/diy/ - Do It Yourself

I have not heard a DIY attempt at this, as it most likely would have been huge news. The resources required would cost you a great deal. Fabs cost $5 billion for a reason.

That and without proper training you would probably kill yourself with all the extremely dangerous chemicals involved, if you could even get your hands on them.

>>459123
Sure. Got a few hundred thousand dollars lying around for start-up costs? Then there's hiring an engineer or two to oversee the fab operations.

Just buy ICs as needed. You can't "DIY" this.

The highest ends of 3d printers should be able to draw circuits that small. /noexperience

>>459163
yeah? idk a whole lot about 3d printing, can they print copper?

>>459163

>>459159
>>459143
I dunno, I think the larger end of microlithography might be doable DIY. like still on a visible scale, but just barely.

>>459166
I havent heard of anything that can print anything other than plastic or aluminum.

Fabrication

The process of creating integrated circuits is called Fabrication. It is a sequence of chemical and photographic steps in which the circuits are constructed on a semiconductor material known as "wafer". The steps are described below:

Lithography

In this step, a layer of photo-resisting liquid is applied on the surface of semiconductor or wafer. It is then backed and hardened.

Etching

In etching process, unwanted material is removed from the wafer. Then the pattern of the photo-resist is transferred to the wafer.

Deposition

In this step, films of different materials are applied on the wafer. It is done by either "Physical Vapor Deposition" or "Chemical Vapor Deposition".

Oxidation

In the oxidation process, the silicon layers on the top are converted to silicon dioxide by oxygen or water molecules.

Diffusion

Diffusion is carried out to anneal the lattice defects.

>that sounds totally doable.

>>459169
You MIGHT be able to get some 100+ um CD chips made...
IF you can get a device to make a uniform light source at the exact frequency
IF you can make a photo mask that costs less than several tens of thousands of dollars
IF you can get the chemicals, which are EXCEPTIONALLY dangerous and can only be bought by licensed people
IF you can construct a class 100-1000 clean room
IF you can get pure silicon wafers
IF you can construct an oven that can reach temps of 1000 C or more

There are way too many IFs in the equation.

>>459176
>something is too hard, so we should just give up.

>>459177
>too hard
I do this every day at work. I know first hand what it takes to make this stuff. I am stating that doing this at home is not a feasible endeavor.

>>459182
not who youre replying too, but can you explain why
>>459176
things are necessary?

enjoy
http://www.youtube.com/watch?v=eFzsyQOTXbM

>>459123

your dvd burner is a nanolithograph

>>459185
https://www.youtube.com/watch?v=F2KcZGwntgg

This should explain most of it. post any questions you have and i will try to answer.

>>459202
did you even read the OP?

bump for a semi-original thread

bump for interest.

>>459123
just cant be done by oneself OP, sorry man.

>>459143

That has been done. Some bitch made extremely crude transistors and a CMOS inverter at home. IIRC she didn't use photolithographic masks, though.
There might have been some dude as well, who did the same with somewhat different process.

I haven't heard of anyone making anything actually usable. That's not surprising.

>>459163

Clearly. They can't even come anywhere close.

>>459123
DIYable, but you won't be able to make very good chips:
http://ajp.aapt.org/resource/1/ajpias/v73/i10/p980_s1?isAuthorized=no

You can probably find it somewhere free on the net, I'm sure of this.

Also, pretty much all lithography is done in first world countries by robots, because humans are too dirty for making chips. Electronics are assembled in China though, because labor is cheap there.


>>459168
>>459163
It's true, pic related. Met a German guy doing exactly this with a home made machine.

>>459166
Can't do copper though.
You can with another process, but that's $$$$ and there's no way you'll be able to diy that.

