Auditioning Ge Transistors For Noise?

[Ginsly]

I have very limited experience with Ge, but have breadboarded some circuits lately just to get a feel for them. I can see why people go through the trouble!

Other than rolling transistors in whatever circuit I’m working on at the moment, I wonder if there’s a good “auditioning platform” to quickly pop Ge in and get a general idea of noise floor before that stage. I realize transistor noise may present itself differently in various circuits, but maybe there’s a specific setup that would help sort for noise relative to the other Ge components being tested.

Looking into this, some have suggested a simple Rangemaster circuit, as it uses one (Ge) transistor.

Any thoughts on this? Is it simply too situational to even do a preliminary test like I’m describing?

 

[PedalBuilder]

I have very limited experience with Ge, but have breadboarded some circuits lately just to get a feel for them. I can see why people go through the trouble!

Other than rolling transistors in whatever circuit I’m working on at the moment, I wonder if there’s a good “auditioning platform” to quickly pop Ge in and get a general idea of noise floor before that stage. I realize transistor noise may present itself differently in various circuits, but maybe there’s a specific setup that would help sort for noise relative to the other Ge components being tested.

Looking into this, some have suggested a simple Rangemaster circuit, as it uses one (Ge) transistor.

Any thoughts on this? Is it simply too situational to even do a preliminary test like I’m describing?

Breadboard the circuit before you build it. That way you can audition different transistors, see how they sound in circuit, and adjust any components that need some tweaking.

 

[Ginsly]

Breadboard the circuit before you build it. That way you can audition different transistors, see how they sound in circuit, and adjust any components that need some tweaking.

Yep, certainly- wondering if there’s a preliminary testing setup before that stage that could help weed out particularly noisy devices in a more general sense. I have the feeling that could be tricky - for example, higher gain/higher leak Ge may have more noise when popped into, say a rangemaster “test circuit”, but would work wonderfully in a Buzzaround/TBIII like you’ve mentioned.

 

[darwin999]

You might check out this thread:

Thread 'Any use for noisy Ge transistors?'

Nov 11, 2023

I've got a small handful of RCA germanium pnp transistors with decent gain for fuzz faces. DCA 55 reports no leakage too, but they're very noisy in every circuit I've tried them in. I tried them in some broadcast builds and fuzz face circuits on the breadboard and they were unusable. I haven't tried using them as diodes yet. Is there anything else I could try?

• Harry Klippton
• Replies: 21
• Forum: General Questions

 

[Ginsly]

You might check out this thread:

Thread 'Any use for noisy Ge transistors?'

Nov 11, 2023

I've got a small handful of RCA germanium pnp transistors with decent gain for fuzz faces. DCA 55 reports no leakage too, but they're very noisy in every circuit I've tried them in. I tried them in some broadcast builds and fuzz face circuits on the breadboard and they were unusable. I haven't tried using them as diodes yet. Is there anything else I could try?

• Harry Klippton
• Replies: 21
• Forum: General Questions

That unity gain stage idea in post #8 is interesting! Looks to be set up for npn, not exactly sure if the pnp version would differ outside of switching to V- power…

 

[swelchy]

I generally socket my trannies so I can pull them if needed to swap a different set for a ge fuzz build... but you absolutely could build a deciated pedal setup just for checking your trannies to see how they sound.. something simple like a sunflower fuzz where you'll be able to dial in bias and such just to audition the trannies.... or a breadboard... Breadboard just seem fussy and unorganized chaos for me...

 

[darwin999]

That unity gain stage idea in post #8 is interesting! Looks to be set up for npn, not exactly sure if the pnp version would differ outside of switching to V- power…

I don't know how to draw a circuit diagrams on this web page, so I'll be lazy and just point to someone else's circuit drawing.
Check out figure 3 in https://en.wikipedia.org/wiki/Common_collector

 

[Ginsly]

I don't know how to draw a circuit diagrams on this web page, so I'll be lazy and just point to someone else's circuit drawing.
Check out figure 3 in https://en.wikipedia.org/wiki/Common_collector

I only see the NPN diagram again in that link, not PNP like I’m shootin’ for… maybe I’m missing it?

