Green Ringer diodes and Q2 low gain

[Ell]

I just build my first successful Green Ringer, and I am over the moon. I have a Visual Sound Angry Fuzz which I always wanted to do a full clone of, but now I feel like I have the most important part, the Green Ringer.

Anyway, it sounds great, but I used a few lazy parts.

For diodes I just went ahead with a matched pair of 1n4007's. Is this a bad choice? I used the diode test mode on my multimeter and it gave the same number for the two that I used, is that the correct way to match them? The DMM reads 542. Is it worth seeking out a pair of Schotty's like in Mark Hammer's thread, or is this close enough?

I only had one 2n3906 hanging around, so I just used the one I had. I should probably test the gain of that, and if it's high, buy some more right?

I am delighted with the effect as it currently is, and it basically sounds as good as the Angry Fuzz... but could it get better? I will be feeding it with a fuzz face quite often, with an overdrive afterwards to give it some compression.
Also, I can't decide if it sounds better with the Nurse Quacky Envelope filter before or after, any opinions on that?

[Mark Hammer]

1)  Good on ya!

2) Virtually ANY fairly equivalent two diodes will "work".  But since the diodes each act as a sort of "gate" for their respective half cycle, lower forward voltages will provide more of each half-cycle and a bit less crossover distortion.  Schottky will be "better" in that regard, but germanium or silicon will be fine, if the two you use are reasonably close in Vf.

3) The Green Ringer never really provided any sort of sensitivity adjustment.  What I like to do is stick a 10k variable resistance in series with a 10uf cap in parallel with the 6k2 emitter resistor in that first stage, to vary the gain of that stage and compensate for inputs of different amplitudes.

4)  "Better" octave-up units will often stick a pair of diodes to ground closer to the output.  Folks mistakenly think that's where the fuzz comes from.  But if you lift them, the circuit remains 95% as fuzzy as ever.  Rather, they provide a quick and dirty peak limiter.  Because picked strings generate a lot of harmonic content, initially, the octave gets lost in the doubling of all that harmonic haze, and takes a moment or two to "bloom" as the note fundamental becomes more prominent (this is also why octave up AND down pedals generally sound best when picking above the 7th fret, where harmonic content is reduced and the fundamental is most prominent).  The diode pair to ground attempts to keep the volume level somewhat constant, such that the octave bloom is better heard.  One can use a back-to-back pair of silicon or germaniums for the task.  Digi2t recommends sticking a small resistor - say 100-470R - between the diodes and ground to soften their clipping, and frankly I agree.  A small value cap in parallel with that diode/resistor network, like 1000-2200pf will help to further remove any unwanted fizz.

[antonis]

Diodes serve as 2 X half-way rectification so, IMHO, as long as their forward voltage drop is almost equal, it should be OK..

2N3906 serves as a phase splitter (both Collector & Emitter gain approximately unity) with a working current of about 225μA (ideally)..
That current (sink/source) divided by its gain is needed from Q1 Collector..
(without involving formulas, it's more than OK for any BJT of medium to low h_FE..)

edit: Mark already responded much more extensively,,

[Steben]

1)  Good on ya!

2) Virtually ANY fairly equivalent two diodes will "work".  But since the diodes each act as a sort of "gate" for their respective half cycle, lower forward voltages will provide more of each half-cycle and a bit less crossover distortion.  Schottky will be "better" in that regard, but germanium or silicon will be fine, if the two you use are reasonably close in Vf.

3) The Green Ringer never really provided any sort of sensitivity adjustment.  What I like to do is stick a 10k variable resistance in series with a 10uf cap in parallel with the 6k2 emitter resistor in that first stage, to vary the gain of that stage and compensate for inputs of different amplitudes.