>>459366
That'd be Jeri Ellsworth:
http://hackaday.com/2010/05/13/transistor-fabrication-so-simple-a-child-can-do-it/

If you have the money to buy an Electron Microscope you can make circuits of that size on silicon. It's just slow as shit and of fairly limited use. And expensive and involves having to learn how to fix and convert old electron microscopes, and source or make silicon wafers and crap, and prepare everything so the electron microscope can work it. You also want a TEM, and they come up far less than SEMs on lab garage sales. That would get you in the 10 to 100 nm range.

bump

>>459388
this is the answer to your question. start from there.

>>459202
>your dvd burner is a nanolithograph
this is true, and the Lightscribe ones can already draw pictures.
so then all you'd need to do is make the "wafer" on a round shape that could be spun in the DVD burner.... :|

It would perhaps not be possible to do *anything* for making a microprocessor, but it is much more than nothing....

>>459741
its got the engraving part down at least.

>>459123
anyone think that something along the lines of millilithography scale IC would be possible?

>>459765
it sounds doable but the chip itself would end up huge.

>>459765

Yes, but that's not going to help you much. You still need to do the diffusions, etchings and whatnot.

>>459772
>>459774
bump

>>459123
cant be done

>>460319
it can be done, and I will be the one who does it.

>>460320
post pics if u make progress.

>>460320

Semi-seriously, if you want to try it once or want something with decent performance, you could consider just designing the chip and buying the actual manufacture. The price depends on the chip size and the process, but the cheapest options cost just a couple of thousand dollars for a dozen or so chips.

>>460331
thats not /diy/ though.

>>460359
it is partially.

You could make a one or two transistor ic on a slice of germanium.

>>460359
>pic
>not keeping the piglets and raising them for sale, breeding, and slaughter

This shows the typical mindset of people that go hunting.

Farming > Hunting and gathering

>>460384
that would be illegal
>hippies in charge of knowing what the fuck they're talking about

That hunter is just a dumbass. You don't kill the sows, you kill the boars. You kill the sows and piglets and you hurt the population and reduce the number of animals left to hunt. If all he could find is a sow hes either a bad hunter or a lazy hunter. Boars give more heat anyway. Tards like this ruin it for the rest of us that hunt correctly. Just like all this wing nuts that wont enact sensible gun control laws because all of the money the NRA passes out and mah freedoms. Soon they'll get voted out of office and then shit will swing the other way and some really onerous laws will get enacted instead of the logical stuff like universal background checks. Sometimes being an American just sucks.

>>460407
>doesnt know what the fuck hes talking about
>doesnt know that feral pigs are a huge fucking problem and the only way to keep them breeding like rats in a barrel of oats is to kill the females and piglets.

sage for you being a thread derailing idiot.

>>460359
Design it yourself.

>>460413
fair enough

>>460412
this. fucking feral pigs can absolutely ravage native ecosystems.

>>460389
Raising pigs is illegal? Feral pigs are just pigs that escaped into the wild and started breeding. It isn't illegal to raise your own on your own farm.

>>460426
Maybe the guy lives in Israel and doesn't have the appropriate flooring.

>>460426
selling pigs without documentation is often illegal. breeding some species of pigs is illegal.

>>460407
>we need sensible library controls
>we need to keep people from learning chemistry just in case they decide to make a bomb one day

>>460407
Dunno what the fuck you're responding to, but reducing the population is the fucking point. Feral hogs are an invasive goddamned species that breeds extremely quickly.

>>460417
what is operatorchan?

>>460451
free SoL!

>>460451
/k/-chan basically.

>>460436
Selling meat without documentation is illegal in the first place. Selling some species of anything is also illegal.
Moron

>>460479
ah, thanks

>>460439

Reductio ad absurdum. Doesn't apply. I'm not who you were responding to.

Sage because we're off topic.

>>460587
actually thats a pretty good parallel to the entire gun control argument. not who youre replying to btw.

>>460587
reduction to absurdity my asshole.
>>460439
is right on.

>>460672
>Keep crazy people from having guns. Cool.
>Keep criminals from having guns. Cool.
>Let sane people have all the guns they want. MORE guns than they can have now. I want full-autos back.
>I'm okay with background checks, collecting illegally owned guns, and tests to get this.

