Buddy's Breadboard and Circuit Design Notes

[Chuck D. Bones]

If you want to maximize gain, then bypass the source resistor. I usually pick a drain resistor and then dial-in the source resistor to get the desired drain voltage. For maximum headroom, the drain voltage should be 1/2-way between the source voltage & Vcc. It's counter-intuitive, but for a given Vcc, FETs will produce more gain with lower drain current. There are diminishing returns so I wouldn't run the drain current much below 100μA.

Here's what I did with Sandy's example circuit. The clipping diodes are Germanium. At low FUZZ levels, the clipping is symmetric with just a few odd-order harmonics. At higher FUZZ levels, the clipping becomes increasingly asymmetric because Q1 drives the diodes asymmetrically. Q2 runs clean with LEVEL below noon. At higher LEVEL settings, Q2 adds some even-order harmonics. If you want Q2 to run dirty all of the time, then swap VR3 & R9. Any JFET will work, just tweak the source resistor. As usual, all component values are negotiable.

 

[BuddytheReow]

So, yesterday I put together in the resources section a handful of Devi Ever snippets that she uses in a lot of her fuzz pedals. For Devi, it's usually a mix-and-match type of situation and slap a new name on the circuit. As a due diligence process for myself before posting anything I have to try them out for myself. Here's my Devi Breadboard project for mix and match. The top left is a pre-gain or fuzz section, top right is a volume pot/output section, bottom left is high and low pass control that I cooked up. There are 4 Devi snippets in the middle that I wanted to mix and match to see how everything reacts together, link them to the filters and then output. I'll be sharing my Devi Baby shortly...

 

[BuddytheReow]

Taking a few pages from the Devi Ever playbook I hooked up a few blocks together and really liked it, but didn't love it. If you keep the circuits as is they can get waaaay too bass heavy and muddled depending on how the circuit flows. To counter this I tweaked both a variable high pass and low pass filter for some decent variability in tone. The one circuit block that really caught my eye/ear was the block with a 2n2222a acting as a diode in the feedback loop. It's more of a clipping circuit than anything else, but it also takes out a bunch of muddy bass.

After mixing the circuit blocks with some slight tweaking (I added C2 below) and tone controls I found something I'm actually really happy with. I love this circuit so much through my amp that this will be going on top of my list in an attempt to put it on vero and box it up, though it may need a slight tweak or two to make it shine. For starters I may need to tweak the BASS control since it gets too treble-y when turned CCW, but that's a simple fix.

I'm hesitant to fully call this a fuzz pedal since you can dial back the FUZZ control and it sounds like a pretty good distortion pedal IMO. Max out the FUZZ control and some gnarly fuzz tones are at your fingertips. SO yea, it's a FUZZTORTION pedal.

I call this one the HEAVY DEVI. I hope some of you will try this out on a breadboard.

 

[BuddytheReow]

If you want to maximize gain, then bypass the source resistor. I usually pick a drain resistor and then dial-in the source resistor to get the desired drain voltage. For maximum headroom, the drain voltage should be 1/2-way between the source voltage & Vcc. It's counter-intuitive, but for a given Vcc, FETs will produce more gain with lower drain current. There are diminishing returns so I wouldn't run the drain current much below 100μA.

Here's what I did with Sandy's example circuit. The clipping diodes are Germanium. At low FUZZ levels, the clipping is symmetric with just a few odd-order harmonics. At higher FUZZ levels, the clipping becomes increasingly asymmetric because Q1 drives the diodes asymmetrically. Q2 runs clean with LEVEL below noon. At higher LEVEL settings, Q2 adds some even-order harmonics. If you want Q2 to run dirty all of the time, then swap VR3 & R9. Any JFET will work, just tweak the source resistor. As usual, all component values are negotiable.

View attachment 27999

Chuck, why do we need R10 in your schematic since it's so low?

 

[Chuck D. Bones]

Chuck, why do we need R10 in your schematic since it's so low?

We don't absolutely need it, but it provides some filtering of the external power and ensures that the current step from switching the LED on & off will not couple into the audio signal. I put R10 in because I could. The current demand is low and the voltage drop in R10 is about 300mV. The circuit works fine with no resistor there.

 

[BuddytheReow]

Tweaked the low pass filer a touch. THis is much better when turned fully CCW.

 

[Chuck D. Bones]

Maybe try 4.7K for R5 so you don't lose so much volume when BASS is dialed back. Does R4 do anything?

 

[BuddytheReow]

Sigh...R4 does nothing. The 2n2222a does all the biasing. I'll keep it in there for the true Devi spirit.

Changing R5 starts some weird oscillation. I'm keeping it at 10k.

I added a 10k in series with R6 to the pot and the volume stays when dialed back. I'll update the schematic tomorrow

 

[Chuck D. Bones]

That oscillation should not happen. It is an indication of an underlying problem. Do you have a 100uF power supply bypass cap?

 

[BuddytheReow]

That oscillation should not happen. It is an indication of an underlying problem. Do you have a 100uF power supply bypass cap?

I do. Should I increase or decrease the capacitance?

 

[Chuck D. Bones]

Neither. I don't see how power & ground are connected between the top & bottom proto-boards.

 

[BuddytheReow]

Neither. I don't see how power & ground are connected between the top & bottom proto-boards.

The two boards are not connected. I probably should try that with the filter cap thrown in there. I remember having an issue with my boards a few months ago and grounding everything to everything else pretty much fixed the issue. Not sure why this is creeping back up now.

 

[Chuck D. Bones]

Are you carrying signal from one board to the other? Hard to follow what's going on, your breadboards are starting to resemble mine.

 

[BuddytheReow]

I'll cleanup my boards of the unused circuit blocks and share them here. The short answer is yes.

 

[Chuck D. Bones]

If you carry signal from one board to another, then you need to also carry power & ground from one board to another. Those power and ground connections need to be short & sweet. Doesn't hurt to put 100uF power supply bypass caps on each pair of rails.

 

[BuddytheReow]

I only had a short amount of time tonight. Wasn’t sure whether to paint a pedal or clean up my board. I chose both .

Now the sound is very different and the treble control doesn’t work. I’ll troubleshoot it in the morning. The oscillation is gone.

 

[Chuck D. Bones]

One step forward, two steps back...

 

[BuddytheReow]

Fixed. I had the second stage going from base to the output cap instead of collector to out. The treble pot wasn’t working but I just wired it incorrectly.

In terms of the oscillation it’s gone, but a long-standing problem with by board has come back. For some reason I have to power each rail separately (ground and +). I’m thinking this may be a problem with the board itself and not the circuit?

 

[Chuck D. Bones]

All power rails (of the same voltage) used to power a part of the circuit must be connected together and to the source. Same goes for ground.

Here's an example. I breadboarded a modd'ed Klon. All blue rails are ground and they're all jumpered to each other. Top red rail is -9V. Next red rail down in +18V on the left half and Vref on the right half. Next red rail down is +9V. Bottom red rail is unused. The power jack connects to the +9V rail and one of the ground rails. The charge pump circuit connects to the -9V, +9V and +18V rails.

 

[BuddytheReow]

Here's the updated schematic. I'm going to make an attempt to put this on vero. I fooled around with DIY Layout Creator on my lunch break today and need to fire up the iron. I'll circle back here once my layout is confirmed.