4)  "Better" octave-up units will often stick a pair of diodes to ground closer to the output.  Folks mistakenly think that's where the fuzz comes from.  But if you lift them, the circuit remains 95% as fuzzy as ever.  Rather, they provide a quick and dirty peak limiter.  Because picked strings generate a lot of harmonic content, initially, the octave gets lost in the doubling of all that harmonic haze, and takes a moment or two to "bloom" as the note fundamental becomes more prominent (this is also why octave up AND down pedals generally sound best when picking above the 7th fret, where harmonic content is reduced and the fundamental is most prominent).  The diode pair to ground attempts to keep the volume level somewhat constant, such that the octave bloom is better heard.  One can use a back-to-back pair of silicon or germaniums for the task.  Digi2t recommends sticking a small resistor - say 100-470R - between the diodes and ground to soften their clipping, and frankly I agree.  A small value cap in parallel with that diode/resistor network, like 1000-2200pf will help to further remove any unwanted fizz.

If I read this correctly, the ideal effect in front of an octave (fuzz) is a compressor or even better a limiter?

[Ell]

1)  Good on ya!

Thank you. I found your past posts invaluable, for this and for every other pedal I have ever built. I had a few problems building this, but I used all of the debugging techniques that I have read about on this forum, and I figured out the issues. This forum is the best.

2) Virtually ANY fairly equivalent two diodes will "work".  But since the diodes each act as a sort of "gate" for their respective half cycle, lower forward voltages will provide more of each half-cycle and a bit less crossover distortion.  Schottky will be "better" in that regard, but germanium or silicon will be fine, if the two you use are reasonably close in Vf.

I have found a pair of 1N34A diodes. They won't sit still when I measure them, but they seem to hover around a similar range, so maybe I will throw them in and see how they sound. It is difficult to get Schotty diodes where I live.

3) The Green Ringer never really provided any sort of sensitivity adjustment.  What I like to do is stick a 10k variable resistance in series with a 10uf cap in parallel with the 6k2 emitter resistor in that first stage, to vary the gain of that stage and compensate for inputs of different amplitudes.

Ah, so that is this part that you said three years ago:

...Stick a 10k linear pot in place of the original 6k2, one outside lug going to ground, the other going to the emitter of Q1.  A 10uf cap goes from the wiper of the pot to ground (+ side tied to the wiper).

I will have to try this too!

4)  "Better" octave-up units will often stick a pair of diodes to ground closer to the output.  Folks mistakenly think that's where the fuzz comes from.  But if you lift them, the circuit remains 95% as fuzzy as ever.  Rather, they provide a quick and dirty peak limiter.  Because picked strings generate a lot of harmonic content, initially, the octave gets lost in the doubling of all that harmonic haze, and takes a moment or two to "bloom" as the note fundamental becomes more prominent (this is also why octave up AND down pedals generally sound best when picking above the 7th fret, where harmonic content is reduced and the fundamental is most prominent).  The diode pair to ground attempts to keep the volume level somewhat constant, such that the octave bloom is better heard.  One can use a back-to-back pair of silicon or germaniums for the task.  Digi2t recommends sticking a small resistor - say 100-470R - between the diodes and ground to soften their clipping, and frankly I agree.  A small value cap in parallel with that diode/resistor network, like 1000-2200pf will help to further remove any unwanted fizz.

Is the effect of this completely different from just placing an overdrive after the GR?

I'm now thinking that maybe I should incorporate all of these mods on switches. When I decide on the combinations which suit me best, I'll just remove the switches and make it permanently that way.
The thing is, I do LIKE the sound of the stock version, I think I'm just getting greedy with the idea of ultimate fuzz power. Unnecessity is the mother of reinvention.

[antonis]

I have found a pair of 1N34A diodes.  It is difficult to get Schotty diodes where I live.

Try to heat them..

[CheapPedalCollector]

Good work, great sounding circuit.

