PNP Fuzz/Negative ground = problems

polifemo · Post by **polifemo** » 03 Jan 2012, 20:04

A little while ago I knew nothing about Fuzz designs, today I know little bit more.

I´ve understood that a PNP Fuzz "fixed up for negative ground" means problems, but I did not know this when I ordered a Tonebender Mk-lll kit of such a kind.

Here´s the schematic

I built the pedal today, and it does work and sound quite OK, but something MUST be wrong...

I can't bias it any lower than a little below 6V no matter what I do.
I´m using a 22k trimpot, instead of the 18k resistor, but even when I turn it down to "min" the bias remains about 6V... (I´ve tried fixed resistor but with the same results)

I get these voltage readings:

- Q1

Collector = 5,73V
Base = 7,37V
Emitter = 7,48V

- Q2

Collector = 5,74V
Base = 7,48V
Emitter = 7,68V

- Q3

Collector = 5,72V
Base = 8,88V
Emitter = 7,99V

I'm new to this, but something is really wrong, right? (This is my first germanium fuzz and I really don´t know much about leakage and such...)

The pedal also does oscillate at the highest gain settings with the mine toggle switch in the "fat mode"... (In this mode there´s also a noticable low frequency/ground hum)

I really would need, and apprciate, some help!

DrNomis · Post by **DrNomis** » 03 Jan 2012, 21:13

It's possible that the reason why you're getting those voltage readings is that you're measuring them with respect to the negative grounding system, rather than with respect to Q3's Emitter, if you convert the circuit back to a Positive Ground, you may find that most of the problems will be cured, particularly the oscillation problem, I've read stories where some people built a PNP Germanium Transistor Fuzz Face, and had problems with the circuit "Motor-Boating" (oscillating at a low-frequency) at high Fuzz settings...

polifemo · Post by **polifemo** » 03 Jan 2012, 21:34

DrNomis wrote:It's possible that the reason why you're getting those voltage readings is that you're measuring them with respect to the negative grounding system, rather than with respect to Q3's Emitter, if you convert the circuit back to a Positive Ground, you may find that most of the problems will be cured, particularly the oscillation problem, I've read stories where some people built a PNP Germanium Transistor Fuzz Face, and had problems with the circuit "Motor-Boating" (oscillating at a low-frequency) at high Fuzz settings...

Sorry, but I dont know how to convert it to positive ground as this is a pre made PCB

Later on I´ll probably use the box, pots, trannies and stuff and build a veroboard version of the pedal.

I HAAAAAATE PCBs...

DrNomis · Post by **DrNomis** » 03 Jan 2012, 21:40

It's pretty easy to do, have a look at the schematic you posted, see where Q3 has it's emitter (the terminal with the little arrow) connected to?, well that part of the circuit becomes the positive ground system....

And see all the points in the circuit that have the ground symbol on them?, well they all get connected to the battery's negative terminal....

This is assuming that all your Germanium Transistors are PNP types...

Also note that the ground terminal of the 100kA Volume pot gets connected to the positive ground system....

earthtonesaudio · Post by **earthtonesaudio** » 04 Jan 2012, 00:52

it does work and sound quite OK

So why are you convinced something is wrong?

Those voltages look normal, with the exception that most of us are used to seeing them the other way around. Oscillation/hissing/whining at max gain is a very common malady of fuzz circuits. The hum may just be ambient hum being amplified a lot.

The peculiar problems that can only be fixed by switching back to pos. ground are RARE.

The issues you described seem more in line with high gain circuits in general. I would not worry about the limited bias range. For the other problems try limiting the gain and/or bandwidth (this is a very general purpose approach to your problem and it applies to many other similar problems as well).

