Hand craft transistors...

[Gilles C]

Was this the secret?

These OC44 were "almost" made by hand.  

http://www.thevalvepage.com/trans/manufac/manufac1.htm

[Joe Davisson]

It would be easier to try to copy the transistor from the Lillienfeld patent #1,745,175. It operates kinda like a FET, but uses metal-oxide. I made a point-contact version and got it to bias, but it was horrendously noisy and didn't really do much amplifying. The noise level was higher than the signal It was a basic test with lots of problems, though.

Had I actually followed the directions it might have worked better. The guy filed multiple patents on similar contraptions indicating that it should work if built right. He apparently thought it did.

-Joe

[petemoore]

From Ge reads I found an article on the early production models that were sliced from a chunk'o Germanium, and that later a process of reducing the Ge to a fine rubble, removing [what] oxides or impurities were in it to some extent, the pressing that into wafers, these sliced or pressed wafers were then made into transistors.
 I wonder if the OC44 or other sought after types were the sliced type.[?...]
 Seems the Ge in the earth is often riddled with impurities, I suppose the slicing technique is all but forgotten, finding the chunk of Ge with the right properties certainly doesn't lend itself to fast production or consistancy, I'm wondering what the different tonal qualities are between the sliced and pressed ones.
 Am I understanding that you actually Made a transistor? if so WOW, do you have any PICs?

[Paul Perry (Frostwave)]

So much for transistors.. who is going to be the first to make an OD using catwhisker diodes?? I can just see being down on the floor at a gig trying to find the 'sweet spot' on the galena crystal lump

http://www.acmi.net.au/AIC/CRYSTAL_SET_3.html

[Joe Davisson]

I don't have pictures, what I tested wasn't in a permanent form. It was a point-contact FET which biased but didn't really amplify. The goal was to try to obtain the action presented in the patent, without going through as much trouble.

Apparantly the right way to make it involves using a pure oxide coating and heating it so some of the gate metal dopes the oxide. I've made metal-oxide diodes which were once very common, using Cu/CuO and Ti/TiO2 combinations. The titanium version is interesting because the forward voltage is rather high, measuring in at about one volt. And it has the strange ability to rectify about a volt from your hands, even though you can't really measure that much from a voltmeter. Might try some different combinations with this FET idea sometime.

Funny that the guy invented the FET long before the bipolar transistor came out. Maybe the performance was too far below electron tubes at the time or it had flaky reliability, etc. Wouldn't it be cool if they sounded good?

-Joe

[Hal]

talk about doing it by hand, one of the guys in the lab im working in this summer is doing some project with spraying thin films in christiline structures and trying to make transistors.  If anything works, i'll ask him to take one home, and put it in a guitar effect :-D (jK)

wonder how they would sound, though.