Substitute germanium for silicon NPN

[TimWaldvogel]

I know when you sub a modern silicon FOR germanium not TOO much needs to be changed.
But when going the other direction for an old school feel, what changes do I need to make to drop in a NPN germanium in place of a NPN 2N3904/2N5088 in something like an lpb-1 circuit?

[TimWaldvogel]

Outside of measuring the hfe.
From what I've read germaniums can practically self bias cause they have such high leakage where as modern silicons are much more efficient. I was wondering if I can leave a bias resistor or if I should remove it.

[R.G.]

It varies too much with the individual germanium transistor to give a guide.

You have to vary the bias for two reasons, both the leakage you've mentioned, and the fact that the base-emitter drop is 1/3 to 1/2 of that for silicon. This makes almost no difference in properly designed circuits (i.e. those designed for device independence, "modern" practice since the 1970s) and huge differences if the circuit is not designed to be Vbe independent.

The second reason is leakage, and that is highly sensitive to both the individual device, how well it was packaged and sealed, and whether it has been overheated or otherwise mistreated over it's life so far. If that sounds discouraging, it is. And it's one of the fundamentals for why germanium was abandoned by the industry as soon as they could get away from it.

One observation: there is no "old school feel" to germanium separate from what it does in a circuit. It will sound different, but that's a result of how it lives/fits into the circuit and signal, not because it's germanium. You can't rub germanium on a circuit to get magic tone.

[TimWaldvogel]

Well this is actually for a multistage mic preamp circuit. The flaws in some germanium transistors do something to the frequency response and I just wanted to try a circuit out without having to spend the extra $ for NOS germanium transistors to burn them out from lack of knowledge