Good luck!

polifemo · Post by **polifemo** » 04 Jan 2012, 08:34

Well, I found it so weird that the bias did not adjust lower than it did and as I´m not all that experienced in building pedals (this is my 5:th build or something) I figured that I had screwed something up. (I usually do...)
My previous build was a hybrid Fuzz Face (Si/Q1 and GE/Q2 and when adjusting the bias trimmer on that one the voltage readings adjusted nicely thru the pots entire range.
Maybe aTone Bender Mk-lll is a more sensitive creature?

Today I´ll try a large cap from the power supply to ground as well as tying to adjust the wiring a bit.
Should one run the power supply and LED leads on one side of the box and the rest on the other?
I´ll also try shielded cables on the in/output + in/out from the board.

Does this sound like reasonable idea?

DrNomis · Post by **DrNomis** » 04 Jan 2012, 09:12

polifemo wrote:Well, I found it so weird that the bias did not adjust lower than it did and as I´m not all that experienced in building pedals (this is my 5:th build or something) I figured that I had screwed something up. (I usually do...)
My previous build was a hybrid Fuzz Face (Si/Q1 and GE/Q2 and when adjusting the bias trimmer on that one the voltage readings adjusted nicely thru the pots entire range.
Maybe aTone Bender Mk-lll is a more sensitive creature?

Today I´ll try a large cap from the power supply to ground as well as tying to adjust the wiring a bit.
Should one run the power supply and LED leads on one side of the box and the rest on the other?
I´ll also try shielded cables on the in/output + in/out from the board.

Does this sound like reasonable idea?

Wires carrying power should be fine to run anywhere, but you need to be careful with wires carrying signals, especially the input wiring since you're working on a High-Gain Fuzz circuit, try and keep the input wiring as far away as physically possible to the output wiring, this will reduce the possibility of oscillation, you don't really need to use shielded audio wiring for the output, but shielded audio wiring for the input is a good idea and can help reduce hum levels, you only need to ground the shield at one end of the wiring....

polifemo · Post by **polifemo** » 04 Jan 2012, 13:21

Thanks again for sharing your knowledge!

Today I tried soldering a 220uF cap from power input to ground input on the pcb, but the motorboating still was there...

I´moved some cables around, but that didn´t do much either. Next step will be shielded input/output cables + a serious attempt at cleaning up the wiring.

I also reduced the 68nF cap to 22nF which led to that I know can go a little bit higher on the drive knob, but past 4 o´clock the motorboating begins.

I a way the pedal is fine as it sounds really good, but I´d like to lower the noise floor, be able to use max gain as well as be able to get it to bias at 4,5V just to hear what happens to the tone

The pedal already compares well to a Skinpimp Mklll, which is´t all that bad, but it is slightly less "Fuzzy" and with more of a distortion character due to the higher bias voltage.
All n´all quite good BUT I want to learn about Fuzzes, not just only have another great sounding pedal.

Thanks again.

DrNomis · Post by **DrNomis** » 04 Jan 2012, 14:15

There's some good info on Fuzzes on this website:

http://fuzzcentral.ssguitar.com/

polifemo · Post by **polifemo** » 05 Jan 2012, 06:57

The pedal is now neatly wired using shielded cable for input/output of the pedal and in/out, to/from the circuit board but the motorboating remains from 4 o'clock and up on the drive knob.
I can live with this as there's plenty of gain at tap within the usable drive knob range and when having the toggle switch in the 6,8nF-position I can use max gain.

The pedal sounds awesome, but as I'm a tonefreak/tweaker nerd I'm a bit disturbed by the fact that I can't control the bias with my trimmer knob (by the way, would a trimmer on Q1´s 47k have been a better idea instead of my trimmer instead of Q3´s 18k?) and that this motorboating won't go away

But as I said, the pedal kicks ass

DrNomis · Post by **DrNomis** » 05 Jan 2012, 07:11

Replacing Q1's 47k resistor with a trimpot won't really affect the biasing of Q3, since Q3 is isolated from Q1 and Q2 by the coupling capacitor (0.22uF) on Q1 and Q2's collectors, but it will allow you to tweak the biasing of Q1 and Q2, I think I remember building this circuit last year and with the stock resistor values, I had Q3 biasing-up with about 4.5V on it's collector, what you could try doing is see if changing the value of the 10k resistor that's between the 18k resistor and the negative supply (the one just above the 0.22uF coupling cap in the schematic you posted), say something like 22k, or 27k, and see if that gets Q3's collector sitting at 4.5V...

