[sci.electronics] Homemade transistors, etc.

heskett@titan.tsd.arlut.utexas.edu (Donald Heskett) (07/28/90)

A long, long time ago, I would guess it was the late 60's, the
Scientific American "Amateur Scientist" section was all about making
your own transistors at home.  They were FETs, created by oxidizing
organometallic compounds on a hotplate.  The connections to the
semiconductor were with made by hand with small pieces of osmium.
Most of this is based on very stale rememberances; take it with a
large grain of salt!

I have some personal experience with crystal radios.  Originally they
used a point contact with a piece of galena, which is naturally
occuring lead sulfide. I tried it using a chunk of galena from a rock
collection; I wrapped the rock with bare copper wire and then made the
point contact with a breadboard mounted safety pin.  It seemed to work
about as well as a commercial germanium diode.  Up until the late '50s
or early '60s you could buy a ready made galena detector from
electronic suppliers like Allied and Lafayette.

 

mmm@cup.portal.com (Mark Robert Thorson) (07/28/90)

The connections were made with indium wire, not osmium.  The transistors
themselves were made by vaporizing different stuff onto glass microscope
slides heated on a hotplate.  These were thin film transistors, like what
is today used for making active-matrix LCD displays.

berryh@udel.edu (John Berryhill) (07/28/90)

Clean rooms?  Bah.

I've made BJT's with beta>700 in the filthiest lab you can imagine.  I'm
talking about a lab that my office mate occasionally takes his dog into
with him for company.  We use a planar process that could be done at home
except for the evaporative metallization.  FET's need a little bit more
careful handling, but the primary demon isn't dust as much as Sodium
contamination. As someone has pointed out, alloyed junction transistors
ought to be fairly simple to make at home, but you've still got to
figure out how to get the contacts on.

--
							      John Berryhill
					   143 King William, Newark DE 19711

dleblanc@joplin.mpr.ca (David LeBlanc) (08/01/90)

In article <25947@nigel.udel.EDU> berryh@udel.edu (John Berryhill) writes:
>  . . . . . . . . . .  We use a planar process that could be done at home
>except for the evaporative metallization.  FET's need a little bit more
>careful handling, but the primary demon isn't dust as much as Sodium
>contamination. As someone has pointed out, alloyed junction transistors
>ought to be fairly simple to make at home, but you've still got to
>figure out how to get the contacts on.
>
>--
>							      John Berryhill
>					   143 King William, Newark DE 19711


I was just wondering how you would get the high temperatures required
for solid state diffusion let alone oxidation in a home.  I suppose you
might get away with spin-on-glass cured in your oven for the oxides but
the diffusions have me stumped.  Yes, I know you can send away for
implants relatively cheap but thats cheating.  Hmm, maybe a solar
furnace could do the job in the right part of continent.

Re. sodium contamination problems - these would cause flatband shifts
but should not prevent the transistor from sort of working.  I recall being
taught that the biggest impediments to MOS FET development were the high 
interface state densities due to lousy oxidation processes, lack of post 
oxidation anneals and wrong crystal orientations.

Kitchen sink CMOS - the cottage industry of the 90's

Dave LeBlanc
dleblanc@joplin.mpr.ca

berryh@udel.edu (John Berryhill) (08/01/90)

You don't have to go above 950C in order to make decent devices.  You
can buy small (about a foot long and 6 inches high) quartz-tube furnaces
that you can run with a variac for a couple of hundred dollars (and they'd
have a zillion uses for the home lab).


--
							      John Berryhill
					   143 King William, Newark DE 19711