>>460683
wut

>>460587
the problem with "reasonable" gun laws, is that reasonable is always defined by the people who hate guns.

>>459166
>>459173
I know a quite a bit about 3d printing. And it's leagues away from what OP is talking about doing. Modern electronics are made on the nano scale, the best 3d printers at the smallest are on the micron scale.

There are people trying to develop a system for making electronics in a "3d printer" fashion, but its all in university research phase. I'm talking extreme prototypes working with completely untested designs. They are only doing it because of the massive cost of making chips the current way. You have to setup entire production lines and make millions of them to justify costs. A machine capable of custom printing circuits would be an amazing cost reduction for making short run and replacement chips. But a lot of the technology and techniques required to do so simply don't exist yet.

It's certainly coming. But when it does arrive just one of these machines will have a tens of million dollar price tag. The tolerances, scale, and materials involved are at the limit of what entire production lines are currently capable of.

>>460426
You do realize there is a massive feral pig problem in the south? They destroy crops, damage property, and harass livestock. Billions of dollars of damages. Many states have enacted year round limitless hunting on them for that reason.

They are also a huge pain in the ass to hunt. They are more likely to attack and injure a hunter than a bear. And your whole point is fucking retarded to begin with. Who would try to re-domesticate feral pigs when we still have domesticated pigs? What do you gain by suddenly capturing such an animal?

>>460727
right on.

>>460730
here here

>>459169
>on a visible scale
That would be millions of times slower than anything we have today, what would even be the point?

>>461172
the point, good sir, is DIY.

>>460730
>What do you gain by suddenly capturing such an animal?
You gain such an animal, for free.

>>460326
>before lawmakers outlaw them

>>461182
You are wasting your breath trying to explain to a subsistence hunter.

Let's settle in place and build some civilization. I used non-lethal traps to catch these chickens and pigs, so now we have a reliable food supply for tough times.

>>460730
>What do you gain by suddenly capturing such an animal?
based on my experience in minecraft a feral pig is just as delicious and useful for breeding as a domesticated pig. what's the difference to the state so long as they're out of the wild?

>>461194
11/10.

the point is that he doesn't want the pigs. He's got no shortage of pigs. Intact he has too many pigs. Why would he want to breed more? He's hunting them for population control and sport, not for subsistence

>>461204
>minecraft
hehe

the funny thing is, many of these posts say it isn't feasible or possible. and a few mention 3D printers and equipment that costs a lot of money. However, what I don't see is a lot of support for the question. yes, the equipment is expensive and 3D printed boards would melt when u tried to solder them, however, using 3D printing and parts you can buy online or find at radio shack, it is feasible to make your own equipment for such a project. O coarse it would still be an expensive set-up, but it is still a fraction of what you would pay for corporate made equipment. and before any one puts down my comment, let me remind you that Stephen Wozniak, creator of the Apple computer and inventor of the floppy drive, created the first personal computer in his garage when "experts" said it couldn't be done.

>>461220
I'm certain old G Dubs had such terrible grammar. Seriously, did some moron really type that damn apostrophe TWICE? I don't want THAT person to own a gun. Seriously.

>>461302
>grammar nazis
>2014

>>461222

You did not just actually suggest trying to build fab equipment with parts from Radio Shack, did you...

>>461322
its possible

>>461330

Under the strictest literal definition of possible, yes, I suppose it is...

>>461332
>they said the same thing about going into space

>>461172

Not everything needs insane speed. 1960-era transistors are fine for many kinds of analog chips. I, personally, would be quite happy with that /diy/ level.

>>461222
>3D printed boards would melt

This thread is (was) about integrated circuits, not PCBs.
The small, simple computers became easier and easier to build over the time and are now something even a newbie can build.
The IC manufacture, on the other hand, is still rather difficult even at the lowest level and the state of the art chips require multi-billion dollar factories.
This is not going to change in the near future, at least not for the silicon-based chips. I don't know about the (printable) organic semiconductors, but they're currently very bad, even when compared to the 1950-era semiconductors.

