Why are germaniums so varied and tempermental?

[skiraly017]

Why are germanium transistors so varied in gains (within the same type) and tempermental to work with (biasing, sensitive to heat, etc.)? I know this is probably a sweeping generalization but it seems sometimes they are more trouble than they are worth.

[jlullo]

from what i understand, germanium is a very unstable element....

this may have been a rhetorical question, though

[jonathan perez]

they werent hugged enough as kids

[skiraly017]

they werent hugged enough as kids

Now that's funny.

[Rick]

I think it is more what this type of transistor does poorly than what it does best that gives you and all what were looking for.
Some germs are useless, and some are WoW. Experiment with several types.   ...Rick

[R.G.]

1. Germanium has a low band gap energy. It's easy to get it to semiconduct. It's also easy for temperature to push charge carriers over the barrier, so it's kind of sloppy and leaky at normal human temperatures.
2. Germanium was mankind's first semiconductor, at least partially because it's easy as in 1. So we did most of our early learning on germanium. The first germanium devices were baked on sheets like cookies in laboratory ovens. Then we found out that we needed to keep them clean. Ooops. I guess we oughta wash the trays off first before the next batch, guys.

Contrast that to today's hyper-refined single crystal silicon wafers cut from single-crystal silicon boules 300mm in diameter and meters long, and clean rooms where there is one dust particle in the ENTIRE CLEAN ROOM.

We did our learning on germanium, then abandoned it for silicon, which is tougher, more temperature stable, more predictable, and about 17 more mores.

What we have left today is the leftovers from our learning experience, kind of the worn-out semiconductor training wheels. We've moved on to cars, trains, airplanes, rockets, and space stations. But in this neck of the woods (guitar effects) we LIKE the way germanium works, even if it is an anachronism.

So there you have it. Germanium WAS loved and hugged as a kid. But it was dumped on a mean street in a bad part of town when a cuter, brighter kid came by. It never recovered.

[mac]

The air-conditioning system used to get that "clean" levels are impressive. A manufacture room is several times more clean than a surgical room.

To see the dependence on temperature T,

ic=is(T).[exp(e.Vbe(T)/(k.T) - 1]

ic=collector current, is=saturation current, e=electron charge, T=temp, Vbe=base-emiter voltage, k=boltzmann constant.

As is and Vbe depend on T and the material used, you have different behavior between Si and Ge.

http://hyperphysics.phy-astr.gsu.edu/hbase/solids/trans2.html
http://ece-www.colorado.edu/~bart/book/book/chapter4/ch4_2.htm

Some quantum statistical mechanics concepts are needed, like fermi levels, etc.

mac

[Steben]

However, the germanium kid was retrieved and salvaged again, once mankind discovered GeSi composite semiconductors. Suddenly the old kid was bright enough to serve again. 

Ah... Humans!.....

[ambulancevoice]

they werent hugged enough as kids

haha

[zjokka]

Supposing that Ge sounds different fro Si, isn't it just because it is leaky and doesn't perform according to modern standards, that it sounds the way it sounds? I tried to formulate this statement in a recent thread, but it might be kicking in open doors...

zj

[squidsquad]

Wonder if ya could...with today's knowledge/technology...study what makes the perfect/beloved qualities....& then somehow re-create...perhaps with new/different semiconducting elements...& be made stable?

[mac]

Wonder if ya could...with today's knowledge/technology...study what makes the perfect/beloved qualities....& then somehow re-create...perhaps with new/different semiconducting elements...& be made stable?

Sounds like "with today solid state technology try to emulate a vacuum tube".
I wonder if some guitar player at Motorola, Fairchild, ST, etc tried to re-dising a fet, if possible, to behave like a tube, or attempted to create something new more tube-ish.

However, the germanium kid was retrieved and salvaged again, once mankind discovered GeSi composite semiconductors. Suddenly the old kid was bright enough to serve again. 

I always wanted to put those SiGe transistor in a Fuzz Face, but I could not get them.


mac

[R.G.]

Absolutely!

All you need is
(1) access to a modern fab for the project; this is a little like saying you want access to the Ft. Knox gold reserves storage vaults.
(2) either enough money to do the experimentation or already knowing what you need to do.

Actually, the latest-made germanium devices were remarkably consistent and useful. I think you could save yourself a lot of money and just have one of the few remaining makers of germanium devices make you a batch. Last time I looked, Germanium Power Devices would run me 2000 NKT275's for about US$9.00 each, prepaid.

[analogmike]

I had GPD make me up several types of transistors, they were very consistent and consistently sounded terrible

Dunlop has been using these since 1993 on and off, and some of their new fuzzfaces don't sound too bad as they have been tweeking the circuit quite a bit. Still not nearly as good as even the generic germaniums you can get from small bear though.

have fun!

[zjokka]

I had GPD make me up several types of transistors, they were very consistent and consistently sounded terrible

Dunlop has been using these since 1993 on and off, and some of their new fuzzfaces don't sound too bad as they have been tweeking the circuit quite a bit. Still not nearly as good as even the generic germaniums you can get from small bear though.

have fun!

How did the leakage of these newly-manufactured devices compare to that of old germaniums? Were they very very low leakage?

sorry but I'm very curious,
thanks
zj

[mac]

I had GPD make me up several types of transistors, they were very consistent and consistently sounded terrible

I read somewhere in the forum that some "new" Ge transistors are not truly Ge. Also that some "new" Ge diodes like 1N60 are just schottky diodes. True or false? BTW, I compared old 1N60 with new ones, that look like 1N4148, and they sound a lot different. The new ones sound like 1N5817. Even comparing with other Ge diodes like OA90, 1N34 and AA117, the feel of these new 1N60s is more like a Si diode.

Actually, the latest-made germanium devices were remarkably consistent and useful. I think you could save yourself a lot of money and just have one of the few remaining makers of germanium devices make you a batch. Last time I looked, Germanium Power Devices would run me 2000 NKT275's for about US$9.00 each, prepaid.

I recently bought a bag of 100 2SB176 Matsushitas transistors in a small electronic shop 10 blocks away from home at us$100. The owner, an old woman, has a lot of japanese transistors in their original packages so they look like as if they were manufactured yesterday. I have enough Ge transistors for at least 10 lives more.

South America was invaded by japanese Ge, so if you contact with tech schools, military, etc, you may find some well-known Ge like 2SA49/52/53, 2SB54/56 (most likely Toshiba), and 2SB172/75/76 family and 2SD352 (most likely Matsuhita)

mac