Unwanted gating/note decay on Colorsound 1 knob fuzz build

Started by Widows, July 10, 2013, 11:53:52 AM

Previous topic - Next topic

Widows

Hi folks, I've recently built a coloursound 1 knob fuzz as per the layout here: http://tagboardeffects.blogspot.co.uk/2013/04/1-knob-fuzz-bonanza.html It's a great pedal but I'm having a couple of issues with it.

1) I'm getting some unwanted gating/note decay creeping in especially on the higher strings and even more so if I'm just playing a single note.
Basically, the note will sustain for a while then I hear a weird crackle and a drop in volume as though a transistor is shutting off. I had a similar problem when building a Bazz Fuss using a single transistor instead of the prescribed Darlington style set up, and this issue sounds very much the same as the problem with the Bazz Fuss.

Does anyone know what I can do to go about increasing the sustain and eliminating this annoying decay problem? I want to be able to get as much sustain as possible as this pedal is going to be used to play a lot of slow, doom-y stuff that makes a lot of use of long notes etc, and to have them decaying early will kind of defeat the point. If I can beat this issue I'll be boxing it up and selling it but I can't really put it up for grabs until I've sorted it....

2) The effected sound seems to be very heavy on distortion in the low end of the freq range, but almost non-existant above the mid/high mid range.
Is there any way I can increase the distortion on the higher freqs, or simply shift the freq band that is distorted so that it covers less of the low end and covers more of the high end?
Could it be that all I need to do is reduce the value of the input cap to stop the bass freqs swallowing the high end quite so much?

Thanks
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

GibsonGM

I went and found the actual schematic (necessary to see what's going on in the circuit)....

Posting what voltages you have on your transistors might help know which one is most unbiased/what's going on...maybe audio probing the collector and base of the BC108 would, too (seeing where the 'misbiased sound' starts, but you may not hear any difference anyway). 

No way you got the transistors switched, is there??    You should have about 4.5 volts on the collector of the BC108.  If it's truly misbiased, you'll be off, and might have to put a trimpot on the emitter. The '109 should be close, too.

2) See if the above changes things (it might)....once you get close to something you like, tinkering with the input cap would be the next thing, IMO.   You could try a .033 uF or something....or even .01....to taste. 
  • SUPPORTER
MXR Dist +, TS9/808, Easyvibe, Big Muff Pi, Blues Breaker, Guv'nor.  MOSFace, MOS Boost,  BJT boosts - LPB-2, buffers, Phuncgnosis, FF, Orange Sunshine & others, Bazz Fuss, Tonemender, Little Gem, Orange Squeezer, Ruby Tuby, filters, octaves, trems...

thehallofshields

Thanks for posting that link; I got interested in this circuit after hearing a demo of the Ritual Fuzz, and I made my own comparison chart. This one is much nicer. I've had this circuit on breadboard for the past couple weeks, and though I'm no expert, I've got some notes to compare.

1. I couldn't believe how much the gain of the BC108's I got off ebay swung. - I had some as low as 38hfe and some as high as 470hfe. My goal with this circuit  was to play with mis-biased sounds, so I've noticed that the gain of the transistors really effects the gating and decay. Maybe you could plug your transistors into a multimeter and test the gain.

2. Also, I noticed a high-gain BC109 in Q1 can really saturate too much bass in the circuit. I had to turn my guitar volume down to keep it from being a total mess. To compensate, I've added a trimpot between Q1's emitter and ground which I keep at 500ohm to dampen the gain.

3. One thing that really attracted me to this pedal was that the treble is voiced perfectly for me; I don't want anymore or any less, so instead I've worked on thinning out the bass. I've tried blending between input caps in this circuit. .68uf on one end for the Ritual Fuzz sound, but I had to go as low as .01 to get a significantly leaner sound. That combined with dampening the input volume gave me overdrive like dynamics. Unfortunately, removing the low-end saturation kind of killed the sustain for me, so I'm back to the drawing board on that subject.

Widows

Cheers for the info guys.

