Swapping Silicon for Germanium - design considerations

[Rectangular]

I recently came into possession of a good number of germanium devices, and am interested in converting some silicon-based circuits, into using these germaniums, mostly as an experiment; I'm hoping for some interesting nonlinearities to crop up. While I'm aware that in really generic applications, swapping Ge for Si won't require any changes to get the circuit to fire up, I was curious as to what the design differences are in Ge circuits, compared to Si ones (ie: what kinds of things were engineers in the 60s doing, to compensate for the high leakage, and other imperfections of Ge ?).  some examples of what I mean: larger or smaller pull-up or pull-down resistors, larger/smaller decoupling caps, filter caps, feedback resistors, etc.. 

any advice, or rules of thumb, would be much appreciated

[raulgrell]

I can't really help, as I have no experience with Ge. But keep in mind:

Part of the great sound of Ge is actually DUE TO the imperfections of Ge devices. Which imperfections these are... I have no clue whatsoever.

[Rectangular]

hey raul

just to clarify a bit further : I'm not trying to null the imperfections of a Ge device. instead,  I'm interested in what measures can be taken to adapt a Si design,  to retrofit it into functioning with Ge semiconductors.  I'm not talking just about fuzz faces and simple fuzz/boost/amp circuits... but with filters and oscillators, plus other subtractive synthesis devices

[raulgrell]

Ah, I see! I just assumed you were going for the distortion of Ge since it's what is most reputable about it.

In terms of everything else, I couldn't think of a reason to use Ge instead of Si, but you might find something interesting. What I can comment on is the temperature instability of Ge. This guy found that a simple diode can counter the effects of this instability: http://www.geocities.com/sunsetstrip/studio/2987/britface.html . Just scroll down to the bottom of the page. He explains exactly what it does, maybe the theory could be applied to other applications of Ge transistors.

[R.G.]

Given that you are only looking for how to adapt ge transistors into si circuits, there are only two things you need to know:

1. The Vbe of germanium is about 0.15-0.3V instead of silicon's 0.45 to 0.7.
2. The leakage of germanium MUST be contended with. One reason the industry moved to silicon is that its leakage is intrinsically over 1000 times lower than germanium and can be ignored in almost all cases. Germanium transistors, for instance, cannot be turned off by opening the base. The leakage self biases the transistor slightly on; some early effects used this. Germanium transistors cannot even be turned off by connecting a resistor between base and emitter. Leakage again. You must reverse bias the base to turn it off when switching. The temperature coefficient of the leakage means you cannot ignore temperature,either.

[mac]

As I'm addicted to Ge, I found a way to adapt Ge in Si circuits, with no tweaking of bias resistor (or very little):

http://www.diystompboxes.com/smfforum/index.php?topic=56703.0
http://www.diystompboxes.com/smfforum/index.php?topic=63010.0

In these two examples I did not change bias resistor, except fine tuning the JBT as explained. I guess it could be used in many circuits.

To me it worth the try, even though, as RG said above, leakage and temperature are a BIG problem.

mac

[Rectangular]

Hey RG, great to hear from you, very insightful as always !

Mac, this is exactly the kind of stuff I'm looking for.  much obliged, both of you.  From the looks of it, I'm going to be buying a lot more trimmer pots if I want to get some of my circuits running on Ge.

also, that diode trick may also come in handy, Ill see if I can put it to good use.