How hard can it be to make a Reed / Armature Speaker?

I had pondered this even before getting a 1929 radio where the reed speaker had been replaced with a transformer and moving coil speaker.

This article is a good introduction to Loudspeaker Development.

First we need one or two convenient to use coils, probably about 1000 to 4000 ohms depending on if one or two.. I looked at a Daisywheel driver (too low). The actuator coils in €2 clocks look promising. Small and about 350 Ohms each. Using four as two pairs in a 4 pole balanced armature or cantilevered "motor" might be feasible.

I have also a 2000 Ohm P.O. Relay coil and electro-mechanical bell from a phone (about 1200 Ohms), with a single coil.

Experiment 1: Using an unmodified telephone bell mechanism.

Finally I was given a slightly damaged "extension bell". These put less load on the phone line so are higher impedance. One terminal had pulled out snapping the fine enamel wire. But I was able to repair this. It's two coils of about 1800 ohms either side of a bar magnet. The armature is pivoted with a spring above the bar magnet and the two end above the poles at the two coils.  The pole & magnet to armature distance is adjustable. A long rod attaches to centre of armature plate and has the "hammer" for the bell on the end. I took it apart to see if the rod came off or could be reversed, it can't as it's riveted.

I made a shallow cone from A3 size 80 gm paper.

I glued and tie wrapped an emery board to the "hammer" to mount the cone. I mounted it inverted as that seems to be how the UK  late 1920s Portable Radio sets do it.

It really needs stiffened, But this is just the FIRST experiment.

Here we see the ABSOLUTELY un-modified bell mechanism with cone support tied & glued in place.

It seems to naturally sit to one side. So I connected the bench PSU and established that about 10V to 14V of the "correct" polarity would bias the plate (armature) equally spaced. I marked a + (red + visible).

 

So does it work?

My McMichael Super Range Portable Four conveniently has a 1/4" jack socket between Anode and HT that disconnects the internal reed speaker (or in my case the replacement transformer and moving coil speaker).

Amazingly it works quite well. The "motor" needs to be held so that when the set is on (DC Bias moves plate) that gravity centres the arm.

Did Telephone bell mechanisms inspire the more complicated "reed" speakers? Obviously the original kind is developed from the simplest single pole moving iron earpiece.

 

Next stage

Fabricate a "proper" armature and pole arrangement of the 4 pole balanced kind or the cantilevered (or do both).

Figure how the mechanical adjustment to compensate for bias works. It has to be adjustable as the bias current of the output triode will vary with state of A, B and C batteries condition and valve (tube) ageing,

Make a pair of full size (for the McMichael SRPF) cones, one washed wiith PVA and one sprayed with Acrylic paint to compare stiffening methods.

To thank the Author because you find the post helpful or well done.

I've never read of any connection between telephone ringer and reed speaker designs.  One would expect ringers to work poorly at high audio frequencies since they were intended to work around 20Hz and have a lot of moving mass.  But you can't argue with results.

 

The first balanced-armature headphone was invented by Nathaniel Baldwin about 1915, but the later reed speaker drivers are not fundamentally different.

To thank the Author because you find the post helpful or well done.