Using Ge transistors in "reverse" (swapping collector and emitter)

[brett]

Hi
Here's what I understand so far:  In some Ge transistors, the emitter works quite well as a collector.  The hFE falls (by 20% to 80%, depending on the type), which can be handy. e.g. if you have a hFE=120 and a hFE=160 and want to build a fuzzface, use the hFE=160 the "wrong" way around for Q1.

My question is about leakage.  Does anyone know whether leakage is consistently lower or higher when the transistor is reversed?
thanks for any information

[R.G.]

The difference between forward current gain (beta) and reverse current gain (alpha) is a function of the asymmetry of the two junctions: collector-base and base-emitter. Early transistors, especially alloy-junction germanium transistors were made very symmetrical - they had to be, the techniques for making different doping densities and profiles were not yet invented, or were not used at least. It was only after people found out that non-symmetrical junctions were a good thing that the two junctions were differentiated. Eventually, they became quite different, accounting for today's silicon transistors having radically smaller alpha than beta (that is, much lower gain when reverse-connected than when forward-connected.)

In germaniums where the emitter and collector are nearly interchangeable, I suspect that is because the two junctions were formed in a way that makes them very similar. I would bet - but do not know for a fact - that the leakages will be very similar as long as one junction is not more exposed to surface contamination in the packaging.

[gmr1]

interesting stuff. I just got a batch of very nice NOS 2N167A's. Mil-spec NPN germaniums. They work very well "reversed" in a fuzz face circuit. The fuzz is a little bass heavy, and not as saturated, but the clean up is incredible and halfway up the volume pot on the guitar yields a very nice overdrive that's very "tube amp" sounding. I'd love if they had a little more gain (they're typically 50-75 hfe), as they'd likely work even better reversed.

Anyone have any other transistor model numbers come to mind that also work well in reverse?

[mac]

Anyone have any other transistor model numbers come to mind that also work well in reverse?

Matsushita 2SB172/5/6 family. They have sub classes that have b=+<200> and leakage=<200uA>.

mac

[gmr1]

Anyone have any other transistor model numbers come to mind that also work well in reverse?

Matsush*ta 2SB172/5/6 family. They have sub classes that have b=+<200> and leakage=<200uA>.

mac

I had no idea. I have tons of those. Have to give them a try. Is there any potential for damage of the transistor by swapping the C & E?

[Scruffie]

I built a friend a fuzz face a while ago with a few extra pots and a booster in it using some BC558B Silicons and he used it for about 3 months then said it broke, so he gave it back I had a look inside and realised that i'd somehow put the 2 PNP transistors in backwards, swapped them back around and it started working again... which makes no sense to me after it worked fine for all that time, but the transistors didn't mind being swapped around after all that time atleast.