I wound up swapping them out for various combinations of 3904, 5088 and 2222 in the Q1 and Q1 slots and still got the same problems. I then put em in my DMM and check their HFE and tried to use ones that had similar values rather than wildly different ones, but still the same.

Not had chance to take any voltage readings yet, when I do I will put em up here.

Thanks again
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

Widows

Ok this is the first time I have done any measurements of transistor voltages so I may have got something wrong here:

First up I set the circuit up with no power into it, set the DMM to DC voltage and placed the DMM probes across the B - E legs of the transistor to the left of the diagram, then the B - C legs, and repeated the process for the other transistor. The results are as follows:

Left transistor:
B- E = 0.6V
B - C = 0.93V

Right transistor:
B - E = 0.6V
B - C = 4.30V

I then did this again with the power connected and the guitar hooked up. I gave it a strum and took the readings. They were virtually the same however they fluctuated and slowly returned to the levels listed above as the notes died out.

^^If I've completely got this wrong, please advise ;-)

It's worth noting that I tried this circuit with a bunch of different transistors n had the same decay issues across the board which makes me think that not enough voltage is getting to one of the transistors and it's closing before it should. Perhaps I've got a level controlling resistor value too high somewhere and it's choking the signal into one or both of the transistors. The transistors in the circuit at present are 2N2222's in both slots, these particular ones have HFE values of:
Left: 284
Right: 275

Both these values flutcuated a little when in the hfe reader on my DMM but they sat around the values given

Cheers folks
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

duck_arse

I still can't see the circuit, but from the layout it looks like a fuzzface circuit. if you search for "silliface" and silliface 2, you'll find lots of stuff regarding tuning a silicon fuzz face. there is the definitive page over at geofex, will tell you a lot of stuff, how they work, how to tune, mods, etc.

to take voltage measures, you need the supply connected. we need to know the supply vltage for anything to make sense, and when you take the measures, don't input a signal. the dc measures tell most about what is right and wrong around the circuit.

and find us a link to, or post the circuit you built .....
" I will say no more "

Widows

Cheers Duckarse. So I connect power but not guitar and take the readings as before?

Supply voltage is 9V from a Voodoo Labs Pedal Power 2+

The circuit is built using the layout in this link: http://tagboardeffects.blogspot.co.uk/2013/04/1-knob-fuzz-bonanza.html

I can include some pics of the one i built as well if that's what you're after :-)
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

LucifersTrip

hey...

firstly, the voltages are taken with one lug on ground and the other on the point you want to measure. see the debug thread:
http://www.diystompboxes.com/smfforum/index.php?topic=29816.0

secondly, Q2C voltage should be around 7V, not 4.5V like a Fuzz Face

http://www.diystompboxes.com/smfforum/index.php?topic=51351.5;wap2

This is one of my favorite Fuzz Face variants, but I made mine germanium
http://www.luciferstrip.com/fuzz/colorsound-oneknob-schematic.jpg


good luck
always think outside the box

Widows

Awesome,  thanks mate. I'll get on that today and get back to you with some proper results haha. I literally googled "testing transistor voltages" and took the info from the first couple of pages in the results. Turns out these were pages to test if it was faulty before putting into the circuit, derrrrr. Must remember not to get into debugging when I'm tired ;-)
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

LucifersTrip

...and don't forget, the 820 in your layout is 82 in an original



upping it from 82 to 820 will change the nature of the sound in addition to making it louder
always think outside the box

Widows

Ahhhh, legend. Cheers for that man. Just waiting on a spare half hour to get round to it, gotta cook dinner now. Damn life, getting in the way n shiz ;)
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

nocentelli

Quote from: LucifersTrip link=topic=103452.msg930383#msg9303

upping it from 82 to 820 will change the nature of the sound in addition to making it louder

I think replaced that resistor with a 47r and a 1k pot in series in one build I did, acts as a useful "character" pot, or you can dial it to just how you like it and replace with a fixed resistor of that value.
Quote from: kayceesqueeze on the back and never open it up again

duck_arse

windows, back to your original question 2, you can reduce the value of the (I couldn't get anything to load at tagboards, so I'll use the circuit lucifer posted) 100nF/0.1uF input cap to reduce the amount of bass getting into the circuit, but the bass tends to knda carry the sustain. you couls search the "easyface" which puts a blend pot between 2 input caps, but it would then be a 2 knobber.

