Design Challenge #2

Chuck D. Bones

Circuit Wizard
Boss has built millions of pedals using the CD4007 as a flip-flop in the engage-bypass switching. The challenge here is to build a pedal using a CD4007 IC in the audio path.
Hint: Catalinbread has already done it, so take a look at their discontinued pedals for clues.

Here's my 2nd version. 1st version had too much gain (yeah, like that's even possible). 4 knobs (L-R): Bass, Volume, Treble, Gain. Remarkably quiet for a CMOS-based circuit. No NOS parts here either. Still tweaking, but it's close. It's not a copy of a Catalinbread pedal.

CMOS distortion breadboard 02.jpg
 
Last edited:
The three I know of are the Hyperpak, the V8 Fuzz Engine (a higher-gain version of the Hyperpak) and the Supercharged Overdrive (SCOD for short).

Here's what I have so far:
CMOS overdrive rev 1.4.png
This thing is crazy loud and has a shit-ton of sustain. Unity VOLUME is around 9:00. Even with GAIN at zero there is plenty of dirt. Crank the GAIN for some sweet feedback, even at bedroom volume. Very quiet even with GAIN dimed. The JFET preamp is similar to the 1st stage in the Fuchsia. The GAIN control is similar to the MI Audio Crunch Box and Marshall Blues Breaker. We don't say "plagiarize," we say "research." The BASS control ranges from tight to roar. TREBLE goes from dark to sizzle. I didn't spend too much time fiddling component values. The DMM in the top pic is reading the voltage on the CD4007's power rail. Total current draw is under 2mA.

For the next iteration, I'm moving the GAIN control to the 2nd & 3rd stages and the TREBLE control between the 3rd and 4th stages.
 
Last edited:
that is good to read that you are not getting a lot of noise from the 4007. and no buffer for the output seems to make sense, even though I've read that some of the catalinbread circuits used buffers on each side of that chip.
 
Still messing with this one. I'm on the 6th or 7th revision. Tried a bunch of stuff that didn't sound good: building a Tweed Sound, Box o' Rocks or a WIIO using the 4007 in place of the FETs. Tried the ThunderZone tone control arrangement. Rev 1.3 above sounded the best so far. The latest rev is pretty close to rev 1.3, but has a variable mid boost on the 2nd stage using a gyrator. There are some component value changes as well.

CMOS overdrive rev 6.1.png
 
Nice to see these possibilities compared to the fuzz circuits using the 4049 chips.
The two ICs are pretty similar. For me, the advantages of the CD4007 are
1) I can install current-limiting resistors between the upper and lower FETs on 2 of the 3 stages.
2) I can use just the upper or lower FET on some of the stages.
3) I'm not tempted to use all 6 inverters in the 4049. The simplicity of the 4007 enforces a bit more discipline.
 
Are there many CMOS overdrives out there? Are CMOS chips good for other kinds of effects? I have a few different types of CMOS ICs in my collection that I'm planning on playing around with on some Beavis projects (I'm interested in DIY synth stuff) and I had been wondering if they could have other uses as well.

Also, are these projects you work on here going to be available once you get them to a place you like?
 
There are a few. Runoffgroove has a dual distortion called Double-D based on the 4049. I built a modd'ed version. The problem with using CMOS in an analog application is there are no guarantees how they will perform. The 4049's made these days are pretty noisy. I have a handful of NOS 4049s from Fairchild and RCA that are quiet. Fortunately, the TI 4007's I procured recently are quiet. TI could change the recipe any time and then they might not be so quiet.

4013 flip-flops get used as frequency dividers in octave-down pedals. You can certainly build synth modules using CMOS to make freq dividers, freq multipliers, sequencers, oscillators, etc. Some CMOS parts, like the 4046, contain a VCO, so it's actually part digital, part analog.

I breadboarded a version of the Maestro FSH-1 and I used a digital pseudo-random sequence generator made from CMOS gates & shift registers in place of the noisy transistor / sample & hold circuit. A pesudo-random generator appears truly random until the pattern (eventually) repeats. The one I built is 33-bits long and will run for about 27 years at full-speed before it repeats.

Now and then, I come up with something that someone, PedalPCB or one of the guys on the forums, finds good enough to merit a board layout.
One of these days I'll take the time to learn DipTrace.
 
Last edited:
Stick a fork in it, it's done.

CMOS overdrive rev 6.2.png

I liked having the MID control, but I wanted it to go boost or cut. Now it does. All of the controls are usable across their entire range. Not too many dirt pedals can claim that. There is no clean setting unless you back the guitar volume down. Other JFETs will work, just tweak R3 to get Q1's drain around 5V. The MID control is tuned to about 600Hz. The MID freq can be tweaked by changing C5 or C6. Increasing C5 lowers the freq and makes the mid hump broader and the dip deeper. Increasing C6 lowers the freq and makes the mid hump narrower and the dip shallower. Decreasing the caps has the opposite effect. If you don't want the MID control, then build v1.3. As usual, all component values are negotiable. The TREBLE control is tuned to a "Goldilocks" range of not too dark, not too bright. R13 and R17 are pretty well optimized for the most headroom. This thing gets quite loud. Total current draw is just over 3mA.
 
Last edited:
Hi Chuck, a couple questions about Q1, the PF5102:
  • Will other JFETs work here? In particular, a 2n5457, since I already have some of those on hand?
  • I read your post on JFETs. Previously I've always used JFETs in a paint-by-numbers manner (not knowing any better). For Q1, does the JFET need to be tested for some particular parameters?

I whipped up a PCB for the circuit in post #15 above, waiting on delivery. Now working on the BOM.

Thanks!
 
  • Like
Reactions: fig
Other JFETs will work. My PF5102s have Vp around -1.2V. The Vp spec range for 2N5457 is -0.5V to -6.0V. If yours land near -1.2V, you can use them with no circuit mods. Otherwise, you can adjust R4 to accommodate different Vp within the range -0.5V to -2.0V. Use this formula and round up to the closest common value (use a positive number for Vp in this calc).

R4 = 4.7K * Vp / 1.2.

Example: if you measure 2.0V for Vp, then use 8.2K for R4.
 
  • Like
Reactions: fig
Back
Top