The Sonic Tonic - A new design

[Michael Weidenauer]

The first idea came into my mind when I first read R.G.'s article about using leaky germaniums as a buffer: Why not overdrive an emitter follower running on lower supply voltage?
The second idea came when I read, that EFs clip the negative cycle if the following load gets too low (So let's load it!).

Here is the circuit:


The signal gets amplified through a non-inverting OP and overdrives a NPN-germanium emitter follower with a supply voltage of less than 1V. The transistor is biased by a variable voltage-divider that sets the input symmetry (Sym1-pot). The output can be loaded via the Sym2-pot to get further asymmetric clipped tones.

The mode switch:
The normal setting is Uc stabilized with the 220µ cap;
The loose settings using 220n or no cap, produce overshots on the positive signal swings.

The sound is very versatile and covers many shades of overdrive and fuzz, including Fuzz Face-like sounds and nice octave-up sounds.
I made some scope pictures to demonsrate the function of the sym-knobs:


All NPN germaniums I've tried worked well (including leaky, low- and high-gain ones) with slightly differing sounds.
The circuit also works with silicon-trannies but you have tweak the bias network (raise the 4k7 resistor) and maybe also raise Uc.
But anyway: Germanium is the real deal!

Have Fun!

[bioroids]

Sounds interesting!

Do you have some sound samples so we can listen to it? You may have found use for hundreds of useless ge transistors!

Regards

Miguel

[Michael Weidenauer]

Don't have any sound samples yet, but as stated above it's very versatile:
It can sound quite like a Distortion+, it can be mean hard clipping like a RM Axis Fuzz and can sound like a good germanium Fuzz Face (also the scope pictures look quite similar).

It is still a pre-version(0.3) that has not left my breadboard yet. Haven't even been fooling with the input- and coupling-caps.

I am allready thinking about a second version, with 9V supply for the tranny that only (load-)clips the negative part of the signal, then feeds an inverting OP driving another loaded EF-stage. The load-clipping seems to be quite round edged (soft clipping) so I have to try.

Build it, tweak it and post it!

[Michael Weidenauer]

I reposted the pix to another host for better viewing:



and here the scope pictures:



The negative-supply schematic for PNP Germaniums (didn't test this but I think it should work):

[mills]

I know this thread is pretty well buried, but  I'd been meaning to build this circuit since I read about it and finally did a variation.

Simplified it to have gain, tone, volume, and the pot that loads the emitter follower.  Left the sym 1 pot as a trim, and removed the other switches.  Ran the Ge transistor off a few voltages from 1 to about 5, and thought that about 2-3V was best.  Good mix of distortion tones and output off that.

It sounded great on breadboard, and the description is pretty accurate... distortion to fuzzy tones, an acceptable low gain sound, and plenty of octave-y and broken sounds.  without being too gated or squelchy.

The one I did got tweaked a bit to be boxed with a rat, but I think I'm going to build one "stock" because it was so much fun. 

[~arph]

Nice little circuit!

Had not seen this one. Might use a dual opamp, to give it a buffer after the tone control. not that much extra components

[petemoore]

  Good to see you're following up on the idea and getting nice results on the concept of distorting Ge buffers.
  Good one ! Now maybe we can make more useful time with our out-of-spec-for the usual Ge circuits: Ge transistors !
 
 
 

[Derringer]

I BB'd this concept tonight with a 4558 and Si NPN and got some really cool fuzzy sounds.
It's late, so I can't turn up much, but there's just a really cool way that the basic design lets you distort the actual waveform on the fly


great idea man

[askwho69]

sound clip is convincing

[Derringer]

sound clip is convincing

working on it

it definitely responds differently with different gain transistors
I haven't tried a FET yet, but it's on the list

I'm getting a lot of noise at different settings though so that may be an issue ... need to tinker further

also, I'm not starving the NPN, I have collector tied to V+ (~9V)

