Historical cat's whisker designs

Crystal radios from the 1920s are still possible to get hold of, but can be quite expensive. They contain a wide variety of approaches to mounting a crystal (usually, but not exclusively galena) and the cat's whisker, and attempts at making them adjustable and adding stability at the same time. This is a collection of photographs and diagrams I've managed to find online. The materials used correspond with what was available at the time, glass and metal but also the first plastics, ebonite and bakelite - which would have felt very futuristic in these years.

On the right is a diagram of a vertical set up found on radiomuseum.org. This design doesn't seem to allow for positioning of the point contact, only the overall pressure - which to be fair is generally more important.

Historical variants found on ebay listings

Ebay is a good resource for images of different variants of cat's whisker apparatus. This is the galena point contact diode described in Semiconductor history and is a typical form found in (usually I think) British made crystal radios.

It's a horizontal setup and protects the crystal with a glass envelope. You can clearly see a direct link to modern components like this germanium diode, which still uses the point contact method - but is fixed and much smaller:

Mister rf, CC BY-SA 4.0 <https://creativecommons.org/licenses/by-sa/4.0>, via Wikimedia Commons

Here is another variant of horizontal construction, there were probably as many varieties of these as there were companies manufacturing radios (several hundred) including home made designs. Sometimes the crystals are fixed mechanically with set screws, more often with wood's metal, a toxic alloy that melts at a low temperature to avoid altering the crystal material.

The other common approach was to mount the crystal at right angles to the 'arm' and allow it to be adjusted in a similar way, but pointing down on the crystal. These tend not to be enclosed in glass, and seem (as far as I can tell) to be slightly more common in the USA.

A vertical arrangement, that seems (although it's difficult to tell) to align two crystals together, presumably a perikon like setup. "Perikon" was the name given to combining chalcopyrite and zincite, which apparently provided a very effective diode behaviour.

Multi crystal sets

Many radios feature multiple crystals and a switch to select which one to use. This is presumably an indication of their unreliability and drift over time - something I have also experienced in use in audio circuits. You could use one for a bit and then switch to the other, perhaps for different frequencies - although I feel the advantages may have been more imagined if they were using galena for both. Given that the mechanism causing semiconducting were not understood at the time, this may have been understandable.

This is a piece of equipment specifically for testing crystals. It's unclear but possibly the one on the left is for testing two crystals brought into contact with each other.

Sometimes different types of crystal could be switched between in a radio. This Marconi Marconiphone "Crystal A" type RB3 radio set from 1922, which seems so different in it's construction methods to those of today it almost invites a comparison to be drawn to quantum computing technology. It has two rectifier options: galena, as we have seen already - and carborundum (silicon carbide), which is very rare in nature but is created via industrial processes for many uses due to its relative hardness.

Instead of a cat's whisker, the carborundum is fashioned into a point that makes contact with a metal plate - so the inverse of the galena point contact. The contact pressure can be adjusted with a thumb screw - this image show the parts disassembled. It seems a common feature of these devices to be easily taken apart and put back together, something we are only now rediscovering is important.

This radio also features something very unusual - a battery! This was used to provide a DC bias voltage to improve the galena crystal. I think to push the conductivity into the active region to remove the voltage drop that stops it passing current below a certain threshold. In my experience this varies across different point contacts, so would have been tricky to 'tune' correctly.