>This thread is (was) about integrated circuits, not PCBs. The small, simple computers became easier and easier to build over the time and are now something even a newbie can build. The IC manufacture, on the other hand, is still rather difficult even at the lowest level and the state of the art chips require multi-billion dollar factories. This is not going to change in the near future, at least not for the silicon-based chips. I don't know about the (printable) organic semiconductors, but they're currently very bad, even when compared to the 1950-era semiconductors.

The reason I brought up 3D printing circuit boards is because, if you actually took the time to read through the comments you would see, some one suggested printing out the circuit board. Now it maybe true that making micro sized integrated boards would be very difficult without the precision of the computer driven equipment used in multimillion dollar labs, but that does not mean it cannot be replicated, it simply means it will be a little bigger. However, anyone with the knowledge to do so can get to the point of building micro sized circuit boards by the same proses in which we used in the first place, by building machines that make smaller and smaller machines. Building such a machine can be done with parts from radio shack, and pre-built project boards such as Raspberry Pi or even a desktop for that matter. all you would need is some knowledge of integrated circuits and robotics. anything you need to build the robot can now be 3D printed and as for tracking for accuracy, that can be achieved with a standard webcam in combination with sensors and pre-programed movements.

>>461443
I don't think you understand how small ICs are. The best I would hope for from a 3d printer would be the size of the smallest surface mount component.

seriously. No chance of making an ic, ancient or modern, it's too specialized

>I don't think you understand how small ICs are. The best I would hope for from a 3d printer would be the size of the smallest surface mount component.
>seriously. No chance of making an ic, ancient or modern, it's too specialized

Integrated circuit
(IC)

A circuit that incorporates multiple transistors and other semiconductors to a single circuit, sometimes called a chip.
A complete electronic circuit consisting of numerous transistors and other solid state devices, and formed on a single silicon semiconductor chip.

Integrated circuit board
Electronic circuit made from transistors, resistors, etc., all placed into a package referred to as a chip, since all circuits are on one base of semiconductor material.

https://www.youtube.com/watch?feature=player_embedded&v=rvpUp7-dwh8#at=273

everything you need to know about it can be found online. once you grasp the concept it isn't as hard as it seems. and transistors, resistors, timers, so on and so forth can either be found at radio shack or online.

>>461222
>inventor of the floppy drive
A namefag, what a surprise.

>>459123
90% of modern ICs is IP.
You could design that (provided you have the proper software and libraries), but you could have hard time with the production.

>>461464
>since all circuits are on one base of semiconductor material
>post video of a phenolic PCB
Maybe you want to have another go?

>>461480
>since all circuits are on one base of semiconductor material
>post video of a phenolic PCB
>Maybe you want to have another go?

What do you mean by phenolic PCB?

https://www.youtube.com/watch?v=tWnfnt2rNO0&NR=1&feature=fvwp

>>461495
A phenolic board is a phenolic board. For a thread about IC manufacturing you have a spectacularly poor idea of what a semiconductor even fucking is.

>>461495
>since all circuits are on one base of semiconductor material
>post video of a glass-epoxy PCB
Third time's the charm?

>>461501
>461501
>since all circuits are on one base of semiconductor material
>post video of a glass-epoxy PCB
>Third time's the charm?

Phenolic paper is a material often used to make printed circuit board substrates (the flat board to which the components and traces are attached). It is a very tough board made of wood fibre and phenolic polymers. It is most commonly brown in colour, and is a fibre reinforced plastic. These PCB materials are known as FR-1 and FR-2.

What exactly are you looking for?? there are many types and the boards they used were phenolic PCB.

http://www.alibaba.com/showroom/paper-phenolic-pcb.html

>>461521
>>since all circuits are on one base of semiconductor material

>>461525
That doesn't answer the question. What exactly are you looking for??

