Hand Coil Wave-Winder

[Nick]

When you'd mentioned your winder before Nick, I'd assumed that it was a homebrew one - not the genuine article. Very Nice!

My apologies. I know I should have made it myself out of sticky-back plastic and a bit of left-over cheese.

Let the side down and all that

[Andy M0FYA]

That posher feed mechanism certainly looks like an improvement over the somewhat crude version on mine.
Andy

[Yorkie]

I've made a bit more progress on the project, which is all but finished, though I've yet to see if it will take off and land! I do have some fairly fine cotton covered wire (the winder won't work with enamelled wire as the turns will slip off), so hope to give it a spin before too long.

I've modified the original design somewhat.

I'd mentioned earlier that initally I'd turned a 'drive pulley' out of plastic and had fitted an 'O' ring as the tyre, but it was pretty hopeless as it didn't have enough grip and kept stalling. I then replaced it with a plastic bevel gear I posted about recently, overlooking the fact that this would be quite hopeless for wave winding as the ratio of the gear was of course 1:1 so each turn would fall on top of the preceding one.

I was able to source a sewing machine 'cotton bobbin tyre' used for winding the bobins of sewing machines, which is a much thicker section than an 'O' ring. I turned a brass pulley onto which I fitted the tyre, and I also fitted a spring on the winder shaft to keep the friction drive wheel pressed against the tyre in the event of wear. It now works fine and by moving it inwards or outwards on the friction disc, alters the turns ratio and hence, the wave winding, but I won't know how finicky that adjustment is till I try it.

As stated in an earlier post, I'd already dispensed with the turns counter (a 'tally counter' clicker) as featured in the construction book, and replaced it with a Poundland pocket calculator. This operated by using a brass cam on the coil shaft to operate a microswitch which shorts out the = button on the calculator. If you press any number on a calculator then press + twice, it becomes a constant, so each time you press the = button, the number you entered as a constant will increment. In this case, the number to enter is 1 + +. Then as the microswitch shorts out the = button with each turn of the winder shaft will count upwards.

I've made the brass wire guide mechanism, which guides the wire from the back wire feed-reel shaft around the guide on the cam shaft at the front, which moves the guide back and forth with each turn of the handle, and the wire goes through the front wire guide to lay turns onto the coil former.

I've attached a few pics which make things a bit clearer.

Hope that's of interest.

[Yorkie]

As it’s now more than 18 months since I started this thread, for anyone who hasn’t seen one of these little machines in operation, there’s a youtube video of the David Gingery winder going about it business here:

http://www.youtube.com/watch?v=FIOocMoRsYQ

For anyone who has built but not yet used one of these little winders or who might be thinking about building one, a few more comments and pics about my experiences might be of interest.

My original intention was to wind a replica Repanco DRR2 coil to build the 1959 BBC Focus Radio out of curiosity, but instead, I wound that coil by hand, having borrowed a DRR2 from forum member Alan Beckett to check the inductances. I built the 'Focus' which confirmed what I'd suspected - basically, it's a crystal set with a two transistor amplifier tagged on, so while it had sufficient volume for headphones, it had extremely poor selectivity. As the BBC layout was very untidy I designed and etched a PCB for it. The coil did it's job fine, but I must say that any lad who bought the bits out of his pocket money back in 1959 - if he succeeded in building it and getting it to work from the untidy instructions - would have been greatly disappointed, and far from being encouraged to get into electronics, would I think have been put off. At that time, circuits for simple one-valve TRFs abounded and could have been built more cheaply, and with much better results. I elaborate on the BBC Focus in another thread some time back here:

http://golbornevintageradio.co.uk/forum/...p?tid=2606&page=2&highlight=BBC+Focus+Radio

Be that as it may, over the last few days I've finally got round to winding a few test coils on my David Gingery version of the original Morris winder machine.

The downfall of this – or indeed any machine that will wind wave-wound coils - is the need for either Litz of cotton/silk covered wire, otherwise the turns will just slip off. Such wire is prohibitively expensive. Using the 45 SWG 'cotton covered' wire, with reasonable results. I wound 500 turns onto a 12.5mm Tufnol former, which turned out to be 105 Ohms, and 4mH. (The resistance of the wire is 4 Ohms per Metre, so the coil consists of approximately 26 Metres). I deliberately made the width of the coil narrow (5mm) to see whether or not the turns would slip off the edge of the coil when several layers have been laid down. I suspect that slightly thicker wire might give better results, but that's a matter of conjecture.

The rate at which the 'waves' arise is dictated by how close the rubber tyred drive disc is either towards the outer perimeter of the coil drive disc, or more into the centre. The rubber tyred disc and the one that it drives form what might be termed a simple 90 degree 'bevel drive'. The closer to the edge of the drive disc, the fewer the waves - the nearer to the centre, the more acute the angle of the waves. The shallower the angle, the greater the risk of turns slipping off the edge of the coil as the layers build up, though the type of covering (silk, cotton etc) and the gauge of the wire will have a bearing on this.

With a shallow angle the turns lie closer together, which increases the inductance for a given number of turns, but on multilayer coils the risk of turns slipping off the edge of the coil is greater.

The instructions in the book, which I referred to in an earlier post explain the relationship of the drive, and why it isn’t possible to use a 1:1 bevel drive, which would be a bit more refined. It states:

'The ratio of the crank shaft to the main spindle is infinitely variable by moving the friction wheel to any position along the shaft so that it contacts the friction plate at varying points. This is an important feature because if the ratio was simply 1:1 each successive turn would fall on top of each other and the desired pattern would not evolve. While a gear drive would be more positive, it would not be possible to achieve the very fine adjustment that can be had with this simple friction drive. It requires only a little experimentation to discover the optimum adjustment for each coil job and for very good results to be had'.


