Can anyone identify this honeycomb inductor winding pattern?

[pitts8rh]

I am trying to identify this unusual wire winding pattern shown below for an inductor from a vintage Moog theremin that I am trying to accurately replicate.  The operating frequency is in the 200kHz range, and the ferrite core is 1/2" diameter.  The inductance is believed to be approximately 75mH.  The light colored winding is cotton-covered magnet wire.

At first glance I thought I was seeing a normal, but low-profile multi-section honeycomb pattern that is seen in pi-wound chokes, but the helical gap running down the length of the inductor suggests that it is something different.  Each 1/8" wide slightly raised segment (between gaps) appears to have its own honeycomb wind, but rather than being wound one section at a time, daisy-chained down the ferrite, it somehow has the look of an acme-threaded screw.  I want to machine-wind the same pattern (for aesthetics, not simply functionality) but I can't figure out how it was done.

Has anyone seen this?  Even if I could find a better view of something with a similar wind I may be able to figure it out.

[Beamin]

Too hard to see. I posted a video where this guy found a old army transmitter from the 60's where he uses this special tool to make such windings. Really rare tool from the 30's or 20's that if you had the dimensions you could replicate it with a milling machine and a lathe, maybe even a 3d printer. That's the kind of tool that once you find it you never sell it. It had gears and cams that would go back and fourth as you wind it. I can't seem to figure out how to find all threads started by myself or if that's possible with the forum search. I'll see if I can find it again but 0.000000001% chance of finding this tool.

[chris_leyson]

There is a good universal or wave coil winder here.
and a link to a book on winding universal coils  "https://drive.google.com/open?id=1Fb-Jz1mJH5M7R8J9dY1WPc_skzhdoJVy"

It doesn't look like a typical wave wound coil and a close up photo might shed a bit more light on it.


[PointyOintment]

Too hard to see. I posted a video where this guy found a old army transmitter from the 60's where he uses this special tool to make such windings. Really rare tool from the 30's or 20's that if you had the dimensions you could replicate it with a milling machine and a lathe, maybe even a 3d printer. That's the kind of tool that once you find it you never sell it. It had gears and cams that would go back and fourth as you wind it. I can't seem to figure out how to find all threads started by myself or if that's possible with the forum search. I'll see if I can find it again but 0.000000001% chance of finding this tool.

Here (10:05): https://www.youtube.com/watch?v=MBN8EjnJC9I?t=605

Thread you wanted to find
All threads started by you (Profile -> "Show Posts" on the left -> "Topics" at the top)

EDIT: I can't seem to get the video to embed or even link to the specified time. So just go to 10:05 in the video manually.

[floobydust]

It looks like 20 basket-wound sections that are partially overlapping each other.



I get large low cost ferrite rods on fleabay from china I use for VLF stuff.

[chris_leyson]

Another link to "Coil Winding" by Geo. Stevens Mfg. Co.
http://www.vintagewindings.com/gen%20pop/8299543VW8335/TransDesign%201/Coil%20Winding.pdf

It looks like a progresive universal winding and details are given in chapter 4 "THE PROGRESSIVE UNIVERSAL WINDING"
Good luck trying to find a machine to do progressive wound coils. I think you might have to build one from scratch with a servo driven wire guide, if it's servo driven and computer controlled then you don't have to worry about custom gears and cams. Good luck.

[coppercone2]

Those coils are meant for low pass filters right?

These coils must be less stable then other wound types (from a metrological prospective) since there are voids between them (in the video winding shown), so the insulation has a easier time creeping and changing the coil properties. It seems kind of 'gooey'.

[pitts8rh]

All of these coil winder videos simply show how single section or multi-section pi-wound coils can be made with a machine such as with a Morris or Gingery Universal coil winder.  But the inductor in question in the picture doesn't have separate sections like most RF chokes.  It almost looks as if you took one of the winders shown above and mixed in a slow cyclic transverse motion running up and down the length of the ferrite in addition to the faster 1/8" cyclic motion.  I'm thinking of a slow, large amplitude triangular-wave motion with a faster and small amplitude triangle wave motion (close to, but not exactly the same as the core rotation frequency) superimposed on it. 

I'm guessing-I've never seen a coil wound like this.  But if this is the case I need to rethink my winding machine...

To respond to previous post: These inductors are not for filtering - they work with the extremely low capacitance of the theremin's capacitive elements (pitch rod and volume loop) to form high-Q resonant structures that linearize the theremin's response.  They are wound like this to minimize inter-winding capacitance and keep the self-resonant frequency high.

[chris_leyson]

I could be wrong but I think you have the cyclic "wave" winding motion coupled with a slow linear transverse motion which you only do once only from start to end. A close up photo of the winding might help. I've never seen a progressively wound universal coil, they've always been single windings or multiple section "pi" wound. Your Moog theramin is the first time I've seen a progressively wound coil.

I've been thinking about building a coil winder and a mechanical solution like the classic Morris coil winder just isn't an option. You've got gear trains and cams to sort out and that's just for a single winding. For a progressive winding the classic solution is to move the bobbin or wire guide in a linear transverse motion, that another gear train and a lead screw. That's one hell of a lot of accurate machining and how the hell do you cut a cam anyway.

You only need two moving parts, the rotating bobbin and a transverse wire guide. DC motor with encoder for the bobbin drive and a linear servo to move the wire guide. Synchronise the transverse motion of the wire to the bobbin rotation and you can wind anything. Been looking at Mighty Zap servos for the transverse motion, available with 27mm, 41mm or 56mm stroke. That's probably not enough stroke for your Theramin coil but there are other linear servos available. Just my 2 cents worth if you were given a clean sheet of paper.

