Transistor Fabrication: So Simple A Child Can Do It

If manufacturing printed circuit boards has become too easy you should try your hand at producing transistors. [Jeri Ellsworth] put together a collection of videos outlining the process. These go way beyond the IC fabrication we saw from her in the past. It doesn’t take much, a 1000 degree oven with steam option, silicone wafers, and a variety of chemicals. We’ve embedded the instructional video as well as two demonstrations of her N-style FET after the break.[youtube=]



60 thoughts on “Transistor Fabrication: So Simple A Child Can Do It

  1. This is silicon:

    This is silicone:

    Sorry, but Hackaday really needs to up its game if it’s going to ever aspire to any form of professional reporting on the subject of hacking. There are spell-checkers you can get for almost every browser, yet I routinely see misspellings here. How can anyone confuse the hard, crystalline form of the element silicon (the basis of almost all of the technology Hackaday routinely reports on) with the rubbery substance used to caulk windows or to give (older) breast implants their jiggle?

  2. tell me when you get a silicone cpu working THEN i will be impressed

    andar: i prefer silicon, silicone is ONLY good for sealing gaps, silicon makes the digital world work though!

  3. At home? My impression was, that this lady is running an electronics company. It doesn’t surprise me, that electronics companies in 2010 are able to produce crappy transis in their even crappier labs.

  4. it is too short, it is irrelevent to the post(mostly) and it doesnt even bash on arduinos!
    if you had posted
    “jeri is hot but she is a supporter of arduinos which i do not condone at all and therefore her hotness is diminished” it would have been accepted and not deleted

    Jeri Ellsworth (born 1974) is an American entrepreneur and self-taught computer chip designer. She is best known for, in 2004, creating a Commodore 64 emulator within a joystick, called Commodore 30-in-1 Direct to TV. The “computer in a joystick” could run 30 video games from the early 1980s, and was very popular during the 2004 Christmas season, at peak selling over 70,000 units in a single day via the QVC shopping channel.[1]


  5. Be very very careful with the HF people if you want to do this… FWIW I think you should by the NMOS for maybe 2 cents. Unless you are going. Then again I do this for a living more or less.

  6. Also for what is worth to P silicon indium solder is an excellent choice. You’d want PMOS I guess for that, there are Boron based spin on glasses…

    For this size device indium solder would work well to either N or P type, you aren’t exactly worrying about matching work functions here…

  7. Jeri Ellsworth rules! The whole point is to show that a working transistor can be made by the average hobbyist who doesn’t have access to a modern state of the art commercial lab. I haven’t seen such a clear, experimenter-oriented explanation of this topic anywhere else.

  8. @ mrgoogfan

    I’ve handled 6 mol HCL in lab plenty of times. If you’re an idiot, you can get burned. If you pay attention, there is no problem.

    Also, NaOH and KOH are just as dangerous (although at the other end of the pH scale) and chances are you have already handled it – the it’s the active ingredient in Draino.

  9. @Natrix

    This is HF, not HCL. The problem with HF is that it often will _not_ burn, so you don’t always know if you’ve come in contact with it. Instead, it easily goes through latex gloves and your skin directly in to your body. Then it reacts with the calcium in your blood, and you die of cardiac arrest. HF is really nasty stuff. Don’t confuse it for HCL or you’ll be in trouble.

  10. @ Natrix

    HCL, NaOH, and KOH only cause external chemical burns, and that damage is mitigated by the fact that you can feel it burning you and wash it off.

    Very strong HF can burn too, but it’s generally a weak acid. Skin exposure may be underestimated or even unnoticed, because of the lack of burning sensation. The real damage comes from HF’s ability to pass through skin, where it:

    * reacts with the calcium in nerves and bones, forming insoluble fluoride crystals which cause excruciating pain and interferes with healing
    * reacts with free calcium ions, causing hypocalcemia which shuts down the heart
    * causes general fluoride toxicity

    By the time you feel anything, washing is usually insufficient and you’ll need medical treatment, consisting of topical and injectable calcium gluconate. Skin exposures of more than 1 sq. foot may be unavoidably lethal – even with thorough immediate washing and transport to a hospital, your heart may stop before you get proper treatment.

    HF deserves special respect.


    HF is evil!!!
    It not acid evil it is nasty evil…
    A drop of diluted HF on your skin will be absorbed rather quickly. Maybe even without your notice. Maybe it will not create great damage to your skin directly… your skin would be red and itchy… but nothing to be concerned (you think)… then after a little time it will become nasty since it starts to destroy tissue under your skin and attacks bones …. and even worse HF is a giant monster calcium and magnesium eater. It sucks it out of all parts of your body (6ml are enough for a adult body) and create serious toxic effects which might ends up in a cardiac arrest (your heart stops working!!), due to hypocalcemia.

    But not enough… its eating up glass and metals which you might use to keep, handle or store it. And it is rather easily volatilize close at room temperature. And HF steam is even more evil then liquid.

    Pros have most and highest respect when dealing with HF… extra gloves, extra lab coat, extra extractor hood, extra safety gogles and a set of extra first reaction aids for HF accidents (bascially calcium gluconate gels and solutions)

    Just be warned…

    Happy hacking

  12. This is the dumbest idea I have ever seen, for multiple reasons.

    First, you will never be able to make a useful MOSTFET without high purity chemicals, ultra-clean lab equipment, and a clean lab. There is a reason BJT’s were used for so long – MOSFET’s have a lot of problems with surface defects and oxide impurities that can only be solved by cleaning up your process.