polifemo · Post by **polifemo** » 06 Jan 2012, 16:24

DrNomis wrote: what you could try doing is see if changing the value of the 10k resistor that's between the 18k resistor and the negative supply (the one just above the 0.22uF coupling cap in the schematic you posted), say something like 22k, or 27k, and see if that gets Q3's collector sitting at 4.5V...

Would that resistor on the schematic be the one marked in red on the layout?

I´ve noticed that if I use 18k as resistor on Q3 the motorboating goes away BUT:

- The pedal gets quite sputtery and grainy
- There´s a huge volume drop
- The bias on Q3 now is about 9v!!!!!????....

Seems like 18k is needed on Q3 in order to get rid of the motorboating.

Do you think that adjusting this 10k resistor - to about 22k-27k - will make it possible to get Q3´s bias down to about 4,5v without these strange behaviours (grainy/sputtery tone/volume drop) of the circuit?

When I´m sure about which 10k resistor to tweak, I will of course give it a try

polifemo · Post by **polifemo** » 07 Jan 2012, 09:02

Hi,

I´ve been going over the soldering/component placement for hours trying to get this pedal right, but I can´t find any mistakes from my side.

With 18k on Q3´s collector the motorboating goes away BUT the sound is now "the sound of a misbiased fuzz" (sputtery, grainy, harsh...) and the collector of Q3 now reads almost 9v!!!???...

I did tweak the value of the 10k resistor that I marked in my previous post, and I found that by lowering it´s value the bias of Q3 goes down. The pedal now sounds much better BUT the motorboating (from 4 o'clock on the drive) comes back...

Every attempt at trying to bias Q3 by changing values on resistors seem to put the circuit of balance, creating this motorboating and hum...

Someone, on another forum, suggested that it´s the diode that biases Q3, and maybe it´s the diode that is bad?
Or maybe I got the wrong one?
Is it the 1N60 GE that is doing the bias of Q3?

I´m really frustrated by all this.
I´m quite experienced at soldering both pedals and amps - so I´m not a total newbie - but right now I can´t get this pedal to work and sound right.

I must chose between a pedal that does not hum or go into motorboating but that doesn´t sound very good, or a pedal that sounds great but has a lot of hum and where I can´t use the full range of the drive knob...

DrNomis · Post by **DrNomis** » 07 Jan 2012, 09:46

polifemo wrote:
DrNomis wrote: what you could try doing is see if changing the value of the 10k resistor that's between the 18k resistor and the negative supply (the one just above the 0.22uF coupling cap in the schematic you posted), say something like 22k, or 27k, and see if that gets Q3's collector sitting at 4.5V...
Would that resistor on the schematic be the one marked in red on the layout?

I´ve noticed that if I use 18k as resistor on Q3 the motorboating goes away BUT:

- The pedal gets quite sputtery and grainy
- There´s a huge volume drop
- The bias on Q3 now is about 9v!!!!!????....

Seems like 18k is needed on Q3 in order to get rid of the motorboating.

Do you think that adjusting this 10k resistor - to about 22k-27k - will make it possible to get Q3´s bias down to about 4,5v without these strange behaviours (grainy/sputtery tone/volume drop) of the circuit?

When I´m sure about which 10k resistor to tweak, I will of course give it a try

Yep, that's the resistor I was referring to, you could also try replacing it with a 20k trimpot and adjust it till you get Q3 biased right....

That 10k resistor and the 18k resistor form a voltage divider which works with the tone control network....

