs.

But here I am overthinking a spinning disc, and John is actually winding real, functioning pickups by double-sticking bobbins on the turny knob of an old sewing machine! John had me over for a couple hours this afternoon and he wound up and assembled most of a humbucker. Thanks, John! You're gettin stuff done in that garage!

Thanks for coming over today, it was great to finally meet you and I really appreciate all the info you shared with me. It was nice to talk shop with a fellow builder, and I look forward to checking out your shop next month. I'm hoping my winder looks more like yours soon than what I've got right now, lol, but it was good enough to get my feet wet with!

[Jason Rodgers]

I just about have my platen disc constructed and mounted. Lollar suggests drilling and tapping a shaft collar for three screws to secure the platen to the shaft. While this is a few extra steps (instead of just gluing it on), the benefit is that it can be shimmed slightly to run true.

When considering wind count, I looked at a bunch of inexpensive counters. I have an old thumb-lever attendance counter, but the button depresses about 3/8" and it often sticks, so that was out. I looked up other lever-actuated counters, but read reviews that their plastic gears wear quickly and would not have accuracy for high-speed counting. There are rotary counters, but you have to match the belt/chain pulley exactly to the shaft you're counting so it counts accurately. I decided to go with this really simple and inexpensive digital counter. It has a battery (like a calculator), so it doesn't need external power. The count advance is simply achieved by contact between two leads that connect to terminals on the back. John gave me a reed switch when I was at his shop, and I think this will work well.
http://www.amazon.com/Amico-H7EC-BLM-Co ... al+counter

When the winder is complete I will post a new thread.

[David King]

I couldn't see a datasheet for that counter. I found this one for the Omron version which is probably close:
http://html.alldatasheet.com/html-pdf/1 ... C-BLM.html
Your counting speed looks to be 1000cps which can get you in a lot of trouble with false counts if you don't debounce your switch. I would definitely use it with a optical or hall sensor rather than a microswitch or magnetic reed switch. The numbers may be hard to read so mount it conspicuously.

The $30 Sestos programable counters on ebay are a nice way to go but they aren't $8.

[Jason Rodgers]

Doesn't matter if you could see any info on that counter, because it's all in Chinese! (Err... unless you DO read Chinese.)

I clearly don't know what bounce is all about, though it sounds bad, so I'm going to hope that I don't have any. I will also wave my stupid flag and admit that I'm probably wrong that cps means counts per second.

In my magnet and reed switch setup, I have a neo magnet on a little cam that was on the main shaft of the sewing machine. It is mounted on the side of this cam and the reed switch is oriented perpendicularly to the shaft so the magnet passes broadside about 1/2" away. I got this idea from one of the winder building sites that is linked on the MEF FAQ. I should have this setup running tomorrow night, so I'll report back.

[Jim McConkey]

I clearly don't know what bounce is all about, though it sounds bad, so I'm going to hope that I don't have any.

You have some. All switches have some. Any time a piece of metal hits another, it bounces off, temporarily making and then breaking contact. Depending on the force of the springs holding the switch in any position, it may bounce multiple times, generally with each bounce coming faster because it is traveling a shorter distance. Because it is reasonable to assume that a human cannot toggle a switch more than a couple times as second, debounce circuits are traditionally set to basically ignore rapid toggling faster than like 20-50 times a second or so. If your motor can go faster than the debounce rate, the debouncer will ignore some of the revolutions, giving you a false count (too low). Without a debounce circuit, your count could be ten times too high, so you really can't live without it. If you are good with electronics, you might be able to find and modify the debouncer to be a little faster, but they were designed for human-operated switches. If you want to use a fast motor, it would be better to use an optical interrupter-type counter, which does not suffer from bouncing.

[Jason Rodgers]

Well, crap. I guess I'll see how bouncy this setup is.