if you reduce the emitter bypass cap, the 10uF across the 1k, you can reduce the gain applied to the lower frequencies, producing less mud in the output. you can combine the input and emitter bypass caps to get a v. nasely wahish response, if that's what you like. you can also reduce the 220nF/0.22uF output cap to reduce the amount of bass leaving, but it is less effective doing it here, imho.
" I will say no more "

Widows

Duckarse: Yeah dropping the value of the 10uF cap across the 1K in Lucifers diagram to kill some of the gain to the lower freqs is basically what I'm looking for.

I got round to doing the voltage tests on the transistors just now so here's what I got:
DMM set to 20V DC scale
Circuit connected to 9V power block (Voodoo Labs Pedal Power 2+)
No guitar leads or instrument connected to the circuit

Left transistor:
C: 3.14V
B: 0.56V
E: 0V <--surely this ain't right?

Right transistor:
C: 2.47V
B: 3.12V
E: 2.46V

This is the layout I'm using, it's the one on the top left:

Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

duck_arse

the emitter of q1 is connnected to ground, 0V. when you do the measures, you have the black lead where? and when the red lead is connected to E, what happens?

your q2 C voltage looks like it might be a short to E. check around that transistor, or post us photos of your work. it's always more interesting and informative.
" I will say no more "

nocentelli

Quote from: kayceesqueeze on the back and never open it up again

Widows

I have the black lead of my DMM connected to the ground wire coming off the ground rail (second row in the layout diagram).
When connecting the red lead to E (with black lead at ground) there's a reading of 0 on the screen

Here's some snaps of the project as I've been building it











What's wrong with Q2? Anyone got any voltages from a working model that I can reference it against?

Cheers
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

duck_arse

windows, first let me say your soldering looks ok. second let me say, I don't wanna sound rude. third, let me say, "but". you seem to be working at cross purposes. I can see you have a big lump of breadboard there, that is where you get your circuit working FIRST. then, it is where you do the majority of your value fiddleing to find what does what, or if you don't like the circuit at all. then when you've sorted yr values, you can plough ahead with the soldering. having a board full of soldered sockets is the same as having a breadboard.

and you said you were using bc10x transistors? they have the opposite pinout to the 2nxxxx series, which I'd be betting is what mr. balls layed-out for. go back to breadboard, check online for a datasheet for the transistors you have, and start from there. let us know what you get on the breadboard.

can I call you windows? I always get names mis-read.
" I will say no more "

Widows

hahah yeah I noticed you'd added an extra 'n' in there ;-)

As for the pinouts, I always look up the datasheet before putting trasistors anywhere near a board however I'll double check this as I may have done something daft. That said, the circuit works fine...it's just this annoying gating effect on the notes when they begin to fade out. Could it be a question of hfe? I was just reading a build report on another site that for the same circuit where the person building it had the same problem. He was using a lower hfe in Q1 and a higher hfe in Q2 (as is apparently the norm), however upon switching them around so the higher hfe was in Q1 and the lower in Q2 it seemed to remove the problem and the notes decayed much more naturally. Up to this point I had been using transistors with as close to matching hfe values as possible, is there supposed to be a difference between them? If so, I wasn't aware of this :-/
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

LucifersTrip

#19
the first thing you want to do is sort out Q2 voltages:

C: 2.47V
B: 3.12V
E: 2.46V

either you're taking readings incorrectly or something is definitely off. yes, C should be close to 7V, but the bigger problem is that B is greater than C  and E is too large. as mentioned earlier, C & E are almost identical (or are), so they are pretty much tied together.

for instance, here's sample voltages from a si Fuzz Face

Q2
E) .8v
B) 1.4v
C) 4.5v

of course, for your One Knob, all three will be different, but still E will be less than B will be less than C.

edit:
http://www.geofex.com/fxdebug/bias_prob.htm
(Q3 fail)





always think outside the box