>>461579
you seem to think an IC is the same thing as a PCB.

PCB on a 3d printer. Sure. Why not.

You could just DIY the lithography or etching on a laser or inkjet printer. You don't even need a 3d printer.

DIY a 3d printed IC, even at 1970s resolution? No. Just no. there is at least 2 magnitudes of accuracy away from even the shittiest IC on the very best protopltype pribter

>>461579
A printed circuit board is not a semiconductor, you fucking numpty. A printed circuit board is not an integrated circuit.

>>461443
Ok you aren't getting it. We arent talking about building boards like RaspPi or desktop motherboards. We are talking about integrated circuits. The chips on those boards. The processors.

It's not just a matter of scale, The efficiency of modern circuits is due to their size and the shear number of connections they have. We are talking nanoscale manufacture. Something 3d printers, even the very best of comercial machines, would be completely incapable of doing.

It would be like using a plasma torch to solder a bread board. You are completely misunderstanding what we are talking about and the scale of the task at hand.

>>461583
>>461586
>>461610

integrated circuit
noun Electronics.
a circuit of transistors, resistors, and capacitors constructed on a single semiconductor wafer or chip, in which the components are interconnected to perform a given function. Abbreviation: IC

In late man's terms, the circuitry is part of the board, which is exactly what PCB is. I'm not talking about a microcircuit.

Does anyone actually read all of the comments?? I never said anything about 3D printing chips other than to say it wouldn't work. I said you could 3D print parts for a robot to construct small complex chips and I mentioned the RaspPi or desktop as a means to control the robot that creates smaller chips. and yes size is a factor, which is why I suggested a multi step process that would allow you to scale down with each step, which is the same process used in the first place.

>>461617
> I never said anything about 3D printing chips other than to say it wouldn't work.

That's what everyone else was talking about. You keep listing off definitions like we don't know what these things are, yet you fail to grasp the fucking subject under discussion.

Look at your definition:
constructed on >a single component or chip

Guess what mother fucker? That ain't a PCB board. It's a chip, which is what everyone else has been talking about and you've been smugly ignoring.

You don't understand the difference between a board and chip. Leave.

>>461337
>The IC manufacture, on the other hand, is still rather difficult even at the lowest level and the state of the art chips require multi-billion dollar factories.
You can do non-commercial, small scale fabrication with an electron microscope. There's been on and off use of this method since the 90s to fab experimental chips in universities, since it's relatively inexpensive to retrofit an electron microscope the university already has lying around. That'll cost a relatively small $150 to 200k all together, and you could reduce it quite a bit with a little hunting, scavenging and luck. That kind of set up can produce chips on the scale of any current commercial production if set up right. Maybe look into that and consider smooth talking a hacker space or something? The EM would be usable for looking at ICs at least, and that alone may be something they'd want.

Bear in mind that this method is great if you want to try something novel out. If you know the circuit will work ahead of time it's boring as anything.

>>461631

>The IC manufacture, on the other hand, is still rather difficult even at the lowest level and the state of the art chips require multi-billion dollar factories.
You can do non-commercial, small scale fabrication with an electron microscope. There's been on and off use of this method since the 90s to fab experimental chips in universities, since it's relatively inexpensive to retrofit an electron microscope the university already has lying around. That'll cost a relatively small $150 to 200k all together, and you could reduce it quite a bit with a little hunting, scavenging and luck. That kind of set up can produce chips on the scale of any current commercial production if set up right. Maybe look into that and consider smooth talking a hacker space or something? The EM would be usable for looking at ICs at least, and that alone may be something they'd want.

Bear in mind that this method is great if you want to try something novel out. If you know the circuit will work ahead of time it's boring as anything.

That is one of the most helpful comments so far. the rest of these idiots are simply dismissing the possibility that it can be done.

>>461629

and as for you... It's morons like you that make me hate places like this. There is always an ass hole setting around bashing on anyone they can.

>>461631

Are you talking about e-beam lithography?
You still need resists, diffusions and so on, with the associated equipment and materials.