I've built them before and put an extra bit of gain in front with a gain control replacing the emitter resistor on Q1 and a bypass cap of say 2.2uf on the center wiper to ground. I've tried anything from 1k to 5k for the control, use what you think sounds best. You can change Q1 collector resistor too to anything between about 15k to 47k to add or remove gain. This will give the fuzz a bit more oomph. Low gain devices definitely work best, I found 120-150 hfe is ideal but 80-200 will work just fine. I prefer to use 1n270 or 1n60 for the diodes, but schottky diodes also work well. The voltage divider is set to have a small forward bias through the diodes to the base of Q3 to provide the base bias. A balance control can be added via changing the two 22k resistors into 20k resistors and adding a 5k linear pot. The input impedance is pretty low too so you can add bass by increasing the input cap for a thicker sound, 1uf is nice to my ears.

Have fun, this circuit is fun and flexible.

[Ell]

I have found a pair of 1N34A diodes.  It is difficult to get Schotty diodes where I live.

Try to heat them..

Nice trick! Thank you Antonis!
I get a reading of 385 from one, and 410 from the other. So, I'm probably better off sticking with my matched pair of 1N4007 diodes at 542 aren't I?
So, as long as I'm measuring the forward voltage correctly, I have the choice of matched higher forward voltage, or unmatched lower forward voltage. It sounds like the equality is important because of how the waveform fold over itself right? (If I'm understanding everything correctly, this is a pretty complicated concept for me).

[Ell]

Good work, great sounding circuit.

I've built them before and put an extra bit of gain in front with a gain control replacing the emitter resistor on Q1 and a bypass cap of say 2.2uf on the center wiper to ground. I've tried anything from 1k to 5k for the control, use what you think sounds best. You can change Q1 collector resistor too to anything between about 15k to 47k to add or remove gain. This will give the fuzz a bit more oomph. Low gain devices definitely work best, I found 120-150 hfe is ideal but 80-200 will work just fine. I prefer to use 1n270 or 1n60 for the diodes, but schottky diodes also work well. The voltage divider is set to have a small forward bias through the diodes to the base of Q3 to provide the base bias. A balance control can be added via changing the two 22k resistors into 20k resistors and adding a 5k linear pot. The input impedance is pretty low too so you can add bass by increasing the input cap for a thicker sound, 1uf is nice to my ears.

Have fun, this circuit is fun and flexible.

I have a lot of homework to do. Thank you very much for the diode alternatives! I will try out what I can.

[Ell]

I'm probably being stupid here somewhere, but I've tried to test my 2n3906 gain using R.G. Keen's transistor test. I made it with 6 resistors adding up to almost exactly 2m2, and 4 resistors almost exactly adding up to 2.46k.
http://www.geofex.com/Article_Folders/ffselect.htm

When I measure the voltage across the 2.46 resistor, WITH the 2.2M resistor attached to the transistor's base I get 2.4 volts. My power supply is set to exactly 9V.
Am I right in thinking that this means that my gain is more or less 240, and that this is higher than the ideal range... and that I should buy some more transistors? When I flick the switch to disengage the 2m2 resistor to base I get 0, meaning that there is no leakage (it's silicon). Please let me know if I've made mistakes with any of this.

Out of curiosity, I measured some Germanium transistors I have lying around.
AC128 #1:    0.34v leakage test, 0.75v total gain test. If I'm right, this means that the true gain is only 41 (75-34), and there is a lot of leakage... This wouldn't make a good fuzz face would it?
Trying to use the maths correctly, (0.34 / 2472) * 1,000,000 = 137uA of leakage.

AC128 #2:    0.72v leakage test, 1.25v total gain test.
So, 125 - 72 = 53 true gain, right?
(0.72 / 2472) * 1,000,000 = 291uA of leakage.

Should I even bother to make a Germanium fuzz face out of these? Are they good enough? I generally use a silicon fuzz face with 2n2222's.

OC44:          0.19v leakage,       1.05v total gain test.   
105 - 19 = 86 true gain
(0.19 / 2472) * 1,000,000 = 76uA of leakage

[CheapPedalCollector]

240 is a little high but they should work fine, its really hard to find low gain silicon anymore.

The AC-128s would be better in a Tone Bender with that much leakage.

the OC-44 is pretty much ideal for a Rangemaster.