How to make a 1920s variable resistance

[SurreyNick]

Having given up all hope of finding an original variable resistance for my 1927 Cossor Melody Maker (my year-long search has been fruitless) I am resigned to the idea of replicating one. My criteria is to make it as near perfect to an original as I can. With this in mind I am hoping members of this forum can help with some information.

1. What are the dimensions of the base (WHD)?
2. What is the inner diameter of the circular winding?
3. What is the width of the former (excl. the wire)?
4. What is the thickness of the former (excl. the wire)?
5. What is the former made of?
6. What gauge and type of wire is employed in the winding?
7. Is the wire insulated, enamelled, or bare?
8. Is there an air-gap between the turns?

Finally, can someone furnish me with some close up photos of the resistance so I can see exactly how it was constructed?

These semi-circular variable resistances were made by several manufacturers so there is likely to be some variation between them. If someone has one in an original 1927 CMM then that would be the ideal.

Thanks

Nick

[Diabolical Artificer]

Sorry if I'm being thick but why not use a wire wound pot?

Re question 6, if you want to make one I think your main difficulty will be in manipulating the very fine resistance wire as used in wire wound devices. We are talking really fine gauge, I've found it impossible to work with. It is bare un-treated wire-has to be to make an electrical connection.

I have some old wire wound tracks/pots in bits you can have to experiment on if it helps.

Andy.

[Cobaltblue]

One of My 1927 CMM's is in the Flat here in Poole I will have a look this evening.

From memory though the wire is un-insulated resistance wire possibly Nichrome.

Cheers

Mike T

[SurreyNick]

A reasonable question Andy and if it were just a case of getting the set working again, then a wire-wound pot would be a perfectly sound alternative. But I have taken the time to find original working components throughout so this is not an option I would consider. I am loathe to replicate a component, the thought of using a component from the wrong era is anathema for this restoration.

I only need a 3 ohm resistance so should be able to get away with a workable gauge. A coated wire should also be fine as the edge upon which the armature runs can be carefully sanded to expose the bare wire. Likewise the terminal ends. But that's why I also wanted to know if there is an air-gap between the turns as that would indicate a bare wire.

Nick.

[Cobaltblue]

Hi Nick

Sorry my steel rule was nowhere to be seen lunch time.

Interestingly I measure the resistance and it showed 30 Ohm.

The base is pressed aluminum the wire is un-insulated and spaced.

The insulation the wire is wound on is the same sort of material like a compressed card that you found on old plug cord grips.

It is in series with all 3 filaments.

I seem to remember that the 1927MM was originally designed for the Stentor series valves which had a nominal 1.8V filament.

Hope these help

Cheers

Mike T

[Mr Moose]

Hello,
From one I have the resistance is about 4.5 ohms
The aluminium base is 2 3/8 in. x 3/4 in. raised in the centre by 1/8 in.
The pivot point of the wiper arm is 1 in. above the bottom of the base. (7/8 in. above the raised portion.)
The radius of the fibre former is 15/16 in.
The fibre former is 1/2 in. x 1/16 in.
92 turns of resistance wire terminated by threading through a hole in the centre of the fibre former from the outside with one end connected to an insulated terminal and the other end unconnected
Yours, Richard

[SurreyNick]

Thanks for the pictures Mike they are exactly what I need.

I am a little nonplussed by the variation in resistances. Bill (AC/HL) has a 1927 model and his is 3Ω, yours is 30Ω and Richard's is 4.5Ω. The only logical explanation I can come up with is that constructors used whatever was available or at hand. The unit was entirely DIY and all Cossor produced was a construction broadsheet. I have had a good look at my scans of this broadsheet and nowhere on it did Cossor specify the resistance required!

I don't know how to determine the value of the component, but it will have a bearing on the gauge of the wire I can use. Any idea how I would work out what value resistance to use?

Thanks

Nick

PS: Richard, thanks for the dimensions

Any idea how I would work out what value resistance to use?

As a first hit I would use the same resistance as all the valves when running. If the whole lot came to 500mA at 1.8V then that would be 1.8/0.5 = 3.6 ohms, a fair match to two of the others. This was my first guess calculation and not reverese engineered!

[SurreyNick]

I have had a good think about the dimensions and I think making a replica is going to be do-able.

I still have a lot to learn so let's see if you agree me with me?

With a former diameter of c.50mm (c.2 inches) the circumference would be 157mm and allowing 60% of this for a winding length I have 94mm to play with. At 13mm wide (c.1/2 inch) and 2mm thick each loop of wire around the former will be 30mm long.

24 gauge Nichrome wire is 0.56mm in diameter (1.8 turns per mm) so wound touching this will give me 168 turns on my 94mm. If I wind with a gap however, using another length of 24SWG wire to separate the loops (which I later remove) I should be able to get 84 turns on my 94mm length.