Since it hasn't been mentioned yet:
Micro imprint lithography and organic semiconductors
are much more DIY-able solutions to IC manufacturing. Granted, MIL will only help you with the replication, you still have to work out mask creation, etching, doping, etc...
Organic semiconductors on the other hand could be prototyped on ink-jet printers.

>>461648
More or less. If by resists you mean putting the pattern on the oxide layer, that can directly inscribed using a TEM or a modded SEM without a mask, which is a big part of the cost saving and how they can go so small but also the main reason why they're useless for commercial application. The stage before that where you'd bake an oxide layer or cure a polymer on top is still not easy, but I'm sure doable. Diffusions are a big part of the retrofitting cost, but there's plenty of literature out there on how it can be done.

>>461643
I'm not bashing on "anyone" I'm bashing on a namefag idiot talking out of their ass. This "helpful comment" has been repeated three times already in this thread. And you were too dense and contrarian to figure it out.

>>461643
>dismissing that it can be done
this is still a dismissal. read the comment you need an electron microscope and 200k worth of mods.

That puts the process far and away from what could be considered DIY construction of integrated circuits.

>>461704
It's more like 100K worth of mods, 200K all in (microscope + mods). You can get EMs and materials for cheaper than that, but it's still going to be around the cost of a house or a really nice car. Then again, some hackerspaces do carry equipment of a similar cost to this so maybe it's something you could set up with likeminded people. The other thing is, once you've successfully built your IC, then what? It's really something that suits continual development, ideas that are modified and rebuilt over and over, otherwise you'll have paid hundreds of thousands for possibly the satisfaction of building one chip. Not sure that's worth it. Chip design is very involved and has large groups of people behind it, you'd have to at least live and breath it to make it even slightly worth it.

>>461704
true but where as alot of the comments are about raising pigs and selling meat and mic. off topic subjects it is one of the better ones.

>>461699

And you can call me a name fag all you want. you'r still bashing. If it wasn't on me it would be on some one else. Besides I like my name it gives me a signature so I'm not just a number, besides that, it helps me keep track of my posts.

>>461719
why are you on an anonymous image board then?

>>461722

Because I have enough accounts as is and too much spam besides some of the most interesting conversations pop up on anonymous msg boards.

>>461722
Because I need to masturbate my ego constantly

>>461743

LMAO petty attempt of an insult. What are you 10??

>>461784
of course I am

>>461784
I'm Drayke I need things explained to me. Waa Waa Waa.

>>461719
Surprise, no-one gives a flying fuck about who you are.

>>461682
Do you still need the cleanroom facilities and stuff? I know there are HAM people doing GHz stuff that have old wire-bonding machines in their garage.

>>461840
You would for transport of substrate to and from ovens, EM and all of that, or something like that. The EM itself acts as a kind of mini cleanroom in a way once you have the stuff in it, but it's getting the stuff inside it that's a problem in a state appropriate for the EM to etch on it. So unless you can work out another way, yes, clean room all the way. But that's something you'd probably like if using an EM to look at normal circuits. It's not necessary, but less headache inducing.

>>462096

So, what is required, all in all?
- a cleanroom facility
- a modified electron microscope
- possibly a spin coater (depending on the process)
- a diffusion oven
- etching baths
- metalization equipment, whatever it is
- possibly cvd equipment, if more than one metal layer is needed (???)
- wafer saw
- wire bonder
- possibly all kinds of test equipment (???)

Then of course all the gases, silicon, chip packages, clothing and the other consumables.
What else is needed? Or what is not needed?

>>461661

What's the current situation with the organic semiconductors? Last time I checked they were ridiculously shitty, transistors with ft in the audio range and all.

>>462105
Last I heard, Isorg is getting ready for mass production, but that's "just" organic photodiodes. I guess that the performance of organic transistors will always suck, compared to their silicon brothers. Their advantages are more in fields where you need small runs, large surface areas and low initial costs.