The Pedal Formerly Known as Chela

Chuck D. Bones

Circuit Wizard
I built a version of the Chela (EQD Talons) and posted a build report. This post describes the mods in detail. The Chela schematic is available in the Build Docs, so I won't post it here. This is the schematic of my build. I kept the ref designators the same, as much as possible.

Chela mod v2.2 schematic.png

I added a CLIP switch (S2) so I could select between the original symmetric LED clipping, no clipping and asymmetric MOSFET clipping. The MOSFET clipping is about the same volume as the LED clipping, but it generates even and odd harmonics. The LED clipping generates only odd harmonics. At high GAIN settings, the two sound pretty much the same, but at lower GAIN settings, the MOSFET clipping is smoother and thicker, not as harsh as the LEDs. The "no clipping" option allows clean boost or opamp clipping, depending on the GAIN setting. I added a VOICE control as an experiment to see how it affected the tone. The VOICE control does three things:
1) It alters the range of the GAIN control.
2) It changes the bass response, in conjunction with C4.
3) It limits the current in the clipping diodes or MOSFET, affecting the harmonic content.
All three functions of the VOICE control make it very useful in tailoring the tone of this pedal. I ended up mounting the VOICE pot where the PRESENCE pot used to go.
The Talons' PRESENCE control would be more aptly named ABSENCE because it weakens the TREBLE, MID and BASS controls and does not do enough to trim off the top end. There is a need for additional high freq cut beyond what the TREBLE control can do, but the way EQD did it was not good. I jumpered R6 and the PRESENCE pot and removed C5. I initially moved the PRESENCE pot to just before the VOLUME control. I found that I only used a couple of settings, so I replaced the PRESENCE pot with a switch (S3). In the center position, there is no additional treble cut. The up and down positions provide medium and heavy treble cut, respectively. The TREBLE control's range provides overlap in the three PRESENCE switch settings. If you don't do any other mod to the Chela, you should really do the PRESENCE mod because getting rid of C5, R6 and the PRESENCE pot between IC1.1 and the tone controls makes a huge improvement in the operation of the tone controls.
I retuned the tone controls a little bit, moving the MID and TREBLE controls down in frequency about 1/2 octave and increasing the BASS pot's boost and cut range. C6 was increased to allow more bass thru. While we're on the topic of the tone controls... the MID control overlaps the BASS & TREBLE controls quite a bit. With this circuit topology, there is really no way around it. What it comes down to is the the TREBLE and BASS controls work best when MID is at noon. When MID is near the ends of its range, it overwhelms the BASS & TREBLE controls and they lose most of their adjustment range.
Lastly, I added the Charge Pump board so that I could run the board on +18V. The Chela needs the additional headroom because advancing the MID pot beyond 2:30 causes the last stage to saturate. D100 is no longer needed since the Charge Pump board provides reverse polarity protection. I replaced D100 with a 100Ω resistor for some additional power filtering. I increased R101, R102 and C101 to get more filtering on Vref. R5 now returns to ground, thru the VOICE pot, instead of to Vref. With +18V power to the board, this pedal is capable of delivering over 13Vp-p at full-tilt, so overdriving a tube amp is no problem.
C4, R5, D1, D2, Q1, C5 and C13 are mounted on a Vero board that is supported by S2 & S3. C4 and R5 could have been left on the board, but the wiring to the VOICE pot is cleaner with them mounted outboard.
I touched a few other component values; you can compare my schematic against the Chela BOM to see which parts I changed.
 
Last edited:
Super interesting, thanks for the extensive explanation and schematic! And I might have to visit that circuit now me thinks...
 
Awesome. I will definitely check these out if I ever build one.

I haven't built a Talons circuit, but I've always been curious about one thing. R5 and C4 (1k and 1u) set the HPF corner frequency at 159Hz (right?), whereas most op-amp based OD/DIST pedals are set much higher. Any comments to why the Talons might be designed like that, or what the effect is? Is the dirt heavier in the bass than most pedals? Is it compensated in the later tone stack? Did you experiment with lowering C4 or R5 to have higher cutoff frequency with the voice knob maxed?

