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

Interesting article, thanks for the link :-)

I went and double checked a few things; mother nature is no liar but she's a stickler for details.
- Pinouts: Correct for the 2222s in there at the time.

- Electrolytic cap orientation: Correct.

- High hfe in Q1 and lower hfe in Q2: I tried a higher gain 2N5088 with an hfe of around 500 in Q1, and a 2N2222 with a hfe of about 280 in Q2. This seemed to improve the gating problem a bit but some notes did still seem to die away unaturally. I tried swapping this combination but it sounded like shite so I swapped it back.

- Resistor values: As per the schematic you posted earlier I removed the 830R resistor (this was the nearest value to 820R I could cobble toegether out of what I had - 360R and 470R in series) and
  replaced it with a 100R, as I didn't have any 82Rs. It did seem to make a bit of a difference but I was keeping the volume down as the neighbours were getting sore ;) I will check out how it sounds
  properly tomorrow and report back.


May also have a go at including a high pass filter to reduce the low freqs hitting the clipping circuit, or just reduce the input cap to cut off some of the bass altogether....once I have this gating problem licked though.

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

Widows

I've swapped the 100n input cap for a 47n which has cut some of the mud out.
Also tried a 22n but wasn't too keen on the sound.
Might revert back to the 100n and add a high pass filter set at around 300Hz (ish) to stop the mud hitting the clipping circuit instead, to see how that affects things.

I've had a play around with some of the transistors, and using a slightly higher hfe in Q1 than Q2 (a few tens higher rather than a few hundred), it seemed to improve the situation, and the gating isn't quite so harsh, although it does still crop up right at the end of the note....and as before, it seems to creep in sooner on the higher notes.

Currently Q1 is a 2N5088 with an hfe of 530, and Q2 is a BC109 with an hfe of 465.
Should I be looking for a close match, or a mismatch (and if so, how mismatched) between the hfe of the two transistors?

Transistor voltages of current vs previous configurations are as follows:

Current Transistors       Previous Transistor Setup
Q1 (2N5088)               Q1 (2N2222)
C: 1.45V                     C: 3.14V
B: 0.6V                       B: 0.56V
E: 0V                          E: 0V

Q2 (BC109)                 Q2 (2N2222)
C: 7.26V                     C: 2.47V
B: 1.98V                     B: 1.98V
E: 0.83V                     E: 0.83V

Measurements taken using DMM set to DC Volts scale, black probe connected to ground lead coming off the board, and red probe connected to one leg of each transistor at a time.As you can see, there's quite a difference at the collector of Q1, and across all legs on Q2.

Can anyone shed any more light on these values? At my current level of knowledge, these are just numbers to me, I don't really understand their significance.

Up til now I've only really been testing it on a tiny solid state amp at home, am off to go and give it a blast on a proper valve stack this afternoon so I will see what those changes do to the sound at proper volume ;-)

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

duck_arse

looks like your Q2 is better. the usual way to tune a fuzzface is with a pot where the 2k2 is. try a 10k to 50k trimpot in there, and see if you can then tune the Q2 collector voltage. if you listen (w/ guit input) while you tune, you'll go from nasty, harsh, bright, square waves, to great, looping, big mushy fuzz, and then to a harder sound again. that is, if it biases correctly.

I had a thought: measure the voltage at both ends of the input cap (100nF), see what is.
I feel sick.

Widows

Done some more fiddling with this.

Swapped a few things so I'll list what changes I've made fromt the original:

- Q1 2N5088, Q2 BC109
- 820R switched for 100R (didn't have any 82Rs)
- 150K switched for 120K (nearest whole value I had)
- 2.2K switched for a 50K lin pot
- 100n input cap swapped for 47n (to reduce bass)

The pot replacing the 2.2K allows control of the biasing, I've had a fiddle and get some much smoother decays now, however I need someone to hold the DMM probes to the bias pot and take readings whilst I play as I can't do both at the same time sadly. I'll have another try once I've accquired someone to help me get some readings when I get it sounding good.

Duck Arse: How should I measure the voltages of the input cap, same way as with the transistor voltages?

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

duck_arse

if you need two people, you doing it wrong!

measure/set your bias voltage at the collector of q2 to about 4.5 volts without any input signal. the exact value is not important. then you can plug in, and give a thrash, LISTEN, tweak a little, see if the sound improves/changes. I dunno about the sillicon version, but the Ge version has a very well defined sweet spot, and goes flatter and squarer/nasty either side.

as for the input cap, again with no input, measure the voltage where the input "tip" wire is soldered. black lead to ground, red lead to the cap, meter on lowest dc volts setting. I was interested to see if you have a dc voltage there. there should be 0.00V, or very close.
I feel sick.

Widows

haha ok, yeah looks like I got muddled up there again ;-)

cheers for straightening me out, I'll give your advice a try in the morning
Gibson SG > Dunlop Cry Baby > Sovtek Big Muff Pi (black) > Digitech Harmony Man > Matamp GT1 > Matamp 4x12 w/Celestion K100s

Widows

Righty, I used the extra pot i wired in place of the 2k2 resistor to tune the voltage on the collector of Q2 to around 4.5V, then tweaked the pot a little bit whilst playing to see how it affected the sound.

I managed to get a range of interesting sounds like you mentioned by dialling the Q2 voltage between fractions of a volt up to 9V. Unfortunately the only suitable pot i had was a 50K pot and that made it quite hard to make small incremental changes to the voltage.

At that point I had an idea and set the Q2 voltage to 4.5V again and measured the resistance across the pot which was around 500 ohms. The closest resistor values I had were 470 and 560 ohms, so I tried both in place of the 50K pot. 470 sounded a bit naff so i went for 560 which sounded better but still not quite right. I swapped it for a 680 which sounded pretty good, lots of point in the high mids. Just for completeness I tried a 1K as well but it went back to sounding tinny and splatty again, so I stuck with the 680 as it resulted in the best sound out of the various values I tried. It's still a little short of sustain though.

I really need to get this back on my stack and have a test with it now i've made these changes n see how it sounds on something that isn't a 10W solid state amp that's had a hard life ;-)

Will report back with what I get.

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