[David King]

In the datasheet I referenced above they refer to two other models of that counter that have counts per second rates of 20 and 30, These would have been easier to deal with but they would have limited your top RPM to1200 and 1500 respectively (which is actually just fine for winding pickups at home.)
The problem with mechanical switch bounce is that it can be really hard to see, it might only happen some of the time and at some speeds and not others. If your wind counts are 10% off and you don't know it that's a much bigger problem than if they are 1000% off and you know it right away.
The other critical factor on the datasheet is the minimum duration or duty cycle of the count pulse. On your counter it's 5ms which is very very short so not a problem. If your counter is internally debounced to count at 20cycles per second then you would probably need a much longer "on" time for the pulse to be counted. Basically you always want to shoot for a 50% duty cycle so that you get the most accurate count. If you have a single magnet that passes by the reed switch once a revolution suddenly the switch is only on 2% of the time and there's no practical way with magnets to keep the switch on longer unless you can find a 1/2 round magnet from some weird 2 pole DC pancake motor.

I think debounce circuits are quite simple and common and they undoubtedly are easily adjusted to do what you need.

Fairchild makes optical sensors and will even send you free samples if you go to their samples page. The sensors are very simple to set up and require very little power. They come in at least two types: interrupt and reflective. Either will work well. Just think of them as similar to the "electric eye" on your garage door closer. There's a light source (Infra red LED) and a phototransistor that senses that infrared light and switches "on" or "off" depending on whether the light is on or off. All you need to do is supply enough current and voltage to keep the LED on (generally 1.7V at 20ma) and then see if your counter's input is sensitive enough to tell the difference between "high" and "low" out of the photo transistor. These types of components were generally designed to work together so chances are pretty good. You can use a couple of AA cells to power the LED and use ohms law to calculate what resistance you need in series to drop the 3V down to your 1.7V. There are also plenty of LED voltage drop calculators out on the web to help you out.

[Jason Rodgers]

Aw, for frick sake, I can't do anything simple, can I?!

Ok, I've registered on Fairchild Semiconductor and I'm looking at their reflective sensors. What's the difference in the three output voltages of .1, .3, and 1? Which one would I want?
http://www.fairchildsemi.com/search/opt ... %20Sensors

And it's no wonder they can send you free samples when the product is $0.3872 each!

[John Sonksen]

I don't know if this is the same thing, but I bought ten of these the other day when I saw the shortcomings of the reed switch.

[John Sonksen]

sorry about the blurry picture, I'm heading to Wilsonville on Sunday, I may be able to swing by your place either on the way down or the way back and drop a couple of these off for you to play around with. I think they're 3V. I was planning on using them with an arduino circuit, and making a program in visual basic to keep track of RPM's, wind counts and control speed and winds. I was thinking of using two servo motors, one to control the speed and the other to control the traverse (if I'm remembering the term correctly). Still a long ways away from all of this but my basic idea is to be able to hook up my laptop to the arduino board via usb and monitor everything simultaneously, or output to a basic LCD that can cycle through the different readings.

[Jason Rodgers]

John, I see why you unloaded the reed switch on me now!

I just got upstairs from the shop after getting the winder all wired up and the platen mounted and (relatively) balanced. Geez, I wish I had a couple bobbins and some wire to give it a go! Again, I'll trust that y'all know what you're talking about when it comes to this phenomenon called bounce, but I think this setup is going to get me in the ballpark for my needs right now (or the zipcode of the ballpark).

You're welcome to stop by, John, but you'd have to fill me in on the application of your little light sensor jobby.

I'll get another thread going soon with the construction documentation.

[John Sonksen]

I'm afraid I can't offer too much with the how's just yet, I'm relying on some friends who are a little more experienced electronically than me to help me put it all together. Right now I'm in the process of gathering materials but between work, school, remodeling my in-laws kitchen and trying to finish my second electric, this stuff is taking a backseat. I'm hoping to ramp up this summer though and get this Frankenstein's Monster up and running then...