[floobydust]

I thought Theremins use two RF oscillators hetrodyning to an audio tone, generated from the one osc. detuned by human body capacitance.
Your pic shows these two identical RF oscillators.
That would make this inductor's purpose an LPF to remove mains hum and filter out AM radio from the capacitance sensor input?

[pitts8rh]

 I agree with most of what Chris said above, although I'm still unclear whether you need a single slow linear traverse of the core or multiples as I described above.  But the rest of what he says is pretty much right on.

What is clear is that while a mechanical winder can be fairly easily made to wind the simple basketweave/honeycomb coils shown in the videos,  the need for an adjustable, synchronized linear movement makes a mechanical winder get very complicate very quickly.  Six inches of linear movement is too much to be practical for a cam;  a lead screw arrangement would be more suited, but it would have to be reversible and be geared to (and in a ratio synchronization with) the core rotation.

I do have the capability to cut cams with any motion contour, linear or otherwise, and a conventional winder was going to be a fun project for me, but I'm now thinking that I need the versatility of a dual stepper motor winder.  One motor to drive a lead screw for both the small and large cyclic movements, and the other to rotate the core.  All of a sudden this little side project is going to be a major distraction.

If I had the original in hand it would be easier to understand, but I don't own one of these - the photo above is not mine.  I want to replicate it not just functionally, but so that it looks original as well.  I was hoping that someone would recognize this wind as something used in vintage chokes, but it seems like an oddity.  BTW, this theremin was circa 1960s-70s.

[pitts8rh]

I thought Theremins use two RF oscillators hetrodyning to an audio tone, generated from the one osc. detuned by human body capacitance.
Your pic shows these two identical RF oscillators.

Correct, except the audio is the difference frequency between the two oscillators; one is fixed, the other shifted around by body capacitance.  I think that's what you meant?

That would make this inductor's purpose an LPF to remove mains hum and filter out AM radio from the capacitance sensor input?

No, it has nothing to do with that.

The variable oscillator is incapable of being shifted much by the fractional picofarad changes in capacitance, particularly when the hand is near to, and masked by your body.  The large inductor and the very small "antenna" capacitance (it's not an antenna though, just a capacitance element) form a high-Q resonant circuit that is tuned to be resonant when the playing hand is very near the body.  The greatly enhanced impedance changes of this LC circuit near resonance provide a strong influence on the variable oscillator frequency, much stronger than that of the antenna alone.  It is a key part of the linearizing circuit that makes octave intervals in pitch approximately spatially equal (within reason) even though the capacitance delta for hand movements far away from the antenna is much smaller than up close.

[floobydust]

I understand. It's just that the inductor looks like a real bear to replicate, with gray Litz wire and a complex winding pattern. Another pic had burgundy wire.
Moog's 1961 construction article (with NPN transistors) he does say machine-wound using special "progressive universal" method and cannot be duplicated by hand-winding. There are alternate parts... of three 25mH Miller's 6308 or Meissner 19-1053, might be worth considering unless the winding machine algorithm can be figured out. Maybe a patent search would find it.

edit: I found this: patents with "progressive universal" and there is  some coil winding machines and old radio antennas wound this way. Must be a lost art.

[pitts8rh]

edit: I found this: patents with "progressive universal" and there is  some coil winding machines and old radio antennas wound this way. Must be a lost art.

Floobydust,

I think you have pointed me to what I was looking for.  "Progressive Universal" is the keyword, and this patent shows the complexity of a mechanical winding machine:

https://patents.google.com/patent/US2305085A/en?q=%22progressive+universal%22&oq=%22progressive+universal%22

I'll have to try to wade through the plurality of patent-ese tonight to see exactly what the winding sequence looks like.  But I suspect that I will have to either come up with a stepper controlled winder, or fake it by doing a many-section pi-wound coil. 

[floobydust]

chris_leyson got the name of the pattern earlier and I stumbled onto it in Moog's article. I guess it's also used for electric guitar pickups. I haven't seen windings like this for decades.

http://www.crankorgan.com/CW1.htm used an Arduino and stepper motors {slow load web page}
" So what did I learn? If you search the Internet about Wave and Progressive Universal wound coils you will find little information. I will give you the trick! Are you ready? As the machine moves back and forth driven by steppers your program will try to wind the wire in the same spot over and over if you are trying to wind a WAVE coil. Winding back and forth on a bobbin is a solenoid coil! A wave wound coil looks like it was woven or sinewave. Here is the Trick! Stepper are even number devices. You need to add extra steps for the spindle motor outside of the Loops in your program. In a mechanical winder they use a gear with an odd number of teeth. With every cycle of the machine the spindle jumps ahead. I did not use a stepper library. I used only Arduino code. It took one loop to go right and one to go left. Those were inside a third loop. In between the two loops I put some extra 17 steps for the spindle. Each loop has a spindle step in it! I was also able to drive the Tip120s as PHASE drive. Much more thinking will be required!!!! The back and forth motor or X axis uses a 1/4-20 rod so it needs 4000 steps per inch. The spindle motor needs only 200 steps to go around once! Have fun! It took my old brain many hours to get the thing working. It was a HOOOOT! "

[pitts8rh]

Yep, that's what I mean when I say ratioed synchronization - perfect sync will end up piling wire on the same spot.  Fortunately the mechanical winders use a variable ratio rubber-tire-on-disk mechanism that is very unlikely to allow this problem, and if so, not for very long.  Gear drives, and of course stepper drives have to pay attention to keeping a slight phase difference between the lateral feed motion and the rotational angle.

[chris_leyson]

Well spotted floobydust, it must be a low pass filter and I failed to spot that very important detail in pitts8hr's post, 75mH. If you know the number of turns required, the wire gauge and length of the coil you could work out the required overlap. Also, a lead screw to move the wire guide is probably a better solution as you get more rigidity.