[klondike98]

I've started looking into the collection of radios and speakers from a local museum. The Thompson Speaker looked simple enough to start with.
[Image: https://www.dropbox.com/s/qjwtkayltquxnz....jpg?raw=1]
[Image: https://www.dropbox.com/s/o42z9erj8e7a6w....jpg?raw=1]

The wires for connection to the radio are missing so the bottom looked like a good way in. The screws were rusted in so an overnight soak with WD40, a slight tap with a hammer on the screw driver and they broke loose. Unfortunately the base also seems to be glued on somehow.
[Image: https://www.dropbox.com/s/91z06qcpnb44yl....jpg?raw=1]

I attacked it from the top next and was rewarded with opening it up and being able to get to the mechanism.
[Image: https://www.dropbox.com/s/gs7pq5nunv6c5x....jpg?raw=1]
[Image: https://www.dropbox.com/s/v9hz91xhrp48wg....jpg?raw=1]

The actual audio mechanism looks pretty good so I'm hopeful it will work OK. The base is rusty so it will need some cleaning later.
[Image: https://www.dropbox.com/s/ao9li586ykbb5z....jpg?raw=1]

Replacing the connection wires looks simple so thats next and then a test. I've never worked on these battery powered radios with this type of speaker before. Any thoughts on a way to test the speaker? Perhaps just connecting the audio output from my signal generator would work (without the RF signal)?

[morzh]

If it is a 2-wire speaker, then yes, hook up to an audio generator and try to see if it sounds.
Bring up the level gradually.

[klondike98]

Yes it’s a 2 wire speaker. Thanks Mike.

[MrFixr55]

These are generally high impedance (in the order of 500 Ohm - 2 KOhm). Ohm it. if very low, remove the capacitor. it may be suffering from silver mica disease.a 1.5V battery can be used to click test it. If you have an output transformer from an AA5, hook the secondary to an amp and the primary to the speaker wires. Slowly turn up the volume. Of course, you can test it with a signal generator. These speakers are certainly not high fidelity.

If hooking up to an antique radio where the speaker connects between the output tube plate and B+ then polarity is important. The wrong polarity will demagnetize the magnet. If it is cap coupled or transformer coupled, then this is not an issue.

Good luck.

[klondike98]

Thanks MrF. It was indeed connected to either the Crosley Trirdyn or the Atwater Kent 32 both of which have the speaker connection at the output tube plate and B+. How do I determine the correct polarity for the speaker wires?

[MrFixr55]

Hi Bob,

I will have to research to see if I have actual instructions from a manual, but I believethe following procedure should work well:

Some magnetic horn speakers have a "volume control" that allows the user to move the coil and magnet assembly closer to or further away from the metal diaphragm. The knob on the top of your speaker base assembly is likely the volume control. Move the coil as close as possible, almost to rattling caused by the diaphragm hitting the poles of the magnet. Then connect speaker, reverse polarity and play again. the polarity that causes the most rattling is the correct polarity. You can fine tune till it rattles with only 1 polarity. That is the correct polarity. Upon determining the correct polarity, readjust the "volume control" on the speaker to loudest volume without rattling. If there is no volume control adjustment (gap adjustment), the polarity that gives the loudest sound is the correct polarity.

Your electromagnet assembly may be marked with a + sign or a red dot. Otherwise the existing cable may have a conductor with a red stripe in the cloth insulation. That will be the connection to the battery.

In some cases the magnet will have completely lost its strength. These are close to or over 100 years old and were not very High Tech to begin with (Although yours looks much more robust than other horn speaker magnets). I have not had to try to re-magnetize one, but one can try. Some people merely add a modern pill-sized neodymium magnet to each end of the "horseshoe".

BTW, I have an A-K 32 that I have to get around to fixing (although it may be good to go right off the bat as the pulleys and belts on this are good). My collection includes an A-K 10C, 20, 20C, 30, 32, 35 (2), 36 and 40. I have a Crosley 601 also, but the tuning caps were made of pot metal, so thay are a total basket case. When I get around to it, I will fit with A-K caps but I need more pulleys and belts, as my 30 and both 35s are missing these.

[Chas]

When the diaphragm speaker is used in a plate circuit with current flowing, generally not over 10ma. The correct polarity is when the diaphragm is pulled inward. In that orientation, a couple of things happen, the magnetic field is increased causing an increase in sensitivity, the diaphragm will rattle so it is adjusted not to.

If the polarity is reversed, over long term, the magnet is depleted, sensitivity is reduced, to reach greater sensitivity in the reversed condition the diaphragm would have to move closer.

Under the correct pull-in current flow conditions the terminal that is to B+ is identified, RED or "+" any identified tinsel cord will have a red or other marker thread for B+.

Gaskets that support the diaphragm are almost always either hardened or missing, without the gaskets buzzing will occur despite polarity.

If the radio to be used with this type of speaker has a 71a output tube, then the horn speaker is couples with a 1 mf cap and a 30hy choke to B+. A period correct audio output transformer would work as well.

Not all horn speakers use the single ended diaphragm, others will use a diaphragm pin driven "motor" these are not polarity sensitive, the diaphragm may be metal, or in some Baldwins, mica. A mica diaphragm in good condition is much more "fuller" in tone.

