Homebrew NMOS Transistor Step by Step - So Easy Even Jeri Can Do It

[4u250ut]

I know this sound lame..But I think you are sexy because of your brain. :)

[christianjb2002]

I hope no one gets decalcified by doing this. That's a horrible way to die..

[douro20]

As I understand it, there are the most common type of power MOSFET, and the first to be widely manufactured. N-channel MOSFETs are workhorses in switching power supplies, motor drives, power inverters, or any electronic device where fast switching of high amounts of current is required. The only problem with MOSFETs is the fragile gate oxide, which can be destroyed by static electricity.

[EKOSMUNiDAD]

.

...OOO... O ...OOO...

.

[markd514]

@capricrow You are absolutely right, because I thought about it, and researche dit a little bit, and I could definitely do it at home, but why would I want to? It is kind of a waste of time. I can just buy one that works...lol

[markd514]

I am sorry, but there is NO way that you have self taught yourself to know all of this about electronics. This is on a physics level that requires many years of engineering. Sure, you can read about it, then "teach" people about it online. But there is no way that you pulled this out of your head and just wrote all this down.

[frac]

I envy you the feeling you must have had with your first success. I couldn't help but imagine the pioneers working their way past insurmountable problems to achieve the same thing without the benefit of prior knowledge. Amazing.

Would putting the wafer off-center make the spin coating a bit better? I got the impression that such a small wafer in the center like that wouldn't receive even and/or high centripetal (centrifugal? ...) forces. Hmm. The "real" machines center the wafer though...

[brianm734]

Sweet.. So in this design the connection of the substrate electrode to one of the outer P-doped regions determines which connection is the source, correct? Have you tried buying pre-surface doped P-type wafer and etching down through the dopant to create the gate channel, then grow the oxide gate insulator? Have you experimented with other P-type dopants like Boron?

Thanks for the excellent videos.

[davidsaunders0]

Thank you. I was beginning to think that this was a lost art. In the 70s and 80s this kind of thing was fairly common place amongst the smaller semiconductor makers and some hobbyist, those most of them were doing BJT devices, still making ICs with out the aid of a clean room. We need to learn this stuff again.

[Technoguy3]

How do you know all this? :O

[Vatsek]

Did you also make a PMOS transistor?

[jeriellsworth]

@Vatsek I have.  It takes longer.

[henriquedante]

I love you

[driscoll42]

Also what model furnace did you use to grow your oxide? I tried to tell from the picture on flickr but couldn't see it.

[driscoll42]

To increase the rate at which the oxide grows, do you think it would be beneficial to create some sort of mechanism to force steam through the furnace? I'm envisioning a coffee machine repurposed to just boil water and have the steam flow into the furnace, perhaps a low rpm fan. Also if the wafer could be mounted such that it was vertical somehow, the steam would hit it much more evenly I think.

[supergenius1994]

I read in popular science (or maybe it was popular mechanics) that if you put strips of paper in conductive ink and then wrap that into a big circle, and put electricity through it, you can make an industrial power air filter that can get rid of dust. It could be useful for making an IC

[jeriellsworth]

@ITGuru0111 Dust is not a huge problem when the transistors are as big your thumbnail.

[supergenius1994]

where do you get the starting wafer?

[jeriellsworth]

@supergenius1994 Ebay and an other companies that offer prime wafers. Very easy to find.

[supergenius1994]

@jeriellsworth I'm sorry, but I can't find anything at all.

[jeriellsworth]

@supergenius1994 They will show up on ebay. Keep an eye open for silicon wafer. There are companies that sell directly too.

[prence1725]

you are a female teaching eletronic, i feel so shy

[tstreib]

I don't understand the process but I was always fascinated with Science. Please, does anyone know where I can go for beginners. web sites or alternatives?

[danhinz1]

HF is indeed nasty stuff, used to manufacture & package it for the semiconductor industry. It may cause cancer(doesn't everything?) bigger concern is it leaches Ca out of your bones. (Look at F & Ca on periodic table) Like already said, wear proper protective equipment & have a place to rinse off w/lots of water (a shower) lastly you could probably cover any splashed skin w/milk of magnesia or other OTC antacid after water rinse. Like the Boy scout motto, Be Prepared. Go for it!

[Dom6161]

First of all, great video! Very inspiring!

Could you give a little more information on the furnace you were using? It seems to be the most expensive bit in this whole exercise and the only real lab equipment needed. Could you recommend a model to copy the project on a tight budget?

[djaa2807]

Very informative video, i'm doing this in Analog Electronics module at the University :)

[moveaxebx]

Woman which knows how to "bake" an NMOS transistor!

Respect, respect, respect, respect...boy some one is so lucky to have you....

[DrMario2007baka]

jeriellsworth, You're in the league - Quartz can even allow Extreme UV in - thus with some attenuation. The only trickier part remains is how to instruct the VLIW graphic processor (Darkstar GPU from TI especially made for the DLP DMDs) with just a video card (You can try ancient video card with DVI port - with some know-hows, as you have to hack both GPUs to do what you want the DMD to do, such as laying out UV photo-lithography imaging) I have done away with DMD, their GPU is hard to use.

