Transistor leakage

[Gumby212]

What exactly are the affects of high leakage in a germanium transistor in a fuzz pedal? Does that just mean that its hfe will vary more bc its constantly leaking volatage? Is that why some take forever to settle when im testing them? I realize temperature plays a large role, but are the more leaky ones harder to test? Seems like their readings vary more. Is there a difference in sound for the leakier ones? Ive read leaky transistors can be used to a fuzz's benefit but how so? I find the leakage and the hfe using a multimeter along with a 2.2m resistor and a 2.472 resistor on my breadboard. All my readings for leakage seem to be between .13 and 2.91, for a wide range of germanium transistors. Ive read that over 300 is too much leakage. Is that .300? Are the ones that give readings like 2.91 way too high? Or am i reading these wrong. Any help would be greatly appreciated, as im new to playing with germanium. Thanks!

[thermionix]

2.472k, right?  The voltage readings for the leakage test are divided by 2472 to get leakage current in µA.

[antonis]

In brief:

Leakage current is considered the current flowing through a revere biased diode (like Collector-Base one)

If that current was clearly specified AND constant at any temperature, it should be no problem at all..
(it could easily be taken in mind when calculating Collector (or Emitter) current..)

Due to its wide spread and its doubling for every 10^oC raise, its value is difficult to be precisely calculated hence the quiescent point of transistor amplifier is prone to be set NOT at the point of load line we wish to bias..
Additionally, even if we would manage to bias our transistor at some specific point, it should commute in accordance with temperature variations..

The greater the leakage current (in relation with Collector working current) the greater the "deviation" from ideal (for our purpose) working condition..

The above result in non-predictable distortion..
(some people call it "Covalent bond forbidden energy mojo"..)