What the diode is doing, is acting as a Temperature-Stabilizer for Q3 as well as biasing it, if you look at the circuit of a Tonebender Mk1 or Zonk Machine, you'll notice that two of the three transistors have a resistor going from their Base to circuit ground, these resistors bias the transistor by diverting some of the transistor's inherent leakage current to circuit ground, the diode works in a similar way but because it's a Germanium diode it's characteristics are temperature sensitive too, so as Q3's characteristics change due to the ambient temperature, the diodes characteristic change opposes that of Q3, diverts more of Q3's leakage current to ground and stabilizes it....

polifemo · Post by **polifemo** » 07 Jan 2012, 10:01

Thanks again for taking your time to reply to my endless questions

I did tweak that resistor, and it was a lower value that reduced the bias of Q3 BUT as soon as I get the bias down a bit the hum and motorboating starts...

With original values there no hum and no motorboating but the pedal IS misbiased which one can tell not only from the harsh, sputtery grainy tone, with that horrible decay of the notes, but also from the fact that Q3´s collector now is close to 9v!

Here are the voltage readings:

Q1
E= 7.51V
B= 7.39V
C= 5.60V

Q2
E= 7.62V
B= 7.51V
C= 5.62V

Q3
E= 8.99V
B= 8.89V
C= 8.76V

Compared to another build, by someone else that was kind enough to send me his readings of the same kit, the readings are quite close until we reach Q3´s collector:

Q1
E= 7.72V
B= 7.61V
C= 6.42V

Q2
E= 7.82V
B= 7.71V
C= 6.42V

Q3
E= 9.38V
B= 9.28V
C= 6.35V

I´ve started to think about the 1N60 diode as a cause of the problem (It does have SOMETHING to do with the voltage of Q3) and when checking the voltage of that one I get 9v on one side and 8,9v on the other.
Is that OK?

DrNomis · Post by **DrNomis** » 07 Jan 2012, 10:42

polifemo wrote:Thanks again for taking your time to reply to my endless questions

I did tweak that resistor, and it was a lower value that reduced the bias of Q3 BUT as soon as I get the bias down a bit the hum and motorboating starts...

With original values there no hum and no motorboating but the pedal IS misbiased which one can tell not only from the harsh, sputtery grainy tone, with that horrible decay of the notes, but also from the fact that Q3´s collector now is close to 9v!

Here are the voltage readings:

Q1
E= 7.51V
B= 7.39V
C= 5.60V

Q2
E= 7.62V
B= 7.51V
C= 5.62V

Q3
E= 8.99V
B= 8.89V
C= 8.76V

Compared to another build, by someone else that was kind enough to send me his readings of the same kit, the readings are quite close until we reach Q3´s collector:

Q1
E= 7.72V
B= 7.61V
C= 6.42V

Q2
E= 7.82V
B= 7.71V
C= 6.42V

Q3
E= 9.38V
B= 9.28V
C= 6.35V

I´ve started to think about the 1N60 diode as a cause of the problem (It does have SOMETHING to do with the voltage of Q3) and when checking the voltage of that one I get 9v on one side and 8,9v on the other.
Is that OK?

Hmmmm...that indicates that there's about a .1V drop across the diode, normally the forward voltage-drop of a Germanium diode is about .3V so you should be seeing about 8.7V (9V-.3V=8.7V) on Q3's base, so this indicates that Q3 is being biased almost fully-off by the Ge diode, what you could try doing is try replacing the diode and see if Q3 biases-up normally, if not, I'd be suspecting Q3, I've done a re-draw of the schematic you posted for clarity, I scanned it into my computer so I can post it here for you:

polifemo · Post by **polifemo** » 07 Jan 2012, 11:26

A zillion thanks

I´havan´t got any diode right now so I´ll have to order some.

I´ll be back (if anyone ever doubted that

)

polifemo · Post by **polifemo** » 23 Jan 2012, 12:57

Fixed!

I swapped out the diode and the Q3 transistor and the pedal now rocks big time