[DrMario2007baka]

Interesting. One problem with common projector optics, they're not even rated for UV functions - some has special coating to eat UV starting at 402nm and shorter, and if so shorter, glass starts to absorb it. 250nm and shorter is the best way, but longer wavelength is fine - just don't shake the whole thing (if anyone uses laser - a shake is enough to ruin the whole transistors - whole fab is full of shock absorbers in there too...) Still, it's a great project!

[jeriellsworth]

@DrMario2007baka Surprisingly two of the DLP TV's I picked up had quartz windows on the DLP unit. (IR compatible)

[ssapp72]

You make homebrewed transistors? IMHO, you are officially the coolest woman alive. :D

[KarriKoivusalo]

Most impressive! Are there readily available photoresists compatible with this process? I'd imagine a HD video projector with proper optics would make a good tool in making small scale integrated chips :)

[jeriellsworth]

@KarriKoivusalo I've been playing with photoresist, but haven't had good results yet. I think part of my problem is not having a good development tank.

[KarriKoivusalo]

@jeriellsworth Interesting, you wouldn't have any sneak preview of materials and processes involved?

[PatheticComputing]

Very impressive! I thought about doing this at one point and even started to shop around for some silicon wafers on ebay. But I had a lot more research to do before even attempting this. Now that I know it can be done, I'm definitely going to try this one day!

[jeriellsworth]

@PatheticComputing It's surprisingly easy, although it took me three years of experimenting to make my first device. Watch out for exposed PN junctions. Keep oxide over them or they conduct.

[PatheticComputing]

@jeriellsworth Hello again. I was wondering if you have tried any phosphorus dopant other than the Emulsitone stuff with any success. Also, can you tell us how much that phosphorofilm costs? Did you buy that kit off their website, cause it's really expensive...

[jeriellsworth]

@PatheticComputing I've tried boric acide roach killer which worked a little. I tried red phosphorous from matches, but it didn't seem to work, but I didn't try more than once.

[DJGrenola]

Amazing. Mind blown.

[CampKohler]

Except as an excercise in the nearly impossible,I can't imagine why anyone would do this any more than one would try to make one's own aluminum. ("Now lower the electrode from a 1000-amp DC supply into your bathtub filled with crumbled bauxite until the voltage across the tub is exactly 5.000 volts. Have a ladle handy to dip out the liquid AL. WARNING; DO NOT LET THE SOAP OR ANYTHING WET FALL INTO THE TUB OR YOU WILL BE STEAMED TO DEATH AND THE BATHROOM DOOR WILL END UP IN THE NEXT COUNTY!!")

[ControlCardPin]

Can this scale up to actual digital integrated circuits? I wonder how long it'll be until someone grows a 6502 in their garage. Might take up the whole garage, though, heh heh

[shiftgood1]

THIS WOMAN IS A GOD ALMIGHTY GIFT to us. Thanks woman. you make it for even the little guy. No one is without knowledge of what you have dome for womankind and youself. of course, be sure to keep aware of others taking advantage of you (and others) by making THE BOX....for your protection from "Freudian Trnasferrences" I speak in liue of more than ten years of par time bosses, workers, slow stickered vhiecles, and SHIFFTERS BOX. Paradigm Shift.

[ashsama2]

I work with MEMS manufacturing all day in a 6"/8" wafer fab, maybe I should try making some transistors :)

We got shitloads of scrap wafers I can try on =)

[Neilidity]

Very impressive! Can I ask, How important is having the Si wafer surface with a given crystal orientation (you mention 100 in your experiment)? Are other orientations OK, or is 100 the best? Also, is it important to have a single crystal orientation, or could a conchoidal surface also be used I wonder - have you ever tried with fractured Si pieces instead of nice Si wafers?

[TzzX78]

It brings Intel memories back...

[TzzX78]

is very dangerous what she is doing...and I am refering to HF is an acid that Intel takes exteme care so empoyees will not get expose to it....Causes cancer of the bone marrow..and once is inside you ...you are done. rip.

[jeriellsworth]

@TzzX78 It's not as bad as you think. All of the etches I show can be done with art store etchant(also HF), which is safe when handled as directed in the instructions. People etch glass all the time, without problems.

[TzzX78]

That was the best. Do you work for a semiconductor company?? if not youshould work for Intel.

[88calls]

A+ keep up the good work.

[apopheniacMCMLXXXIX]

Oh my god. This is unquestionably the most amazing thing I've ever seen.

[zzzSilveRFoXzzz]

Stupid? i think that the answer is yes... why are you wasting a lot of paper sheets?

[ericfontainejazz]

NIce Homebrewed Transistors!!!

[happygamestvfun1]

whats the point i can get them for fuken n 1 cent

[kirkamr]

This is very impressive. I worked for a semiconductor company back in the day and always assumed this type of fabrication was beyond mere mortals.