    Second, these are not the kinds of chemicals anyone should be using in their garage. HF is extremely dangerous – far more dangerous than HCl. Previous posts have touched on this already.

    Third, it would be far, far, far cheaper and more useful to buy commercial MOSFET’s. Obviously, that is not the purpose of the exercise, but the point still stands.

    Now I could definitely see homebrew point contact BJT’s being an interesting educational project, and possibly even useful for high power applications… maybe even a pnp or npn diffused junction BJT… but homebrew MOSFETS? Not a good idea.

  13. @ cdjstyle:
    Its a transistor you can switch a current with it on / off ?! If you ask why he did make it by himself… I guess the answer would be: Because he can…

    @ smaddox
    sorry but you are a bit ranting. Its not dump its a nice proof-of-principle. Sure it has all this kind of problems you mentioned but hell… its working. BTW I really would like to see some input-output characteristics.

    1. I like the idea to break-down “high-tech” to kitchen-level. This proofs that nothing is that complicated that you can not do it by yourself. And its somehow nice that things work (even if very imperfect and rough) without a million (or in terms of semi-fabs billion) dollar fund. Rather important and one of the basics in terms of hacking isn’t ?!

    2. This shows how all this started. I mean look at the first transistors and at the first FETs… they were not much different to what he did.

    However, I agree (and posted already) that HF is not for ordinary usage… but I guess this is Hackaday and not Safteyaday… Every Hack contains potential danger. As long as people make this very very clear it is ok for me.

  14. All I have to say is that Jeri has built racing cars, run a computer company, helped create a Commodore 64 in an FPGA/ASIC, pinball machine and now this. I can’t wait to see what else she has up her sleeve.

  15. HF – It’s not a huge safety hazard in low concentration which you see in off the shelf products. You can use art store glass etchant (source of HF)or Whink which old ladies use to clean rust off tile and sinks.

    Why do it? – Surprised to hear that in this forum.

    BJT’s – Harder to make at home with furnaces that vary in temperature(emitter diffusion end point problems), no way to make good collector diffusion without epitaxy and boron diffusion rates.

  16. Holy fuming HF Batman, Jeri’s back and i’m still
    reverbing over that audio floppy hack.

    Good to see the video updates. Hope the published
    writeup mentioned in an older video appears.

    “Why do it?”, her Flaming Furbies might be one
    high power app for this.

  17. Hats of to this lady! I shall remember it if I survive a nuclear holocaust :) Joke aside, very interesting. I’m keeping this together with the French guy’s method of making vacuum tubes.

  18. * chuckles at silicone.

    CircuitGirl is cooler than all of you simultaneously wirelessly connected to the mother of all Arduinos and armed with enough blue LEDs to clog up the oil leak in the Gulf of Mexico.

    and @ smaddox : If you understand the point of the exercise then why do you sound just like that post menopausal english teacher I had in middle school? Get laid ffs.

    did i mention CG is hawt? (►_◄)

  19. This was awesome. The final product looks beautiful! NMOS turns me on, or is it an intelligent hot self-made EE? BTW yall need to stop bitching about the spelling and focus on the facts of the post. Last I checked none of you have a hacker blog thats as popular as HAD, and they seem to be doing fine following their own ways.

  20. Hey Jeri, I appreciate you thinking you have to prove you did what you said you did, but your creds are to the point where you don’t have to break it down for us to believe that you did it. Would you please make your next project the cure for cancer?

    To those who asked why? I am so ashamed for you. A transistor, the single most world changing idea to have happened since the wheel, screw, and lever. It changed radio, television, computing, telecommunications, medicine etc.. And to have built one in your own house, amazing. When you next feel the need to ask why, I suggest you stop and think of reasons for why not first. I could understand this on some other forum, but that question should never appear here, at least not for these kinds of demonstrations.

    Most people do not even know what a transistor does. If you had to ask why, I am thinking you might be one of them. If you are, go and download your porn, and leave the rest of us alone.

  21. interesting.
    wonder if Jeri has tried making a “negative resistance” device yet?

    this process could also be useful for making memristors, in this case you start with blank silicon, etch so it becomes porous then implant with silver or gallium ions and install a second porous silicon sheet on top.

  22. I work in a wafer fab here in KC, and I think this is awesome, although a lot more crude than what I’m used to seeing.

    All these comments about HF — We go through a pretty expansive chemical handling class before working with all of our chemicals. Honestly, the HF doesn’t bother me at all. Maybe I’ve become complacent. We wear another set of gloves and a special chemical apron when we handle it, so we are pretty well covered.

  23. the gel is agar-agar and is pronounced aygarr not ahguh
    I agree that things like that (silicone/silcon)diminish what is brilliant efforts on Hackaday.

    Jeri Ellisworth is a true scientist.

    1. No, no and no! They do not last 1000s of years!
      Large 1960s vintage ICs have predicted lifetimes as high as 10,000 years while current minimum geometry have lifetimes in the low tens of years.
      The processes of aging are also accelerated by stress energy.

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.