hfe doubler for ge q's

[pinkjimiphoton]

found this surfing the net, looked cool... appears to double leakage current, too

[antonis]

appears to double leakage current, too

and increase noise floor, too..

IMHO, a Darlington (or better Sziklai) pair should do the same job while saving one BJT..
(but less predictable current gain multiplication..)

[Rob Strand]

I like Mr Hoshuyama's stuff.   The extra diode that results in saturation will have an effect on the sound, but it should be small.

The success depends on the degree of matching of two "mirror" transistors.

The main scheme to to make the circuit less dependent on the transistor matching is to add emitter resistors
to both of the added transistors.   This improves the match at reasonably high currents.   You need to tune
the emitter resistor values so that "reasonably high currents" are the currents in the circuit!

If you use different emitter resistors values it's possible to get more than 2*hFE.

You can also do things like put a resistor on one of the transistors to get more gain.

Keep in mind the idea of this circuit is to replace a single transistor.   If the effective gain of replacement is too variable then the final circuit might won't work.   VBE matching helps prevent this problem.

The Sziklai circuit is probably the worst in terms of producing a consistent gain.   It produces a high gain at high-ish currents.   It's OK to use it in circuits which don't care about the transistor gain, like a buffer, but for an existing amplification stage it might not work without adjustments.

Here's a whole heap of possible gain boosting circuits.   There plot shows the effective transistor gain, "overall hFE", of the combined circuit.

It's fairly obvious matching helps.    What might not be so obvious is just how much the overall hFE varies with VBE mismatch.   That's what the plot shows.  It also shows you need to tweak the RE resistor to suit the operating current of the circuit.

I flipped the polarity of the circuit but the idea still holds.

Where there's three circuits the first is the matched case and the other two are when the VBE mismatch is approx +/- 40mV.

[Axldeziak]

I found the original page it was posted at:
https://understandcircuit.seesaa.net/article/200604article_6.html

And here is some rather humorous (and bad) google translating....

"I thought about how to double the Hfe of a Germanium transistor in a circuit.
（ Published on SNS in June 2005. I haven't tried it myself, but it shouldn't move, so I'll publish it on Blog. )
Germatra is often with dozens of smaller Hfe. In Faz Face, there are more than 120 big Hfe stories that are good, but such stones are hard to get. If you allow it to be adjusted for each stone, it will be full of semi-fixation and will become a circuit that cannot be reproduced.
Aron's bulletin board and others have announced some fuzz face-like circuits that have circuitically forced the Hfe of the Silicone Tra, and vice versa.
Double the collector current of Germatra with a talent mirror. It will be a equivalent circuit with a diode in the collector. The germainess may decrease a little （ or violently ）, but for reference, there may be such an evil path."

edit:
I should have looked further down in the comments. It seems it was already talked about here in 2007!
http://www.diystompboxes.com/smfforum/index.php?topic=55717.0