I've wound another coil with the drive disc more into the centre, which increased the rate of the weave and makes the angle of weave back and forth more acute, and I had no difficulty winding a coil of 500 turns, without any slippage of turns. I'm sure I could have increased the number of layers without any problems. However, the downside is that because the turns are more widely spaced one from the next, this has a bearing on the inductance. I'd fitted another cam which affects the width of the coil - this one enabling a 12mm width of coil whereas the one referred to in my first post was just 5mm wide. Though this second coil more than twice the width of the first, and had 500 turns like the first one, the resistance of the wire was almost the same (100 Ohms) but due to the less dense weave pattern, the inductance was only 3mH as compared to the smaller, but more tightly woven first coil, which was 4mH.

I've attached a few pics to help make more sense of the above. The first pic shows the weave of the coil taking shape after 50 turns have been wound on.

In the second pic, two black rings caused by the drive tyre will be seen on mating disc - the outer one shows the positing for the tighter weave (closely spaced turns) - the inner black ring shows the position of the tyre for the looser weave with wider spaced turns. The extent to which the cam disc is offset from the centre dictates the extent of the left and right hand throw if the wire guide, and hence, the width of the coil. The one shown in the last pic is for a 12mm wide coil.

I've also attached a pic of this latest coil.

I'm not winding these test coils to any great purpose - just to familiarise myself with the capabilities of this handy little gizmo and to confirm that it isn't just a novelty but does work as well as the youtube video suggest. I should add perhaps that the inductance of these test coils is far higher than that required for say a replica Repanco DRR2 coil (DRR2: 'Dual Range Reaction 2 Waveband'), in which the highest inductance - that of the long-wave winding, is only 1.5mH, the MW being 170uH and the coupling coil 145 uH.

A bit of a minority interest I know, but I hope this update is of interest to someone.

[Andy M0FYA]

David,
I'm following this thread with great interest, as I have a need for a couple of coils wavewound on Aladdin formers for a restoration project (although I'm thinking I might be coming to you cap in hand to see if you would wind them for me...............)
One thing I have read somewhere is that it is beneficial to sparingly apply shellac varnish as the coil build up, to prevent the turns slipping off, and then apply a layer after the coil is finished. Have you tried this idea, or do you have any thoughts on it?

Andy

[AlanBeckett]

That's a splendid job young David. I've got other Repanco Coils if you want to practice.
ebay Item 121353070909 looks promising. They don't say so but I expect you could buy it in longer lengths. It works out at about 60p/m I think.
I'm sitting very firmly on my hands for fear of starting something I won't be able to finish in the current climate.

Alan

[Yorkie]

One thing I have read somewhere is that it is beneficial to sparingly apply shellac varnish as the coil build up, to prevent the turns slipping off, and then apply a layer after the coil is finished. Have you tried this idea, or do you have any thoughts on it?

Andy

A good idea Andy.

Cellulose sanding sealer (rather than shellac sanding sealer) dries very quickly - in seconds rather than minutes, and likewise would clear dope as used for aero-modelling tissue, so could be applied say every 50 turns, but I purposely didn't try applying anything during winding as they didn't in the film and I wanted to see if the turns would stay on. If you look at the film, you'll see that the turns are widely spaced - not close together and that's what happens when the rubber drive disc is closer to the centre. That has the effect of lowering the inductance, just as it would on a conventionally wound coil with spacing between the turns. If the turns are wound closer together the angle is less acute, and that's when there's a risk of them falling off the edges, so applying a fixative every few turns would I'm sure help.

I did dip the finished coils in hot beeswax and left it in for a little while to heat the coil up so that as much of the wax would drip off the coil rather than solidify in a thick coating.

I think it would take quite a lot of experimentation to get the inductance of the coil just right. I suppose the best way would be to wind a coil with say 200 turns, check the inductance and if too high, take a few turns off at a time till the inductance is what is desired. However, once that experimentation was done, I think consistent repeatability would be possible if the turns and settings of the discs are noted. Of course, if a coil has a slug rather than air cored, it leaves some scope for adjustment.

I've do idea what gauges of wire was originally used by the likes of Denco, REP, Osmor etc, or whether it was Litz or cotton covered. All a bit of a leap in the dark really.

I still don't quite understand how the original Morris winder managed to lay down the waves when the ratio of the bevel gears seems to be 1:1 with no scope for adjustment, risking the turns falling on top of one another rather than side by side, but I may be wrong on that ratio. It must have worked because they did sell lots of them back in the 1930s! I've no intention of going into business winding coils and ought to get out more but it was a rainy day when I had as furtle around with the winder in the week.

[Andy M0FYA]

I'm not sure that the Morris winder was intended to wind wavewound coils of the type you've wound - the instructions refer to Honey Comb coils and Spider Web coils. They are both bigger diameter types for crystal sets and early valve circuits.
Andy

[Alan Douglas]

The Morris (mine anyhow) has drive gears of 16 and 42 teeth, and driven gears 39 and 44.

[Nick]

Lovely video, David - most impressed. Love the music too!

The Morris (mine anyhow) has drive gears of 16 and 42 teeth, and driven gears 39 and 44.

Mine has 4 gears - 39, 40, 42 & 44-tooth and 3 cams, labelled 1/8", 3/16" & 1/4" (these are the default standard items originally shipped with the unit).

You can also get a 15-tooth gear and 3/8", 1/2", 5/8" & 3/4" cams (the 15-tooth gear is needed for these larger cams).

The instructions I have say it winds spider, honeycomb & solenoid coils...