[Derringer]

see if this link works

http://www.megaupload.com/?d=J8UVNDHA

Input is a non-inverting 4558, 100KA Gain pot and 1000 pf in the neg-feedback loop,
I'm using a 500Kb pot for sym-1 with a 100 ohm limiting R on the V+ leg and 1K limiting R on the GND leg
50KB pot as a voltage divider on the collector of a 2N3904 gain of about 160 (I say 2n5089 in the mp3 but that is wrong)
Collector V is at about 5V.
emitter load pot is 10KB with no limiting R, it can go to zero ... but should have a limiting R
AC coupled to the 2nd half of the 4558 set up as a voltage follower
into a SWTC, 10K pot and 0.033uf to GND
0.1UF output cap to 100K volume

[kurtlives]

How does the 27pF connected to the op-amps offset null work?

Cool design!

[Derringer]

no I've got a mojo laden 2n2369A set up as a gain stage getting hit by the first 4558 amp with leds in the FB loop to prevent hard nast opamp clipping,

the NPN has 10K on the collector, 10KB pot on the emitter, 330 pF B-C, and a 1-Meg pot as the bias divider on the front end
out into the voltage folower, tone, volume

the 2n2369A sounds a bit cleaner than the 2n3904 I had in earlier

I have also tried J201, 2N5457, 2N5306, 2n5089 in the traditional design implementation and I think prefer the "gain-stage" design here better with the 2n3904 and much more so with this 70hfe 2N2369A


love this design ... all I'm doing is overdriving an NPN stage with an opamp I suppose ... but this really is a tone that I like a lot.
not real high gain, chunky, ballsy, lively

[mills]

Allright, I was hoping that we could talk about some modifications to this circuit. 

I'm trying round two on this and simplifying the circuit a bit, and tweak it to be something easier to use.  No switches, and this is the current ciruit I have on a breadboard (pot values are because thats all I had handy).



After the ? I've got a second op amp stage to compensate for volume loss on the more extreme settings, and was going to toss a simple low pass filter in after that.  I have a few questions...

1) the resistor thats going to provide vref to the opamp... I think thats going to mess with my loading pot?  I had a big resistor in there, but got less range outta the load pot, and a small resistor gave less range the other way.  Makes sense, but not what I want.  I went to a medium value, but is there a better way to figure out (approximately) a value to use other than just trial and error?  maybe a different way to isolate the two sections?

2) for a volume control, since theres a pretty big range of signal strength coming out of the transistor, I thought it was best to do volume control by changing the gain of the second opamp stage rather than run a constant gain and the "standard" volume control.  Is there a more clever way to do this? A reasonable way to minimize opamp clipping?  I thought about putting a "standard" volume control before the second opamp, but again, its going to mess with my "load" control. 

Thanks for any help or comments!

[Derringer]

you could try a dual opamp
and have the emitter, load pot, then ac coupled into a buffer and then into the following gain stage of your choice using the next opamp section. I imagine that would help isolate things.
are you using a germanium transistor?


what i settled on and what I have boxed up now is pretty much what I described before, except I went with a 470K R from base to V+ and 330K from Base to ground in place of the sym 1 pot. I still have 10K from collector to V+ on my 2n2369A and a 10K pot wired as a variable resistor with a 1K limiting resistor on the emmiter.
That goes through a 2200 pf cap to a volatge foloower with an impedance of 1M, then to a SWTC with a 1K, then 10K and 0.022 uf to ground, 0.1Uf then the 100K vol pot.

I love it.

[mills]

...Well, you're all probably tired of random resurrection of this thread, but almost done with exams so I finally tweaked a few more values and got it to a point that I'm happy with it. 

Here's the (messy) updated schematic



Sym 1 is a trimpot.  Kinda changes the character and range of the sym 2 knob, but I don't really think it does enough to put as a control on the enclosure.  Tone knob at the end the SWTC version two.
Only thing I want to think about trying is running the Ge transistor off Vb as well.  Save a few parts.  Also, I'd like to suck less at layouts, but thats another topic.

I'll try to post some clips this weekend since I should finally have time.