The Double-D

Chuck D. Bones

Circuit Wizard
This is a 4049-based distortion, inspired by the runoffgroove Double-D. For those not familiar, the CD4049 is a CMOS hex inverter that is normally used for digital applications. Long ago some clever analog engineers figured out that this particular CMOS device (and a couple of others) can be hijacked for analog purposes. I built my first 4049-based distortion back when... well let's just say that Keith Moon was still alive. The 4049 makes a distortion sound that is unlike bipolar transistors, diodes or opamps and can be very tube-ish. These days there are quite a few boutique pedals that use the 4049, including the Bloody Finger (Chrome Dome). To make a long story even longer, I was fishing around the 'net for 4049-based distortions and came across the ROG Double-D. The 4049 contains six inverters (amplifiers) and it's a challenge to see how many can be used in one pedal before the gain and noise get out of hand. This one uses five to the six inverters, arranged as two separate distortion channels. The Jiggle channel uses two inverters, has moderate gain and a tight bottom-end. Clean boost it attainable at lower gain settings. The Bounce channel uses three inverters and is designed for high gain and maximum sustain. The ROG Double-D has four knobs: two for GAIN and two for VOLUME. Pretty simple. Maybe too simple. So I threw in four more knobs: a TONE control for each channel, a PREGAIN control for the shared JFET preamp and a BIAS control to throttle the power to the 4049. At high BIAS, the 4049 gets almost the full 9V, yielding tons of headroom and 6Vp-p max output. At low BIAS, the gain drops, the distortion is more compressed and a bit gated. Less volume is available, only 3Vp-p :p. In the middle, there is a sweet-spot where the gain peaks and the tone is nicer. The TONE controls are high-cut that vary from 580Hz to 14KHz. In the ROG circuit, the JFET preamp runs at maximum gain. I added the PREGAIN control to reign it in a bit when needed. I also re-arranged the Bounce channel GAIN control to give more range and less noise. When I was first checking this out on the bench, it was motor-boating. After a couple of days of swearing, troubleshooting and more swearing, I came to the realization that the pedal was fine, it was the power supply that was oscillating. The 4049 presents a "funny" load to the power supply and this one didn't like it. It was fine with battery power or with a more stable power block.

The parts count is pretty low considering how much this beast does. I socketed the 4049 because some are noisier than others.
Double D - CCA 02.JPG

Yeah, the inside's a bit of a rat's nest. I was lucky: no squealing. The board faces down and the bottom cover has a layer of duct tape for insulation. Those resistors on the stomp switches are the ballasts for the LED rings.
double-D innards 02.jpg

The stomp switch on the right selects the channel: red for Jiggle, blue for Bounce.

Double-D front 04.jpg

Double-D front 02.jpg
 
Madbean has the Snarkdoodle that was a giveaway board a few months ago, based on the Red Llama. It uses the 4049 chip and is made for a 1590a box.
 
Nice Build Chuck.

I built a Custom Reezafratzitz V.2 with Mids control using my Eyelet Board Design a few years ago.
It uses the Texas Inst. CD4069UBE Hex Inverter but its prone to white noise.
I replaced it with an Fairchild CD4069UBCN & it is Great now.

Awesome Pedal !!!

FRATZ  Custom Distortion Pedal 1.jpg

FRATZ  Custom Distortion Pedal 2.jpg
 
Last edited:
That is some fun mods to what is already an interesting build. I first built Craig Anderton's Tube Sound Fuzz from EPFM, and more recently the Bloody Finger and Snarkdoodle. I like the idea of using more of the 4049 channels for the ROG build. It might be fun to try feeding each channel into the outside terminals on a pot so the wiper could dial some in-between sounds from both channels.
 
Music6000 reminded me that not all 4049s are created equal. Some perform better than others when run in a linear mode. Several manufacturers have made these parts over the years and they all have different recipes. A new TI chip I tried was a little noisy, a NOS Fairchild chip had low noise, but liked to motorboat at certain bias settings. I put in a NOS RCA chip and the circuit was quiet and stable. YMMV.
 
@Chuck D. Bones - I just built the Madbean Mysterioso, which is another cd4049ube-based overdrive/distortion (based on the Blackstone Appliances Mosfet Overdrive). I really like the sound of the Mysterioso, so naturally I went looking for other 4049-based overdrives, and I stumbled on your modified ROG Double-D. I'm actually working on PCB for it. A couple questions before I have it fabbed:
  • The stock ROG Double D has an output JFET buffer, your version appears to omit that. Any reason? If I wanted to add the buffer to your design, can I just follow what ROG did, and basically copy the circuit just past the Jiggle/Bounce output switch?
  • In general, have you made any more tweaks or updates to this design? Any other considerations I should keep in mind before having this fabbed?

Thanks! I'm happy to post any of the KiCad files if there's any interest.
 
Hi Matt,
In my opinion, JFET output buffers are unnecessary unless you're going to drive a long cable. You can easily add the source-follower back in. I'd omit the 220nF cap at Q2's gate, it's unnecessary. CD4049s are noisy, so be sure to install a socket. No other tweaks to the design except you should consider making the BIAS control an internal trimmer. I just set it somewhere in the middle and never touch it.

You might want to check out the CD4007 OD I posted in Chucks Boneyard. It's only one channel, but it's much quieter than the Double-D and has better tone controls. FYI, the first version has more gain.
 
Here's the schematic if anyone is interested.

View attachment 2856
Chuck,
A quick question on your schematic for the Double D, as I'm very interested in building it. I'm wondering about the tone controls, but maybe/probably I'm looking at it cross-eyed... Since VR2<<VR3, it looks to me that C10 || VR3 yields substantial mid/upper frequency rolloff for the Jiggle channel regardless of how the tone knob is set. (Same Q applies to Bounce channel.) What am I overlooking? Regardless, many thanks for your contributions here - cheers!

(added a little later) Hmm, so if the input impedance of whatever comes next in the signal chain is much less than VR3, then that makes the rolloff less severe (i.e., than the 1/2*pi*VR3*C10 frequency), but then we're impacted by whatever comes next. Maybe this is why ROG's original design had that output JFET buffer, eh? (Or am I violating the first rule of holes here - ok, I'll stop digging.)
 
Last edited:
The low-pass corner freq is determined by C10 multiplied by whatever resistance it sees looking out into the rest of the circuit. When VR2 is dimed, tyhat resistance is R12 || (VR2 + VR3) = 328Ω. When VR2 is at noon, it's (R12 + VR2 * 0.15) || (VR2 * 0.85 + VR3) = 1.8K. when VR2 is at zero, it's (R12 + VR2) || VR3 = 8.6K.

One other thing, the BIAS pot is really unnecessary. You can delete VR7 & R21, connect C20, C21 & U1-1 to R20. change R20 to 1K.
 
Ahhh, got it, THANK YOU - I overlooked the virtual ground at pin 15 of the chip...
I'll also look at BIAS changes you recommend, I had previously planned on using a trimpot there and not externally exposing that pot.
 
Back
Top