Leaky transistors?

suspiciousminds

Apologies if this is a daft question but what exactly is a leaky transistor and whats the effect of using them in a pedal?
For example, if you had one in a fuzz pedal would it result in a drop in gain, sound loss or what exactly?

randomhandclaps

Virtually all germanium transistors leak current.  Leakage is important to know as most multimeters will give you a Hfe (or gain) reading of the apparent gain of that transistor.  However leakage causes a reduction in that gain so the true gain of a transistor is total Hfe minus leakage.  So for example if you had a reading of 120, but a leakage equivalent to 20, then the true Hfe (gain) of that transistor would be 100.  Too much leakage can also be indicative of a faulty transistor.  Some circuits are more tolerate to high leakage trannys than others.

Silicons do not suffer from leakage unless they are faulty so we do not need to account for this.

suspiciousminds

Ahh, so the overall result would be loss of gain?
Also how do you measure the gain using a multi meter then?

randomhandclaps

Yes, the major factor would be loss of gain.

With a couple of resistors (a 2.2M and a 2.4K) you can rig up a small circuit which using a multimeter will give you both the total gain and the leakage measurement allowing you to minus one from the other and find your real gain.  It's not as accurate as or as quick as purpose built semiconductor testers (like the Peak Atlas), but provided you are testing all your transistors using the same method it is perfectly reliable.

Some multimeters have a specific socket for this, but as I said remember these will give you fairly reliable measurements for silicons but do not account for any leakage so are no use alone for germaniums.

ecc83

A typical small signal Germanium transistor could gave a base emitter leakage current (at 20C) of 1microamp. It could also have an hfe of 200 so that would cause a collector "leakage" current of 200mu amps. Some Silicon transistor circuits don't have working collector current that high!

The leakage for Silicon devices is often less than 1/100th of a microamp and thus for all normal situations be ignored. In both cases leakage rises with temperature but since it is so small for Silicon these transistors can operate with junction temperatures up to 200C whereas "Geraniums" need cossetting and 85C is their limit.

As for testing for gain? Just tack them into a published circuit and check the DC conitions. They will be different for different samples and a bit of Ohm's Law will tell you what hfe is to a first approximation? But really you just want the "best" one?

I suppose peeps want "the magic" and IF you can tell, fine. Personally I would take a lot of convincing that Gerries are worth the trouble!

Dave.