Make Your Own Point Contact Transistor

Beyond the power variant, it sometimes seems as though we rarely encounter a discrete transistor these days, such has been the advance of integrated electronics. But they have a rich history, going back through the silicon era to germanium junction transistors, and thence to the original devices. if you’ve ever looked at the symbol for a transistor and wondered what it represents, it’s a picture of those earliest transistors, which were point contact devices. A piece of germanium as the base had two metal electrodes touching it as the emitter or collector, and as [Marcin Marciniak] shows us, you can make one yourself (Polish language, Google Translate link).

These home made transistors sacrifice a point contact diode to get the small chip of germanium, and form the other two electrodes with metal foil glued to paper. Given that germanium point contact diodes are themselves a rarity these days we’re guessing that some of you will be wincing at that. The video below is in Polish so you’ll have to enable YouTube’s translation if you’re an Anglophone — but we understand that the contact has to be made by passing a current through it, and is then secured with a drop of beeswax.

A slight surprise comes in how point contact transistors are used, unlike today’s devices their gain in common emitter mode was so poor that they took instead a common base configuration. There’s a picture of a project using three of them, a very period radio receiver with bulky transformers between all stages.

If you’re interested in more tales of home made early transistors, read our feature on Rufus Turner.

Thanks [Dr.Q.] for the tip.

21 thoughts on “Make Your Own Point Contact Transistor

  1. Given that germanium point contact diodes are themselves a rarity these days we’re guessing that some of you will be wincing at that.

    There’s somebody on eBay selling 5 gram chunks of what’s claimed to be five-nines germanium for $20 US, and multiple other sellers with not much higher prices. Ripping up diodes seems like the hard way, especially if they’re rare. Although it looks like you can get the diodes for under a dollar in moderate quantity…

          1. The most useful Pu isotopes (238, 239) emit essentially just alphas, with a tiny, tiny fraction (about a millionth) of the energy coming out in < 100 keV X rays, but no gammas to speak of.

            Still, I wouldn’t want it in skin contact. Give me a couple of centimeters of air or a layer of plastic wrap.

  2. A 200 pack Ge diodes (real) of russian D9, or a 20 pack of D2 go for 20-30€ respectively and these series have sub-types such as D9G, D9B, etc. Many outperform or equal western Ge diodes. The boxes have 2000 diodes, but to make a profit the BLOC countries sell 200pcs at the aforementioned prices. I have about 200 D9, 20 D2 and a mix of 20 D9 of all types.

    Useful circuits are rare, but a half-wave rectifier for an IR transimpedance amplifier for SPDIFLink is trivial to breadboard from 15 parts. At 25 MHz Bandwidth and data rate of 12.5 Mbit/s (SPDIFLink is 50% duty cycle) .

    Useful to exotic circuitry.

  3. I don’t know if this is completely relevant to the subject at hand or not, but I found that I could make a usable point-contact diode by heating a box-cutter blade until it had a nice layer of blue oxide on it, then using the tip of a needle for the contact point. I say “usable”, not necessarily “good” here, but it was an effective enough diode for receiving AM broadcast radio.
    I bring it up because, as this uneducated buffoon understands it, a diode is essentially 2/3 of a transistor. It might be another potential route toward achieving a DIY transistor, were someone to be so inclined.

    1. I did this as a kid following some kind of …
      guide? Hadn’t remembered until you mentioned it. Neat off the shelf demo at least, but I think the night be a fair amount of meat here for hands-on science education as well as DIY tricks.

  4. FWIW and prolly a bit off topic…. The advent of ICs and dev boards with powerful processors complete with peripherals are absolutely fantastic – a complex project can be got up and running in hours as against hand wiring and soldering with discrete components, sometimes taking days or weeks to complete ……
    At the risk of being labelled a neo luddite, I lament how much understanding of basic electronic theory is missed out on.
    As an example, I was completely flabbergasted by a “Mechatronics Engineer” who misunderstood how a transistor worked, no idea of biasing and the effects of capacitance and noise on long signal wires….. He managed to turn a simple task of creating an optical “paper out” detector into a week long saga using a dev board with very unreliable performance….. Total cost of $3+ k, most of that being his salary. Ended up showing him the door and doing it with a few discrete components in a couple of hours……
    Rant finished.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.