Would this transistor work?

[NunoLava1998]

I am thinking of making a computer from scratch (literally) and i wish to create my own transistor (I have a lot of experience in computers, lots. Not too much in electricity, though) instead of buying transistors like this one:

So i want to know if my PNP transistor design works. Thank you for any information on this.

 

[cjdelphi]

You want to be dope silicon to create a transistor?

 

[duke37]

You have connected your transistor as a diode. You need three connections to amplify.
What is wrong with buying transistors, they are cheap enough.
Are you going to dig up copper ore and refine it?

 

[NunoLava1998]

You have connected your transistor as a diode. You need three connections to amplify.
What is wrong with buying transistors, they are cheap enough.
Are you going to dig up copper ore and refine it?

t-
no, i'm gonna use graphite and this is for things like http://www.instructables.com/id/How-to-Build-an-8-Bit-Computer/

 

[hevans1944]

So i want to know if my PNP transistor design works.

No, it does not work.

 

[NunoLava1998]

No, it does not work.

Okay. I'll maybe just use regular transistors (y' know, 0.09$ stuff)

 

[(*steve*)]

9c each is pretty expensive. In quantity you can easily get way below this.

The cheapest transistors I have cost me around 1/20c each -- but they're in an inconveniently small package.

 

[(*steve*)]

A very quick eBay search reveals you can get 100 2N3904 transistors for $1.48 with free shipping. I didn't look too hard, you can probably do even better.

 

[cjdelphi]

That's still way cheaper than buying a furnace not getting the precise temperature right then coming out with paper clips

Even the japanese struggled without high precision hardware, they bounced light off the temperature gauge which then showed up on the wall where the tinest needle turn resulted in seeable movements on the wall, problem solved! Until they had to dope it and use exact amounts

Point was, if a proffesor in japan struggled you might as well lol

 

[hevans1944]

A few years ago (circa 1996) I had the good fortune to be employed in a lab that had a low-energy, high-current, ion implanter. One of my tasks was to keep it operational, but It mostly stood idle since the people who actually used it to dope semiconductor wafers had long before then moved on to greener pastures. Not once did I ever consider using this machine to make my own transistors... or anything else for that matter... even though highly purified, grain-oriented, monocrystalline silicon wafers were readily available for literally pocket change.

I did toy with the idea of making micro-electromechanical systems (MEMS)... tuning forks, microphones, rate gyroscopes, sensors on itsy bitsy cantilevers, etc. Soon gave up that idea when I discovered I could purchase ready-made stuff right off the shelf at affordable prices. It just got better and better after that.

Being able to build integrated circuits is like having a license to print money... assuming you can come up with the initial billion dollars or so to build and staff the fabrication facility. Next best thing is being able to rent time in such places to become a fabless manufacturer of integrated circuits. Third best thing is being able to purchase integrated circuits for a few hundredths of a cent each in large quantities, provided you don't have a Not-Invented-Here (NIH) mentality. For those that do have a NIH mind-set, there still remains ASICs, and microprocessors with program protection fuses, and field-programmable logic arrays (FPGAs). But make your own transistors from scratch? Not in the 21st Century.