[000Ru000]

I'm amazed that you can home-brew stuff like this; great vid :)

[BARONSCHWARZWALD]

when making this are there any toxic fumes or gases produced?

[jeriellsworth]

@BARONSCHWARZWALD No. The dopant is in glass form.

[TzzX78]

@BARONSCHWARZWALD HF is bad for your skin..and bone marrow

[alphabeets]

Mind blowing!

[kencbk]

Very Impressive. I thought this can never happen, a home brewed transistors. Nice video detailed like it, i would like to try this. thanks for your post

[Thetruthishere11]

Fantastic!

[allmyenemies]

Hi, which kind of liquid you use for doping silicon? You have used boron atoms? If i have a p-type silicon i have to put a n doping for the n-wells, right!? In your bottle there is a label "Boron" but this kind of doping will put a p-well, you have to make a p or n well!?

[jeriellsworth]

@allmyenemies I use spin on dopants from Emulsitone they have borofim and phosphorosilicate. FYI. it's easier to work with phosphorosilicate films.

[kinglonewolf104]

Wow, that's quite alot of work required to make a NMOS transistor. The end result is impressive though :)

.

&eB

[jigglesnap]

Real programmers make their own computers... out of SAND

[mophead6912]

This is great!!! Thanks for sharing, and yeah.... you *are* hardcore

[edtronic]

SEXY PROJECT

[Verm1n]

Excellent! Have you measured the Rds(on)?

[bytesilver]

you fucking amaze me sometimes, your the hardest core chick i know and im in absolute love with you... although im also in love with my amazing beautiful and kind girlfriend. i just needed to express that.. i been checking your vids for quite some time now... but really... this... is beyond anything i could have ever imagined, and THANK YOU SO MUCH for making this possible to the people.

[sylware]

This is great! I want more!

[dissimilate]

I am in awe. Keep hacking Jeri.

[YoLninYo]

this is brilliant! have you ever tried making solar cells at home in your furnace? there is no opensource method on the web that I know of ... is it too hard or just too easy so no-one tries it?

[izzyjb]

@YoLninYo There was someone showing home made solar cells at Maker Faire 2009.

[jeriellsworth]

@izzyjb That was me. I also showed transistors, but you might have missed them.

[jeriellsworth]

@YoLninYo Solar cells are very easy to make with this process. Problem is that they take too much energy, because of the furnace and small batches. They'd never return energy to pay back 6hrs at 1000c.

[YoLninYo]

@jeriellsworth thanks for your reply. I was thinking more in terms of empower users/small industries in developing nations. It is true that the energy costs are high for small batches but the biggest problem facing people is having the "know how". If there is a way to get crude cells using crude tools and kilns then it might spark a whole cottage industry...

many blacksmith shops in villages in africa/india have such furnaces and tinkerers/hackers can build on that.... IF they had a "howto". :)

[robsonnbi]

What's the cost of a silicon wafer?

[jeriellsworth]

@robsonnbi $10 -$15 from vendors or cheaper on ebay, but you have to know what to look for.

[solidacid1337]

i wish i had a 1000 degree kiln, this is AWESOME!

[Gta2CubanPete]

i want to see the led turning on and off

cool stuff

[jeriellsworth]

@Gta2CubanPete Check my other videos. I have a demo of it driving a blue LED.

[fdesmet2nerds]

When are you going to show us how to grow and pull a silicon ingot out of a molten blob of melted sand? :-)

[jeriellsworth]

@fdesmet2nerds Ha! That would be fun. The process of purifying silicon involves a lot of thermal cycles, acid and explosive silane gas. :)

[pumkinvine]

i not sure why I watch these videos, the only part that I knew what you were talking about was "CPU fan" lol

[yugrewohs]

Wow, in a post-apocalyptic world, I want you on the team that helps rebuild society.

[Turrican0815]

@yugrewohs thats what i thought too :)

[HD41117]

Jeri, you mention "the water" beading up on the exposed silicon layer. Is this water used to rinse the etchant from the surface? Sorry if it's a dumb question, just wanting to know where it's coming from and how it's being used.

[jeriellsworth]

@HD41117 This is the rinse water. I've use distilled water and made functional devices.

[HD41117]

@jeriellsworth Thanks Jeri. Your video with the three stage rinse tank was a very detailed answer to this question, too. Great stuff - I'm excited to try it, too.

[Afrotechmods]

XD CPU fan as a centrifuge... you are awesome.

[bonecrime]

It would be awesome if you'd build some kind of working circuit out of your homemade transistors! :D

[jeriellsworth]

@bonecrime I tried a ring oscillator as my first IC, but it had a short. One of these days I will do more.

[bonecrime]

@jeriellsworth I'll be looking forward to seeing it!

[sinistermoon]

Awesome!

[flippinrawks]

excellent presentation, thanks for sharing

[badfishification]

cause everyone is crazy like you and makes chips at home :P