84 turns at 30mm each gives me a total length of 2520mm. 24SWG Nichrome has a resistance of 4.56Ω per metre. 4.56 x 2.52 metres gives me a resistance of 11.5Ω. Yes?

Thank you merlinmaxwell. If I only need 3Ω I could get away with 20SWG!

I will make the former from card coated in Shellac. This will enable me to give it a permanent curve.

Nick

[AC/HL]

I'm afraid I only have a simple camera, so that's the best I can do.
It's Bakelite, 8cm wide, 4cm high and 1.5 deep.
It was an internal adjustment, so would presumably have been used in the initial setting up.

[SurreyNick]

Thanks Bill.
Although they are a bit blurred they are still good enough for my needs.
Much appreciated
Nick

[Cobaltblue]

Yes the Resistance is a bit of a mystery.

I had a quick look at the 1927 Cossor Melody Maker sheet I have here and also a later document (1933) from Cossor recommending valve replacement types for the 1927 Cossor MM and in both cases it shows options for 2V, 4V and 6Volt types.
It could be that the Resistor was chosen dependent upon the valves chosen.

I have another two 1927MM's but they are both in Exeter I would now like to see what those read.

Cheers

Mike T

[SurreyNick]

It's a thought Mike, but adopting the formula proposed by merlinmaxwell will it make much difference? I don't have data for the 4v and 6v valves but is the current likely to be very different?
Nick

[SurreyNick]

I must say Bill, having had a closer look at the pictures of your Burton Components variable resistance it really is a very fine example. I imagine it was a 'premium' product in its day. I can tell you now, I won't be trying to replicate yours

No, I think I will attempt the much more straightforward design of Mike's (albeit using Bakelite for the base). The only thing I'm going to have to hunt around for are two small binding posts with knurled nuts. The rest I should be able to fabricate reasonably easily.

NB: Does anyone know where I can buy a small length of Nichrome wire?

Nick

[ThePillenwerfer]

There's plenty on e-Bay.

- Joe

[SurreyNick]

So there is!
I should have checked first
N.

[Ed_Dinning]

Hi Nick, let me know what you need in terms of length and resistance and I'll put some in the post for you.
Note that there were many resistance alloys; the same wire dia will not always have the same ohms/ foot

Ed

[David G4EBT]

You can buy 2 Metre lengths of nichrome resistance at the link below.

24 SWG is 4.4 Ohms/M so 2 M = 8.8 Ohms.
22 SWG is 2.7 Ohms /M so 2M = 5.4 Ohms.

22 SWG might be a good choice as 3 Ohms would be more or less midway along the track, so you’d have some leeway either side to reduce or increase the resistance.

2M of 24 SWG is £2.40 post free.
2M of 22 SWG is £2.19 post free.

http://www.ebay.co.uk/itm/Nichrome-N...p2054897.l4275


22 SWG = 0.711mm diam / 0.0280" (28 thou)
24 SWG = 0.558mm diam / 0.0220" (22 thou)

As to a former, rather than use card treated with shellac, I think a couple of layers of thin paxolin glued together with rapid setting two-part epoxy might be a better bet. I’ve got some 0.85mm thick paxolin (woven fabric type – not SRBP) and to show its bending qualities have attached a pic of a piece I’ve cut 15mm wide, happily curled to 1.5” diameter. A couple of strips glued together would be 1.7mm thick, and three layers would be 2.55mm.

For comparison to imperial sizes, 1/16” = 1.59mm, and 2.55mm = 1/10”. Hence, this paxolin would happily conform to the dimensions of the former stated by Mr Moose in post #6.

(The thickness of the former in post #5 looks to be more than 1/16”).

Let me know if you’d like a sheet on the paxolin – it’s about 25cms x 20cms. You’re welcome to a sheet FOC.

As you said earlier Nick, when winding on the turns it would make sense to space the wire by using a thinner gauge wire alongside it, then removing that wire when completed. If there is any sign of movement of the turns when completed, it might be a good idea to use a thin layer of expoy cement around the inner perimeter to secure the turns to stop the wiper moving them when it’s adjusted.

Depending on how posh you want to make it, it doesn’t look too demanding to replicate one that looks in keeping.

Good luck in your endeavours!

with rapid setting two-part epoxy

I prefer the slow set stuff, gives time to get it right and sets harder. For a 90 year old set the added day won't matter much. And if you warm it up (with a hairdrier or somesuch) it flows before it sets giving a 'varnished' look.

[hansomcommon]

If I am not to late I have found this original NOS 5ohm unit which I think is just what you need, pic enclosed. Please PM me if it is OK for you.