Also just checking I'm understanding right, the asymmetry of the mosfet clipping is pretty aggressive, right? (0.7v one way vs 2v the other way, according to amz article). When designing for asymmetrical clipping, what does the ratio of the thresholds tell us about the even/odd harmonic content?
 
Good questions!
I did not experiment with the R5-C4 time constant. For me, this pedal shaved off enough bass with VOICE max'ed. I already have a Chinese TS clone and a Frost Drive, so I didn't need another pedal with the same freq response as a TS. Or a Rat/Expandora for that matter. But you do make a good point: C4 could be reduced and the bass attenuation dialed up or down with the VOICE control. As you know, the tone shaping before the distortion has a major effect on tone. More bass = more sludge. Chords get more dense. Reduce the bass going in and chords get clearer, you can hear individual strings. Tone shaping after the distortion is different. You basically use the TREBLE pot to control the harmonic content and the BASS pot to recover some of the bass that was attenuated before the distortion.

I could get into a very long discussion on asymmetric clipping. I've done a fair amount of simulation and breadboarding and I've come to this conclusion: It's very easy to make asymmetric clipping that works well using discrete transistors and FETs; it's difficult to make asymmetric clipping that works well using opamps. By "work well," I mean have a transfer function that produces more 2nd harmonic than 3rd, and so on.
This is the short explanation...
Here's how the single MOSFET clipping works: at low levels, neither the body diode nor the MOSFET transistor conducts and the signal passes undistorted. As the signal level increases, only the body diode begins to conduct, so the waveform is flattened on one side only. This produces a range of even and odd harmonics, with the lower harmonics being more prominent. This has a rich, natural sound, reminiscent of a clarinet or violin. As the signal level increases further, the MOSFET transistor starts conducting and the waveform is flattened (clipped) on both top and bottom, but more on one side than the other. Now the even harmonics start diminishing while the odd harmonics are increasing. It's sounding more like an amplifier beginning to break up. At even higher signal levels, the clipping on the top and bottom is very similar and as a result, the even harmonics fade out. This is hard overdrive territory. At this point, it sounds the same as symmetric clipping, even though the amplitude on one side of the waveform is clipped around 0.7V and the other side is clipped in the 2.0V to 2.4V range. The advantage of a large difference is clipping levels on the top and bottom is it makes the region larger where clipping happens on one side only. When I play hard with the VOICE and GAIN maxed, the symmetric and asymmetric settings sound about the same. As the note fades out, then you hear the difference.

I hope that answers your questions, or at least provides some clues as to what's going on.
 
Perfect, thanks, I hadn’t heard it explained that way. I think I had previously written off the single mosfet clipping, thinking the large difference in threshold was a negative rather than a positive, but I’ll have to try it out on a build.
 
Just a little nitpicking and correction here... following through the schematic above the power pins on TL072 seem mixed up, 8 needs to go to VCC and 4 needs to GND.
 
Just a little nitpicking and correction here... following through the schematic above the power pins on TL072 seem mixed up, 8 needs to go to VCC and 4 needs to GND.
Oh snap! You are right! I just corrected the schematic. Good catch, Cooder. I should also point out that the schematic does not show the (optional) charge Pump Board and that the CLIP switch positions are as viewed from the front panel.
 
I'd like to do the mid pot mod to my tone job. I'd connect a 7.5k resistor from the middle lug of the mid pot to where exactly? Sorry, I need it dumbed down!
 
Cut the middle lead short on the MID pot. Solder the outer pins to the board normally. Install the 7.5K resistor from the stubby middle pin on the pot to the pad on the board where the middle pin would go.
 
I was looking over this one again and was trying to figure out the function of R15 1k. I've seen that before, but most of the schematics I've seen for these Baxandall op amp EQs run straight to 4.5v bias from the positive pin. Is this putting a constant load on the pin in case of the voltage fluctuating?
 
Just built this with your "Absence" mod and I'm getting no distortion in the signal. All I have is a clean signal with a functioning EQ. I jumped R6 and PRES, and omitted C5. Maybe I missed something else? It was a lot of info to take in, but I probably just have a separate issue on my hands.
 
Back
Top