Friday, December 21, 2018

High impedance vintage headphones

Venturing into a new domain with 2000 Ohm resistance per phone.

Something new to discover in the area of vintage 20th century technology. These headphones date from the 1920s - with their very high impedance of 4000 Ohm in total they're meant to be used with a radio of the period. Most likely a crystal wireless set, or perhaps even an early (expensive!) valve radio.

This particular set of headphones survived what are probably decades of disuse in good shape with the nice leather-covered headband fine too. The metal parts clean up nicely, with the odd bend gently worked out.  The earcups are probably not bakelite, but hard rubber. Molded on the inside is the Ericsson wordmark - so these are Ericsson headphones (or at least a make that used phone speakers made by Ericsson).

Nevertheless when they arrived, the phones were broken. This is easily tested by tapping the leads on a battery - the DC pop and crackle of even a 1.5V AA cell will be clearly audible. These headphones however remained completely silent. Open circuit somewhere.

From a browse-around on the web, it becomes clear that these can often have some broken connections - generally where different gauges wire have been spliced together. Also from that browse-around, some instances where the insides get replaced by modern (cheap) drivers - sometimes not even soldered in place, but a gash sawed in the cups (<gasp>!) to get the wire through. What also is clear from some reading up, is that this type of headphones can usually be repaired - like much of the technology of the period, it should last.

Internally, the headphones have four 1000 Ohm coils all in series. By unscrewing the earcups - these come off easily - the iron diaphragms with their gasket-ring come off to give access to the inside. Methodically measuring the resistance from point-to-point, the defective coil was easily identified. Undoing the screws that hold the assembly of coils and magnet in the aluminium cup, the set of coils with the defective one was taken out for closer inspection. Clipped together to not break the connection between the pair, the paper wrapper can be gently peeled off.

Unfortunately, the start and end lead-wires were both still fine and correct - so the break must be inside the coil. (Thought it odd that, with 4 kOhm I should think that not even hooking them up to the mains could damage a coil...)

Then to proceed to look inside the coil, nothing else to do but to unwind it. With a quickly rigged-up little coil-winding tool - to not get all tangled up in the very thin lacquered magnet-wire - unwinding begins. (Coil winder made with Meccano of the same vintage - sort of a 'mechanical breadboarding' toolkit.)

Luckily after only a few meters of unwinding, another paper-tape starts to appear. The wire for this coil was apparently spliced halfway winding. Probably during manufacture, when one spool ran out and another was placed onto the machine.

When gently peeling off this paper tape, the wire-end just fell out. This very likely was the original break - some checking up with the multi-meter confirmed that the rest of the coil-windings were fine.

Gentle rub with fine-grit sandpaper to remove the lacquer, then twist the ends together. To make sure, a small drop of solder to keep it all tight, then a new paper (masking) tape to hold the connection safe. The paper-tape holding the new splice is then wrapped around the coil and the rest is re-wound too. The thicker lead-wire is spliced on in the same manner, wrapped around and held in place with a small extra bit of paper-tape. And thus a fixed coil. A little spot of paper-glue and the original wrapper is placed back too.


The lead-wires are then soldered back onto the eyelets that are screwed to the terminals in the cup and the whole assembly comes together again in its aluminium cup.

Carefully placing back the diaphragms (with gaskets) and screwing the earcups back on, the headphones then simply snap back into the forks of the headband. Making again a complete 1920s set of headphones that are now functioning fine.

These should last for decades yet - now for a crystal wireless set :-)

In the meantime; with a quick soldering of some sockets onto a 3.5mm jack the headphones can already be used with any modern music-player or phone. With the very high impedance of the headphones, the volume must of course be set high for a decent, still modest sound volume. The sound is a bit 'tinny' perhaps, but it does a very good job of reproducing the period music - voice comes through very crisp and clear. Also despite being relatively heavy and with hard earcups, they're not as uncomfortable as you'd perhaps expect.

So not quite modern 'hi-fi', but surprisingly good headphones from the 1920s.


  1. Great work on the headsets. I've gotten quite a few of those over my years of playing radio. Really good ones do not even need a battery; just short the wires together, and if you hear a click you've got a set of working Hi-Z headphones. If you want a real challenge make a Gillette Blue Blade radio with a safety pin for the cat whisker and (if you can find the wax paper & foil gum wrappers to wind your own capacitor, and make a razor blade crystal set like the prisoners in the war. Then there is Galena and a cat whisker or the modern way with a 1N32A or other RF diode.

  2. That's some nice fixin' for your 'phones. Hope you find a nice crystal set to listen to (:

  3. Tnx; it was very rewarding to hear the fixed phones :)
    Before trying with an actual Galena, I think I'll try to get things working with a suitable packaged diode first. Had a look around at old crystal sets, but these seem to have soared even higher than nice typewriters! Will probably have a go at constructing one (my understanding of electrics is 'rudimentary', but should be ok for 1920s stuff ;-)