Germanium ain't cheap

[fig]

Apologies to @giovanni .

arrivedelchi

 

[Harry Klippton]

Next @Big Monk is gonna tell us he can design a transistor amp circuit that sounds exactly like a tube amp

 

[thesmokingman]

I think the best way to see the differences is to take things to an extreme ... for instance, zvex does a demo of the silicon fuzz factory where all the subtle differences between si and ge get very exaggerated and apparent. as far as some sort of challenge, guys have been trying to tame silicon fuzz faces into germanium for 50 years now and nobody's completely gotten there. we're not all suffering mass delusion and the market has borne that out over time.

 

[Harry Klippton]

guys have been trying to tame silicon fuzz faces into germanium for 50 years now and nobody's completely gotten there. we're not all suffering mass delusion and the market has borne that out over time.

 

[DAJE]

My usual setup is my Ge Tone Bender 2 and my Si Fuzz Face (both clones I assembled) next to each other. I use them and switch between them all the time, and I have often said I'm glad I don't have to choose just one. They're quite different and each has a character I enjoy very much.

 

[benny_profane]

I don’t have a terribly strong opinion here, but you’re contradicting yourself:

If we are being honest with ourselves, a well tuned Si fuzz has every bit of the sound qualities we expect from a Ge unit.

With that said, I'm not sure the sonic differences are objectively differentiated if the circuits were tweaked to ape one another. For instance, if i had a primo example of a good Ge Fuzz Face, my opinion is that I could clone that device using Si transistors and circuit tweaks, frequency shaping caps, etc. and have a legitimate chance at fooling someone in a blind test.

But…

I've never seen a circuit shootout with an Si device setup to mimic a Ge device exactly, i.e. I build a Ge device as a control then build a Si device to mimic that exactly, to the best of my abilities, and do a blind test.

I’d be interested to see what you come up with. Beyond the cachet Ge holds—and the attendant price premium—there is something specific going on with those devices. If Si is emulating Ge, then there is something that Ge offers that Si has to be engineered to provide. This is, unfortunately, not a spec listed in a data sheet. So, while I don’t believe the difference between the two is an unbridgeable chasm (see @thewintersoldier’s “95%” statement above), there are differences. Whether those differences are appreciable or even desirable (e.g., noise, unpredictability, frequency filtering), may play a big part in determining the subjective perception of the two.

 

[Coda]

I have no horses in the race, but I did build a Box 66 a few weeks ago. It was good. Over the weekend, however, I replaced the Ge with Si…the pedal sounds waaaaay more like a Vox ToneBender than it did with the Ge…

 

[Coda]

The Tonebender family, specifically the MK II, have shown time and time again, at least in my circuit building, breadboarding, etc. to really thrive with Si transistors.

For a while I used to offer a caveat to those I gave advice on the MK II/Supa about how i thought the Ge units gave the circuit a more tactile feel and that note bloom, feedback and sustain was better with Ge.

In the last few months I've been experimenting, here and there, with full Si MK II circuits. One of the things I always disliked about the schematics floating around was how they used much lower voltages on Q1 and Q3 when using Si transistors. Using the more generally accepted MK II voltages, playing with B-C bypass caps, and tweaking the frequency shaping caps has yielded MK II circuits with almost all of the tactile playing feel, note bloom, sustain and feedback of the Ge units but with an overall tighter feel to the lows and crunch.

With the Ge, it reminded me of a Ge fuzz face; kind of thick, and round…unless you playing into a hot amp, which adds the hair and grit. The Si version sounds like the later example, but into a clean amp.

I think it comes down to two things: are the properties of a Ge transistor important (like in an FZ-1), or is the function of the transistor more importanter?…

 

[Coda]

Did you use the OC transistors?

I don’t remember off hand what the Ge ones were, they weren’t anything special, maybe 2n1303, HFE around 85/130.The Si ones are 2n3906.

 

[Coda]

Having played a number of what i would consider textbook examples, I think those OC-xx transistors, the 75 and 81x in particular, have a certain something in the treble frequencies that give that characteristic Tonebender "kerrang".

I think that's why Silicon transistors work so well in those circuits.

Good observation.

 

[Coda]

Fuzz Faces are where things get more difficult.

Ge FF is great for early Hendrix. That’s about it, in my experience. Si is way more versatile…

 

[Harry Klippton]

 

[benny_profane]

So, people have different preferences and subjective preferences. Circuit design can help make different sounds. Sometimes those sounds can sound like circuits with different parts. Global statements, equivalencies, and equating subjectivity with objectivity will get people riled up.

Did I sum up this morning’s discussion?

 

[benny_profane]

If anyone is really riled up we are doing something wrong...

Plus we need fodder for "It's All in Good Fun" now that everyone is removing their white washers and has thrown their cloth wire in the garbage.

I just meant it sparked discussion—not that people were upset. Although, those things listed above are the kind of things that make people get super bummed out about TGP and other hype ad copy.

Side note: what do you think—beyond lower gain buckets—makes NOS Si better than new production?

 

[jeffwhitfield]

I think the best way to see the differences is to take things to an extreme ... for instance, zvex does a demo of the silicon fuzz factory where all the subtle differences between si and ge get very exaggerated and apparent. as far as some sort of challenge, guys have been trying to tame silicon fuzz faces into germanium for 50 years now and nobody's completely gotten there. we're not all suffering mass delusion and the market has borne that out over time.

Funny you should mention the Fuzz Factory. Been playing with my own build here. Started with a set of MP20a's and just the other day popped in some AC127's. Difference was amazing. Actually preferred the MP20a's. Now..that said, I'm super curious on what a set of silicon transistors will sound like by comparison.

 

[fig]

Funny you should mention the Fuzz Factory. Been playing with my own build here. Started with a set of MP20a's and just the other day popped in some AC127's. Difference was amazing. Actually preferred the MP20a's. Now..that said, I'm super curious on what a set of silicon transistors will sound like by comparison.

Do you think the difference was in the part number or in the measurements?

 

[jeffwhitfield]

Do you think the difference was in the part number or in the measurements?

Might be the difference in the type of transistor. PNP vs NPN. Given that the Fuzz Factory isn't exactly a biased circuit per se, it would seem that a PNP might be a bit tamer than an NPN with similar hFE measurements. I put the same AC127's in a Fuzztone '69 circuit and they just sang. Bias on the circuit makes all the difference.

 

[andare]

Silicon FFs sound harsher (which is great on stage) but a bit fizzy when rolling off some volume. I like that they keep the high end. My Strat lives at 7-8 for cleans with a hint of breakup and I have no treble bleed so Silicon is the way to go in the mix, for me.

Germanium to me sounds too dark and mushy when full on but those gain stages between full fuzz and totally clean are very similar to how my tube amp overdrives. Very organic and just beautiful.

In other words, Germanium is for playing alone or in a sparse mix like a trio, Silicon for denser mixes. Another thing I do is put the neck pickup out of phase with the other two. The OOP sounds lose the woofy low end and gain additional high end resonance. It's an acquired taste but it makes things more versatile and it works for my music.

BTW today I will attempt my first ever stripboard build, a BC108 Fuzz Face.

 

[HamishR]

I built a BC108 Fuzz Face last year. It actually worked out alright. But from memory its better with BC108Cs rather than Bs.

My favourite TB I ever built was a Marshall Supa Fuzz circuit with Russian MP38s so it could run on regular 9V. It sounds fat, vocal and lovely. But I hardly ever use it. I seem to prefer my similarly voiced overdrives.

 

[fig]

2N1308