 

[darwin999]

Maybe I'm imagining it...
Figure 3 -

File:PNP emitter follower.svg - Wikipedia

en.wikipedia.org

 

[owlexifry]

Is it simply too situational to even do a preliminary test like I’m describing?

probably.

lets suggest you do decide a circuit, perhaps a unity gain stage to do preliminary testing to 'weed out' the bad/noisy ones.
IMO, this is probably a waste of time. because as you say, and as many others suggest, whenever dealing with Ge transistors for a particular circuit, it is best to breadboard the circuit in question (or use sockets on a board-build) and audition 'selected' Ge transistors to see which sounds/works best to your ears.

then there's the whole aspect of leakage / bias voltages, and selecting Ge transistors (perhaps of a certain hfe) to suit pre-determined values, and/or adjusting collector (or emitter) resistor values to achieve certain idle voltage/bias targets.

so IMO, if there's any preliminary testing to be done, it would be to invest that time into measuring and sorting the true hfe and leakage of each transistor into lots. (or you could even label each one individually with the measured specs).
so then when the time comes to auditioning transistors in a circuit, you'll have the benefit of selecting ones to suit based on circuit prerequisites (gain, leakage) - as opposed to blindly auditioning a heap of transistors that you may not have any knowledge of.

here's a great circuit for this
(this is exactly what i use):

Germanium Transistor Tester

Collection of vero (stripboard) & tagboard layouts for 100s of popular guitar effects, with over 1000 verified designs. DIY your own boutique effects!

tagboardeffects.blogspot.com

 

[Ginsly]

probably.

lets suggest you do decide a circuit, perhaps a unity gain stage to do preliminary testing to 'weed out' the bad/noisy ones.
IMO, this is probably a waste of time. because as you say, and as many others suggest, whenever dealing with Ge transistors for a particular circuit, it is best to breadboard the circuit in question (or use sockets on a board-build) and audition 'selected' Ge transistors to see which sounds/works best to your ears.

then there's the whole aspect of leakage / bias voltages, and selecting Ge transistors (perhaps of a certain hfe) to suit pre-determined values, and/or adjusting collector (or emitter) resistor values to achieve certain idle voltage/bias targets.

so IMO, if there's any preliminary testing to be done, it would be to invest that time into measuring and sorting the true hfe and leakage of each transistor into lots. (or you could even label each one individually with the measured specs).
so then when the time comes to auditioning transistors in a circuit, you'll have the benefit of selecting ones to suit based on circuit prerequisites (gain, leakage) - as opposed to blindly auditioning a heap of transistors that you may not have any knowledge of.

here's a great circuit for this
(this is exactly what i use):
View attachment 92541

Germanium Transistor Tester

Collection of vero (stripboard) & tagboard layouts for 100s of popular guitar effects, with over 1000 verified designs. DIY your own boutique effects!

tagboardeffects.blogspot.com

Gotcha. Yep I kinda figured it might be a fool's errand. It's fun to audition transistors, and I do that with Si as well - it has "fixed" some circuits I thought I had screwed up!

The transistor noise situation is tricky since I'm already dealing with higher-gain, noisy fuzz circuits - it's tough to tell what's "normal" for the circuit and what is excessive transistor noise. I guess that's the real predicament. Then there's the fact that breadboarding is inherently kinda noisy.

 

[owlexifry]

Gotcha. Yep I kinda figured it might be a fool's errand. It's fun to audition transistors, and I do that with Si as well - it has "fixed" some circuits I thought I had screwed up!

The transistor noise situation is tricky since I'm already dealing with higher-gain, noisy fuzz circuits - it's tough to tell what's "normal" for the circuit and what is excessive transistor noise. I guess that's the real predicament.

i'm really not much of a fuzz expert, i assume these smaller old/vintage fuzz circuits are inherently pretty noisy, even when using silicon counterparts.
(although i may be doing something wrong. i'm really not sure if a silicon fuzz face circuit is supposed to be as noisy as what i've experienced)

Then there's the fact that breadboarding is inherently kinda noisy.

it doesn't always have to be.
a baking tray makes a pretty good ground plane / RFI shield.

 

[Ginsly]

a baking tray makes a pretty good ground plane / RFI shield.

It’s connected to the breadboard’s ground via jumper? Huh!

 

[owlexifry]

It’s connected to the breadboard’s ground via jumper? Huh!

yes.
- input jack sleeve is the central node of a star ground scheme (as it is in all of my builds).
- breadboard ground lead runs directly from this node
(baking tray and aluminium mating surfaces were filed/sanded for maximum continuity)

 

[Ginsly]

yes.
- input jack sleeve is the central node of a star ground scheme (as it is in all of my builds).
- breadboard ground lead runs directly from this node
(baking tray and aluminium mating surfaces were filed/sanded for maximum continuity)

Gonna keep this in mind- thx much!