Simple build, loosing my mind trying to get sound out!

[John Lyons]

I have been working on some complex circuits and wanted to try some simple ones...

I have joe Davisson's Antiquity fuzz on the breadboard and It's not working. I have torn it down and built it back 3 times now with new 2n5089 transistors. Using all stock schematic parts and values except a 10K trimmer for the 5.6K to bias Q2



Here's what I get:

9.5v regulated supply

2n5089 Q1
E = 0v
B = .6v
C = 5.4v

2n5089 Q2
E = 4.6v
B = 5.4v
C = 4.6v

I have gone through everything here, it's a simple circuit, am I loosing my mind?

BIG EDIT
I replaced the .01 output cap with another of the same value and I get sound now!

Here are the new voltages:
2n5089 Q1
E = 0v
B = .6v
C = 2v


2n5089 Q2
E = 1.4v
B = 2v
C = 4.6v

I get a pretty good amount of fluctuating static but the circuit is working. The output is below unity gain though.
Anyone have any experience with this circuit?

John

[Izzy]

Is that diode in right position?
May be it is supposed to be reversed. 

[tcobretti]

I would definitely check the ground for Q2; something looks wrong there.

[R.G.]

tcobretti is correct.

Simple transistor basics tells us that with 4.6V on both the collector and emitter of Q2, it cannot pass audio. With both of them pulled up by the base, it is logical to think that the 2K is not pulling Q2's emitter down. You may have a flaw which prevents the emitter lead from being pulled to ground.

If you built it two times with different transistors, I'd check to see if that 2K pot is open, or whether the connection to the place where you plug your emitters and pots is not making connections.

Breadboards wear out and quit making connection with steady use. When they get intermittent throw them away and get another one. The time you save will pay you back many times. If you have spent all this time on an intermittent breadboard, you will appreciate this.

[tcobretti]

My uneducated opinion is that the voltages still look a little weird.  However, I have seen FF's with a similar layout where the output was below unity gain.

[petemoore]

  to put the Meter right across the 2k and get readings on the pot's R values, testing for a bad connection and bad pot, from an offboard ground wire to Q2E.

[John Lyons]

The pot measures 2.4K
Basically the noise jumps up in the last 5% of the pots travel. If I back it off a hair the noise goes away.
It's not just hiss...
I'll try replacing the pot (it's a very old one) and see what I come up with.

[tcobretti]

The pot measures 2.4K
Basically the noise jumps up in the last 5% of the pots travel. If I back it off a hair the noise goes away.
It's not just hiss...
I'll try replacing the pot (it's a very old one) and see what I come up with.

This is not uncommon with FF circuits.  A shot in the dark, but you might try a 50pf cap across the collector and base of Q1; it helps with oscillation, and I suspect this noise at the end of the pot's travel is the beginning of oscillation.  BTW, you can change the output level of the pedal by messing with the 470R/8k2 resistor values.  I haven't played with it much, but I think if you reverse their positions the pedal will be louder.  You also might try a larger resistance output pot.

[mac]

Download this:

http://geocities.com/guitarfxs/exes/biascalculator.zip

This software will give you an idea of the voltages you can expect from this circuit. It does not include the diode and the 100K to gnd but the effect may be negible.

mac

[jonathan perez]

"NO MIND...NO MIND!"

[ubersam]

"NO MIND...NO MIND!"

Have you been watching The Last Samurai...

[John Lyons]

tcobretti

The noise was oscillation. The 50pf cap cured that. thanks! I'll have to look up some fuzz face notes to get this guy tuned up.
I reversed the 470R/8.2K resistors and got more volume out as well. biased it up and it's sounding prettty good.

My simple build turned into a several hours of frustration. I would have thought that this circuit would be a little more fool proof from the build reports it has gotten... Maybe a fluke of breadboard related somehow.

Thanks for the help folks.

Mac, I'll check out that calculator, thanks!

John

[petemoore]

The noise was oscillation. The 50pf cap cured that. thanks! I'll have to look up some fuzz face notes to get this guy tuned up.
  A little HF rolloff in a Vhigh gain FF...k, another 'trick' is to tune the gain pot, start by turning the 2k pot up full, clip a 10k pot across the 2k pot, adjust the 10k until the squealing stops. Measure the resistance of that setting [don't touch it] on the 10k, put a fixed R of that value across the 2k...now when you turn it up full gain, it'll be just under where the oscillation starts.
  But the way I did it is by starting with the FF pot Small Bear sells [this particular tapered pot adjusts FF Gain very smoothly], and choose a stop resistor value in much the same way as starting with a 2k and working down.
  A 2k gain pot with 2n5089 transistors...I'm guessing that would Have to self oscillate.
I reversed the 470R/8.2K resistors and got more volume out as well. biased it up and it's sounding prettty good.
  I'm not sure I understand this 'mod'. you swapped the 470 and 8k2, hmm.
  My simple build turned into a several hours of frustration. I would havethought that this circuit would be a little more fool proof from the build reports it has gotten
  Works every time, it has to.
  Maybe a fluke of breadboard related somehow.

[John Lyons]

I actually only had a 5K linear pot so I had to strap a resistor across the pot to get 2K. I have been turning the pot up to the "danger zone" and   the oscillation (mainly hiss) stops there.

