Ge and temperature question

[blackieNYC]

Something I have quite found the answer to in a search - when the temperature goes up, what can I expect to happen to the gain of my single NPN booster? Hot=more gain? less? Random? I think it's a 2n 404. I played on a very hot stage last night and got some surprises.

I made this booster for an end-of -chain solo boost.  Perhaps Ge is not a good choice. I could move this to the front on the chain where it would serve (well)  to push dirt pedals harder, but I would lose the consistent 8-10dB boost I want for soloing. (Clean sounds will boost, dirt sounds will just get dirtier).  Perhaps a  Si transistor boost would be better. I do want the boost to be fairly neutral. The Ge boost originally was clipping at its input, which I attenuated enough to a avoid clipping while still keeping a max output gain of about 12 dB

[Ben Lyman]

I breadboarded a simple LPB-1 Si boost and got an unwanted, albeit slight, clipping and tone change.
I put a trim pot in front of the input cap and got some very nice results, I highly recommend trying it.

[Gus]

Do some tests
If you have access to thermocouple and readout tape the bead to the transistor.  If not Infrared thermometers don't cost that much nowadays.
Connect a DMM set to DC volts at the collector and ground
Power the circuit with a regulated power supply to keep a constant supply voltage
Heat and cool the transistor and record the collector voltage vs temp
Or let the circuit stabilize to the room temp and measure the collector at different room temps

Try that with a silicon transistor
Try that with different biasing for Ge and Si

[R.G.]

Something I have quite found the answer to in a search - when the temperature goes up, what can I expect to happen to the gain of my single NPN booster? Hot=more gain? less? Random? I think it's a 2n 404. I played on a very hot stage last night and got some surprises.

Imagine if you had built your $10M mainframe computer out of those, or your communications satellite or your nuclear bomb trigger.   

All of those things were actually done - but it hurt. The industry wanted away from germanium as quickly as possible.

The reason is not gain, it's leakage. A bipolar transistor is two abutted junctions, one forward biased and one reverse biased. The collector-base junction is run reversed, and its leakage is amplified by the current gain of the transistor. While current gain does change with lots of things, leaking more current into the base isn't directly one of them. What C-B leakage does is shift the operating point, unless you have cleverly designed the rest of the circuit to be stable and self-correcting to some degree even with leakage. That's how those computer and stuff worked. The designers knew of germanium's foibles and worked hard to mitigate them.

A bias shift might sound like more or less "gain" in the uneducated guitarist sense of the word, where "gain" really means "distortion".

On as equal-apples basis as you can get, germanium leaks about 1000 times more than silicon, by the nature of the beast. So silicon circuits are about 1000 times as temperature stable, because leakage *is* dependent on temperature. Germanium junctions make great temperature sensors, because while their temperature coefficient in %/C is about the same as silicon's, the leakage you can use to detect this is about 1000x as big.

[Guitar Poppa]

Hello !

You should have joined the schematics of your Ge booster to get accurates answers or helps...

As wrote RG, Ge transistors have a bad reputation in everything about heath sensibility... Heath increasing does not cause a higher audio gain, but shifts the currents in the crystal, especially leakage. The bias being shifted, the transistor does not overdrive at the same level, and not the same way. So you get another tonality. Maybe not so bad, maybe not so good.

I don't think that using heathvariable resistors will regulate your circuit : it is very difficult to find the right value and right variation that will compensate the current variation inside the transistor...

But...
I use germanium circuits for about 10 years : overdrives, preamps, fuzz and little practice amps. With certain precautions, you can widely decrease the effects of thermal drift and get reliable circuits.

--- First : Some components added on the right place will be usefull :
Negative feedback resistors between the emitters of transistors and ground will make the transistor tend to autoregulate its collector-emittor current.
A germanium diode placed in the opposite direction between the base and ground will also help, ans its effect will be amplified by the presence of a resistance as said supra. If I had shematics, I could tell what should be the values. I just can give principles...

--- Two : If you have a trimpot to adjust the bias, don't use it at common temperature and as soon the circuit is on. You should adjust it in conditions close to the conditions you'll have in a real gig : high temperature, and several hours playing. Adjust the bias to give the best sound after a least 20 minutes working in real conditions of temperature.

--- Three : if you'll have to play in very different conditions, you should place the control so that you can ajust it from the outside. That's what Mike Piera did years ago in its Sunface...

Well... The two last points are pragmatics, and easy to realize. The first is more technical, but is very efficient. Visitors on this site should remember that antique stompoxes had very simplified circuits, because their designers and manufactures were often stingy... These are the limits of cloning : you'll clone the defects as well you'll clone the good ideas. In most case, you have to add some usefull components.

Post a shematic if you will more accurate help
Will be happy to give a hand

Guitar Poppa