GL

Chas

[klondike98]

Thank you, very helpful! The AK 32 has pot metal pulleys and are not any good so they will need new ones. I have some coming from eBay but still need one more but that’s another story.

So red or + goes to the B+ battery if I can locate that on the mechanism or wire. Otherwise, test as you suggest to establish polarity.

[klondike98]

Thanks Chas. The horse shoe magnet still has some magnetism in it, as my screw driver is pulled into it. As to the rest I will have to look closer, probably more pictures for you guys! The output tubes on the two radios this would have been with are a 112A and a 201A.

[klondike98]

Thompson speaker update:
The good news is I can see what used to be a red tracer color in one of the wires so would assume that's the + side. The bad news is that checking the ohms across the wires (also checked at the connection points) shows a complete open. I then checked between one lead and the center connection between the two coils shown with the arrows, also open from either end. My conclusion is that both of those coils are open and I think that may be a very difficult fix. I suppose I could disassemble everything and try to rewind the coils Has anyone tried anything like that?

[those lines on the bottom and top right are reflections of the window blinds in my shop. I was using a magnifying lens to light and magnify the image]
[Image: https://www.dropbox.com/s/2mvu3s5fwvw6lt....jpg?raw=1]

I decided to take a quick look at the Stewart Warner 415 speaker and found that the resistance across its two leads was 5 to 7 Meg so something isn't right there either. Getting into this speaker is not as easy as the first. The front and back are fitted onto the body and secured with two rivets. The back was missing one rivet and the other was partially out so I was able to work it out and get the back off. Unfortunately that doesn't give me much access. The workings are on the other side. I was able to re-seat the gasket however. I now have to decide whether to try and dremel grind off the rivets for the front without damaging it and whether its worth it or not.
[Image: https://www.dropbox.com/s/41r4kom1enh9yd....jpg?raw=1]
[Image: https://www.dropbox.com/s/51rvq8yfgnwifu....jpg?raw=1]

[Chas]

Bob,

Open coils are the bane of us radio collectors

Although this particular speaker uses windings in glassine layers, plain scramble winding will work with some loss of high frequency.
Need to make any kind of suitable winding jig that can support the bobbin, of which may have to have "cheeks" made to prevent fall of the winding during the process. The wire size will be at least #40 possibly smaller. Manual tension should be used so it will take practice. Essentially, fill the bobbin. Terminating the wire with .025 solder braid such that the braid is locked under Kapton or similar tape.

There are a LOT of ideas here:

https://www.radiomuseum.org/forum/rewind...hones.html

Especially important is the winding direction in respect to each bobbin, if would incorrectly the forces will be in opposition to that of the magnet.

I would expect that each coil be at least 500 ohms DC resistance, preferred would be 1K DC resistance total for the two coils of 2K. That very roughly equates for a tube plate impedance of 10K ohms.

Replacement with a modern speaker and an output transformer is often met with very low volume and distortion as there is usually a mismatch of the output transformer and very low efficiency from a moving coil speaker.

Chas

[MrFixr55]

I am not sure of the brands, but I have seen speakers with ships and dancers bring a fairly good price on eBay.

Re the coils, the only thing I can recommend is that you can peel away the paper wrap and inspect the coil for a break, start unwinding and hope that the break is only about a few layers down. Otherwise rewind per Chas

+1 on Chas, the "Magnetic Reed' speakers, the coils are not wound around the magnet, therefore polarity is not as much of an issue in terms of losing magnetism.

I have used an AA5 transformer and modern speaker. However, this may not work well with anything less than 71As at 135V or higher, as 201s and even 112s put out <35mW at 90V, but works well on my Radiola 17, 18 and 60, as the 71A puts out about 0.7W at 180V.

[MrFixr55]

How does the reed attach to the cone? Interesting that the cone is held in place by 3 screws, the only screws in the unit?

[klondike98]

Thank you very much guys!! Appreciate the opportunity to learn from your experience. I may have confused things by talking about 2 different speakers in the same thread so I’ll move the Stewart Warner discussion to a new thread. I’ll have to think about the rewinding a bit before giving it a try. The more I consider it I don’t think there is much to lose in giving it a try. Restoration of the exterior can still take place. Will post any progress

[klondike98]

The Thompson speaker has undergone a disassembly to further assess the coil windings. I've taken a few photos as I went along. At first I thought I could get the coils off without a full disassemble but later realized it was all interconnected and one of the screws that needed to come off was completely inaccessible without taking everything else apart

[Image: https://www.dropbox.com/s/afjot906kha2tf....jpg?raw=1]
[Image: https://www.dropbox.com/s/91j3yr509v9tab....jpg?raw=1]
[Image: https://www.dropbox.com/s/kvgi993iijxxrg....jpg?raw=1]
[Image: https://www.dropbox.com/s/re08m6g44yyzhv....jpg?raw=1]

Both of the coils are wound around an iron core that is connected together and the angle you see in the other photos so it would be difficult to wind directly onto the core. The wire does slip off the core so I suppose I could make a cardboard form to start the winding. This seems like a very big task. I need to think more on it