[kropotkinlives]

Indeed. I tried to make a transistor with the reverse side of a pile of 74LS IC dies I extracted with solvents, household chemicals and a blow torch. I managed to make an acceptable diode that lasted about 2 minutes at a mere 200uA of current. I gave up then and decided that the transistor was the base abstraction layer I could be bothered with.

For ref diode recipe:

1. 74LS die. Turn it over.

2. Small pile of borax on one half. Small pile of sulphur on other side.

3. Apply torch until everything is baked nicely.

4. Scrape off surface with a razor.

5. Poke two pins connected to your circuit until you find a bit that works like a diode (can take a few minutes). I used a simple home made curve tracer out of a twin-t oscillator, buffer and an oscilloscope. Don't hit it with much current or it'll kill it instantly.

[jacquesm]

You can do this with a lead crystal as well (that's how in the old days the diodes for crystal receivers were made).

Even a dirty razor blade and a pencil will work as a diode!

[kropotkinlives]

Yeah that was where I started with the idea and then decided to see if I could dope some silicon to make a diode then move up to transistor level, then a simple IC.

You could knock up a point contact diode or even transistor without too much pain but that's not as much fun :)

There's info on how to make point contact devices in here: http://www.qrparci.org/wa0itp/csts_book.pdf