With the 470R/8.2K divider I just put in the 8.2K from 9v and the 470R to the collector and the .01 from the juction of the 8.2k/470R.
The output went up a bit... I guess because of less resistance from the collector and more resistance to 9v???

Here is a quick clip of what I have going on. www.mrdwab.com/john/Antiquity.mp3
Marshall amp set almost totally clean, bridge humbucker.

John

[tcobretti]

With the 470R/8.2K divider I just put in the 8.2K from 9v and the 470R to the collector and the .01 from the juction of the 8.2k/470R. The output went up a bit... I guess because of less resistance from the collector and more resistance to 9v???

If my logic is correct, the overall resistance between +9v and Q2's collector remains the same (470R + 8.2k= 8.7k), so Q2's bias doesn't change, but the resistance between the collector of Q2 and the output decreases, which makes it louder.  My Mondo Fuzz (Si FF) has a 10k and a 1k, and a MPSA18 for Q2 and it is loud as crap.

Man, that sounds pretty good!  I have no idea what the diode does here, but I need to try that out!  I bet if you put a compressor (or maybe a tonebender style gain stage) in front of it it would be out of control.

[John Lyons]

I may have worded it wrong but I meant that the resistance from collector to the output cap was less and therefore louder.

The diode limits the input so the first transistor doesn't saturate,so the text says on Joe Davvisson's site http://www.diystompboxes.com/analogalchemy/index.html

The sample above sounds ok I think but not as good as the sound through the amp. Sounds a little one dimentional on the sample.
The actuall sound is a little too compressed and not as "open" as I'd like it to be. Having a look over at Fuzz Central at the Axis Fuzz makes me want to try that one. I may try some lower gain Transistors at any rate...the 2n5089s are something around 500 HFE !!!!!

John

[tcobretti]

NOTE: self promotion below.

You should build my Socket Face (the layout is in my gallery).  The idea is that you socket every component on the layout; it's a FF breadboard.  You could build any fairly traditional FF variation on it, and the only part of the antiquity fuzz that wouldn't be easily swappable is the 100k to ground after the diode.  You could build the Axis Face (except for the .033u cap across the 1k2 resistor), then try my Mondo Fuzz, then try the Roger Mayer Classic Fuzz.  I spent hours and hours plugging in components after I built mine.  The only problem is that it is a pretty tight layout;  if you wanna get to the collector resistor for Q1 you have to move some stuff out of the way.

About the low gain transistors in a FF.  With all respect to the guys that are truly smarter than I am about this stuff (I have learned a ton reading Fuzzcentral and obviously RG is the man), I am not sure I buy that small hfe transistors are the key to a Ge-like FF.  I suspect that the value of the feedback resistor and it's affect on the interaction between the transistors is a key component of the classic FF sound;  while 100k may be perfect for the old Ge FF, lower values like 47k-33k (coupled with a fairly small cap on Q1 to tame the highs and control oscillation)  will yield results that are very similar to a Ge FF.  I would imagine that if you built a fairly standard Si FF with 5089s, a 33k (47k if you want more gain) feedback resistor and a 50-100pf cap on Q1 that you would have a very sweet sounding pedal that would be very similar to a Ge FF. 

But of course, there is no exact substitute for the real thing.  Ge trannies just sound so good!  The thing we can't seem to get right is the weird compression or "sag" of a Ge FF.  I'm sure it drive many people crazy, but to me it is great and I can't seem to make it happen with a Si FF.  I believe that mastering the sag may be the final step in a close to a perfect Silicon "Ge FF" clone. 

Does that last sentence make sense?

[John Lyons]

Thanks for the info Travis

I'm going to try a few things out on the breadboard.
The Socket face is a cool idea but when I get something I like on the breadboard I'll just do up a PCB of it and commit to one...maybe socket the input cap or make a cap blend pot or switch.

John

[tcobretti]

Let us know what you come up with; I'm a sucker for a good sounding FF.

[petemoore]

  Socket Face...nice board to have around when if you want to hear FF derivations/offshoots.
   Using 4 pins per cap [because they don't fit right with just 2 pin span, 3 pins in a row would work tho....]
  This is a 16pin socket, and makes a sturdy FF Perf'breadboard...a permanent/tweekable FF testboard. If I take care to bend the component ends just right and take it easy on the socket wipers...or you can hotmeltglue the parts in, otherwise I just keep this board handy out of a box for doing FF testwork, I just put it in the testjig, and clip the three wires on.
   iCap            oCap
   Cap            Cap
   -------         ------
   bCap             oCap
   *Cap            cCap
   cCollector      cCollector
   *Base            Base
   eEmitter        Emitter/
    *'100k'>     {Q2E, 100k>Gain Circtuit, pot and cap].
  I used a trimpot of course for Q2C [actually a 4k7 stop resistor and a 10k pot =s 4k7 - 14k7 adjustable], longlegged 47k [or/and trimresistor piggybacked parallel, I leave this trim R tacked on one side so I can just bend slightly and make it touch to reduce Q1C resistor to ~33k]...
  Anyway I'd been using inline sockets and decided to put the whole thing on one many-pin socket