silicon vs germanium

"Splodrod" wrote in news:fbb7n6$k3s$1$ snipped-for-privacy@news.demon.co.uk:

I'll leave the finer points to people who know more, much more, but basically germanium has a softer onset of nonlinearity, so has a smoother kind of fuzzy aspect to the sound. Silicon will be cleaner till it distorts, then it will distort more sharply. I think some people like germanium for being a little more like the effects of vaccuum tubes.

Also, if people are changing from germanium to silicon, they might not be after a change in sound, they might just be ignoring the differences and substituting newer transistors in older designs.

He did say the reverse, putting germanium where silicon originally lie. I wasn't sure if that was a typo.

If it is, the obvious point is that germanium transistors have basically been superseded by silicon, and virtually nothing has been manufactured with germanium transistors in decades. Finding germanium transistors at this point is likely about finding old stock.

If you can't easily find or buy germanium, then obviously replacing bad transistors doesn't leave much choice but to use silicon.

Michael

snipped-for-privacy@FreeNet.Carleton.CA (Michael Black) wrote in news:fbbnoo$h0u$ snipped-for-privacy@theodyn.ncf.ca:

Fair, it might be that way round, I just read it one way as it's ambiguos and I didn't notice.

It might then be that germanium transistors are sought for their sound quality to put in new circuits. One way to do that might be to use op-amps with germanium diodes in the feedback loop or other part of the circuit. I have no idea how viable that is, but germanium diodes are still sold, if not actually still made.

Many thanks for that. I think the "onset of nonlinearity" is the answer. Yes it was to replace the silicon with germanium. Thought I read somewhere whilst trying to find an answer that some germaniums are still made. The common ones but didn't look at the web site date. Cheers me dears BB

You need to be careful as just voltage drop across a Si junction is about

So if not properly compensated... you could end up with distortion or even destroying the device

Probably very little, although I wonder if they might have a slightly different output characteristic that clips more softly.

Germanium transistors are also likely to have lower current gain than modern silicon types and that might affect how they perform is some circuits.

Graham

Yes, the circuit will need to be adjusted to supply the different bias voltage. Just replacing a silicon part with germanium is not likely to work very well.

Also, the germanium part is likely to have lower current gain and that may also require circuit redesign.

Graham

Please don't top-post in these groups. If it's not important enough to appear above your text, trim it out.

Those parts are collector's items.

I guess if one can believe that gold plated wires can change the characteristics of an electrical carrying conductor enough that it can be detected at an audio frequency, then, I guess that person can tell the difference between silicon and germanium. For my self I have changed many circuits with germanium transistors to silicon just to gain the reliability and temperature stability of silicon. I would have never guessed that any waveforms would have changed if they had I would have been very surprised. Germanium transistors have many of the characteristics as silicon of the same size except they are very leaky and continue to increase in leakage with time and temperature. Both silicon and germanium have about a -2mv/c Vbe temperature coefficient it just starts at 0.6v for silicon and 0.15v for germanium. Ray

You would have to have some kind of audio-religious belief to do that since it has no basis whatever in science/physics/materials properties. I would totally disregard anyone who made any such specious claim.

That would be quite different. For a host of possible related reasons.

Graham

"Ray King" wrote in news:gIMCi.78682$ snipped-for-privacy@bignews6.bellsouth.net:

All that's assuming that people want the linear operationg region. You're ignoring that the OP was interested in guitar effects. Germanium junctions have a less sharp diode step change so when that's used to deliberately clip the signal there is a big difference in the sound. It may be that the extra leakage raises and changes the noise floor to contribute to the fuzzy character that people are after when they specify germanium transistors. Usually I think it was just that the onset of clipping was considered less offensive than with silicon, and for some musicians it's important to stay with an original technology rather than try to emulate it.

Did you bother measuring ?

Opinion is cheap. And guesses are inded only guesses. Provide hard evidence and I'd believe you.

Graham

Eeyore wrote in news: snipped-for-privacy@hotmail.com:

To add to this, some very distinct changes in sound may not be noticeable when looking at a wave, especially those that subtly change the timing of a long rising or falling slope. Failing to see changes in a waveform is a very poor basis for judging the sound. That's why FFT and DCT were invented.

Hi,

This is done in fuzz pedals, that generate distortion. When replacing them, you need to adjust several resistor values too!

A silicon transistor is generally a "better" transistor, but it is the "bad" part of the germanium transistor that is being used. It saturates more, the gain depends on the currents more. And it is slower.

The transistors are driven into saturation. The silicon transistor goes up to the power supply, then clips and give sharp edges. this gives a very sharp, mean sound.

But the germanium transistor saturates and can't even reach the power supply voltage, it shows characteristics that look more like a distorting tube, rounding off the peaks. But there is more to it: a germanium transitor is much slower, has higher capacitances (i think that that may also have to do with the older manufacturing processes). This causes the edges of the clipping to get really round. And at the same time, very high harmonics are not generated anymore.

So what does it do: It generates distortion, but only a couple of harmonics (more even than uneven harmonics), making the sound more pleasant. It gives more "warmth". And still "power" as it works in the mid-frequency region. It gives a sense of a "full" sound.

And now I have a great tip for you, and all you germanium distortion enthousiasts out there: you can add a very small capacitor to the base-collector of the distorting transistor to influence the warmth of the sound! Notice that the capacitance must only be a couple of pf as it is being multiplied by the transistor due to the Miller effect.

You can use this capacitor too to use another germanium transistor (one that is cheaper than the ones that are always used) and adjust the sound so they sound the same.

You can also influence the sound by adjusting the bias, so the transistor will either generate even or more uneven harmonics depending on wether distortion is on both up and down voltage, or only one.

If you have any questions about how to calculate resistor values, feel free to ask me. Email me at (and remove both the "byespam"'s) to snipped-for-privacy@byespamhoeben.com

Pieter

Pieter,

You have written a very insightful piece on the basis of the "tube sound" distortion versus solid state and also versus silicon and germanium to regain that tube sound.

The signal characteristics you describe and categorize as a "Good warm" versus "metallic and tinnie". The relationship to favor even and limit odd, while limiting total harmonic offers insight into what process you would need to create a tube sound with semi-conductors.

Best regards,

Marc