FuzzFace and AC127

[arstoffel]

I am building a FuzzFace, and I intend to use AC127 transistors. By now, I bought 3 of them; they have low gain (between 42 and 73) and low leakage (between 24 and 58uA).

I wouldn't say they sound GOOD, but it seems that if I buy a bunch of them, I will find a pair that works well. There is one thing I found strange, though. Playing two (or more) notes produces a really ugly effect, like a Ring Modulator.
(http://www.harmony-central.com/Effects/Articles/Ring_Modulation/ring-modulation-ss1.wav --- it's more or less like that)

In terms of circuitry, there are three great differences between my FuzzFace clone and the original: the kind of transistor, the gain of the transistors and (as a consequence) the DC voltage levels. What may be causing the terrible intermodulation effect?

Yet another question: Do the AC127 have any reputation?

[Fret Wire]

Post a link to the schematic you are building this with. Here's a good sticky to read when you have problems.
http://diystompboxes.com/sboxforum/viewtopic.php?t=29816

Now, in the sticky two things that come to mind are that AC127's are NPN. Are you using a neg. ground layout? Secondly, you should post the tranny lead voltages. Did you attempt to bias the fuzz, or just use the values on the schematic?

I've got gains that low to work fine before. Just use the two highest ones, putting the highest in the Q2 postition. You should still get proper voltage readings even though the gain is lower. And there is nothing the matter with AC127's as long as their leakage is low and they are stable.

[brett]

Hi.  I've tried AC127s before.  My experiences aren't quite as bad as yours, but they haven't been good, either.  I found it hard to get low leakage devices.  Most I bought came in at 300 to 800mA using RG's test rig.  hFE was higher than your range, but that only emphasised the noise of these suckers.  I was never happy with the result and headed down the Si path....

But keep digging, there's no reason why you couldn't "snag" a good result.

cheers

[arstoffel]

My circuit is just a plain NPN FuzzFace; in any case, I should have posted a schematic.
http://www.muzique.com/schem/fuzzface3.gif
R2 is 470ohm, C2 is 22u and Q1 and Q2 are both AC127.

As I wrote, it works well, the only problem is that playing two notes at once sounds bad.

One thing I forgot to say it that playing an open D, for example,  sounds OK. It becomes really terrible if I play these same notes one octave higher. (Sorry not to descriebe the sound better, you can see it would be really difficult to me. But it really resembles that ring modulator.)

Replacing the 33k resistor with 47k improves the tone a lot, but slightly changes the intermodulation effect. The voltages I read are as follow:

With 33k:
Q1 --- base: 0.08V; collector: 0.72V
Q2 --- emitter: 0.61V; collector: 3.72V

With 47k:
Q1 --- base: 0.07V; collector: 0.5V
Q2 --- emitter: 0.42V; collector: 5.4V

The circuit is powered with a power supply, about 8.8V.

[Fret Wire]

Do you know the gain and leakage of the transistors? Btw, post the voltages for collector, base, and emitter for both transistors.

[arstoffel]

Q1 gain is 50 and Q2 gain is 72.

I noticed that the leakage depends on the temperature. Today I measured 133uA and 335uA for Q1 and Q2 respectively. About two weeks ago it was colder, and I measured 24uA and 58uA for Q1 and Q2 respect. (the gain doesn't seems to depend on temperature).

Q1 emitter is grounded, and Q2 base goes to Q1 collector, therefore

With 33k:
Q1 --- base: 0.08V; collector: 0.72V; emitter: 0V
Q2 --- base: 0.72V; collector: 3.72V; emitter: 0.61V

With 47k:
Q1 --- base: 0.07V; collector: 0.5V; emitter: 0V
Q2 --- base: 0.5V; collector: 5.4V ; emitter: 0.42V

[Fret Wire]

Your voltages are proportionate. Leakage on Q2 is a little high. Drop down the 47k down one or two values. Double check your board, jacks, pots, and switches for any cold solder joints. Also check for a slightly bridged solder joint or copper trace. The biasing isn't that far off at all, probably a minor problem elsewhere.

[brett]

Hi.

With 33k:
Q1 --- base: 0.08V; collector: 0.72V
Q2 --- emitter: 0.61V; collector: 3.72V

Something VERY odd here.  Q1 base is always 0.3V in a correctly wired NPN fuzzface.  

I suspect that there's some unwanted grounding going on somewhere in the voltage feedback loop (ie the 100k resistor).  Another possibility is low resistance in the pot (due to a dud pot or 22uF cap or bad wiring).  Check that it's within a hundred ohms of 1k from lug 3 to ground.

cheers

[arstoffel]

I have checked my circuit, it seems to be OK. But there is something wrong about the transistors. I put them in the transistor socket of a DMM, and then measured the base-emitter voltage with another DMM. It was always about 100mV.

When you buy germanium transistors, do they look brand new? The ones I bought have bent and oxidized leads. They do not look new at all (although they have very long leads). If they were removed from an old circuit, they were probably not desoldered with enough care...

Here is a photo (notice the rust). What do you think?

//www.inf.ufrgs.br/~arstoffel/00003.jpg

[Fret Wire]

They appear to be unused, but poorly stored. Take some fine sandpaper, and clean the leads shiny. Try measuring their gain and leakage again. Your ckt may work right now.

Btw, a DMM won't check the gain of germanium transistors accurately, because it doesn't take their leakage into consideration. It will incorrectly show their gain higher than it really is.