Quite amazing stuff this youngster is doing. From literally scrap he has built a semiconductor fab.
That's one way to get around the chip shortage!
I've been subscribed to him for years now.
Guy started fabricating semiconductors while still in highschool.
Whenever I hear the bold (unprovable) claim “Worlds first”, my mind kinda shuts off. Since he’s no way of proving that his was the “worlds first”, it makes him appear a little big headed.
Never say that. You have no way to prove it, and we have no way to counter it. Lose/lose for your image, in other words. Just make the devices and drop the “worlds first” - every man and his dog claims this, for everything from sliced cabbage to satellite phones. I’m sure you’re excellent at what you’ve made, but I’d drop that claim.
We made diodes in one of my college classes. Photo etched each layer and had two furnaces to slide the wafer trays into (one for each doping type). Amazingly, I was able to find them after 45 years.
Here showing different stages of the process.
This is the one I liked the most.
Nanosized photonic structures on butterfly wings. Imaged with an electron microscope.
..and he was never heard from again.
..and he was never heard from again.
Sam's part of the twitter hivemind now. No need to panic when creators stop producing content for a while.
Everyone knows the big challenge is in educating the masses to grow scalable chip farms. Boutique garage chips will only ever be educational, and and $10B Rube Goldberg 3nm factories will only ever be exploitative.
I really don't understand why you are crawling on a road trodden 50 years ago by professionals. Because you want? Because you can? Because no one else was doing it alone in the garage?
Some people lack passion in their life so much that they choose to suck passion away from others as well.
Good thing Sam is able to handle these comments by providing replies reflecting the ignorance of the original comment.
Some of my relatives say this to me all the time "...but why do you need all this equipment?? You should spend your money on
something better".
Then i tell them that i need it and they ask me to explain why. As if i owe them an explanation now?
Explain what? Why i choose to support my interests based on experience and knowledge i gained over many years?
These kind of people assume that what you do is dumb because it doesn not align with their interests, which are obviously superior in every conceivable way.
impressive!
I can think of some classic chip designs to try
with todays high demand for bespoke integrated circuits of the classic retro designs
bespoke music synthesis devices like top octave generators & bucket-brigade devices
& DIP chip red LED seven-segment displays of the 1970s
Yarl Qman is a TSMC shareholder getting nervous
Whenever I hear the bold (unprovable) claim “Worlds first”, my mind kinda shuts off. Since he’s no way of proving that his was the “worlds first”, it makes him appear a little big headed.
Well, the first shown on YouTube, I don't think you can dispute
Other than him, I'm aware of Jeri Ellsworth having made some DIY MOSFET or a few. It's not a common hobby.
The guy's website:
http://sam.zeloof.xyz/I recall him saying that a lof of his work has been made easier by having scored cheap wafers on eBay with the gate oxide and a polysicilon layer already deposited, so he doesn't need to deal with producing those thing and some of the chemicals involved.
it makes him appear a little big headed
Seems to me he has every right to be if he wanted - hands up anyone here who has even vaguely approached the kind of stuff he's achieved. Bonus points for being younger than 30.
But in fact he doesn't seem to be big-headed at all. OTOH, grumbling jealousy... wonder where I've seen that before...
I'd suggest that where this idea could shine is in production of legacy semiconductors which are no longer possible because mass-production tech has changed so much in process that the result really isn't the same device. 2N3055H transistors made using the hometaxial process spring to mind; the later epitaxial designs simply are a different transistor.
If Dalibor can recreate/rediscover the technology that made large-format nixies possible, I see no reason this guy couldn't figure out how to make a 2N3055 that behaves the same as the original in those old much-sought-after amplifier designs.
Before you say it... no, that is not
my rabbit-hole; it's just a "thing" that I'm aware of.
mnem
Just an observation; As always, take anything I say with a grain of salt big enough to pickle a dwagon.
I'd suggest that where this idea could shine is in production of legacy semiconductors which are no longer possible because mass-production tech has changed so much in process that the result really isn't the same device.
Just an observation; As always, take anything I say with a grain of salt big enough to pickle a dwagon.
I am sure there would be a market for the LM3909 LED flasher, among other retro ICs...
I'd suggest that where this idea could shine is in production of legacy semiconductors which are no longer possible because mass-production tech has changed so much in process that the result really isn't the same device.
Just an observation; As always, take anything I say with a grain of salt big enough to pickle a dwagon.
I am sure there would be a market for the LM3909 LED flasher, among other retro ICs...
That would be cool but IIRC that would require having more layers. Besides that the LM3909 is likely made in a bipolar technology. Something else missing is input protection; a usefull chip has some resilience against ESD. IOW: it will take a few more steps to achieve making actual chips like you can buy from a big manufacturer.
Nevertheless I'm very impressed by what Sam Zeloof is achieving in his home lab. It is great to see young people interested in technology that everyone takes for granted.
Bipolar requires running your own epitaxial growth (after buried diffusions), precise diffusion depth control and multiple diffusion on one another. I know nothing about those processes in practice, but it surely sounds like more work if doable at all.
CMOS has some nested diffusions too and one of them needs to go under gates, so you can't start with off-the-shelf wafer with pre-applied oxide and poly.
NMOS is easy mode
Yup, have been following for a while, a Wired article is great success.
Would also be interesting to see affordable fabbed amorphous or poly transistors on glass; the performance might be poor, but when all you need is a few MHz equivalent bandwidth or clock rate, then you can miniaturize the hell out of it. Like I could make this whole module,
in a single chip; the board is only needed because no one makes a chip that does the whole function by itself (it's a not-uncommon function, but chips only ever have it along with a bunch of other inseparable housekeeping functions).
Or epitaxy on sapphire, if you can get wafers of it at good enough purity. (Artificial sapphire isn't too expensive by itself; in fact you interact with it on a regular basis, it's the scratch-resistant windows on checkout machines.) Doing this in Si is still going to require vacuum apparatus, with CVD or sputtering; doing it with compounds like ZnS however could potentially be prepared in atmosphere.
Tim
Boutique garage chips will only ever be educational
I dated a woman who was making me back diodes in the kitchen, on GaAs wafers. I don't know where you can get a back diode these days other than inside some very expensive detector.
Her comment on Jerry Ellisworth video was "She picked an easy target. Any monkey can do this".
That's beyond the point I was trying to make. The point I was trying to make is this: Every generalization is wrong (including this one). Boutique chip shops are doing all right (I'm familiar with more than a couple), and many things can be done in the chicken coop if you know how. When one of the boutique shops I'm familiar with went belly up the principals bought about the half of the equipment, with cash. Pumps, reactors, etc. I wonder why.