The Fractal Overdrive is a clone of the Pearl OD-05 Overdrive, a strangely obscure effect from the early 80’s when Pearl (yep, the drum company) briefly dipped into the effects pedal business. The result was a line of very well-engineered pedals that unfortunately never got the attention they deserved.
The interesting thing is, even 30 years later this pedal still is relatively unknown. There are no commercial clones of this pedal and it’s relatively unexplored in the DIY scene (with the exception of BYOC’s Parametric Overdrive project). But those who have played through one universally praise it for its tonal flexibility.
The OD-05 has a standard feedback-diode clipping stage like a Tube Screamer, but the magic is what comes in front of it: a parametric equalizer that allows either a cut or boost of up to 15dB in a range from 100 Hz to 4 KHz. With this flexible preclipping EQ, you can get some incredible sounds out of it. You can crank the mids for a cocked-wah sound, or tune the circuit to enhance the resonant peak of your specific guitar and amp. You can even get a scooped-mids sound if that’s what you’re shooting for, something that’s pretty rare in an overdrive.
The circuit has a lot of familiar things in common with others you’ve seen, but in this implementation they don’t always behave how you expect. This is not quite a Tube Screamer where you can tweak just about any part and get something good out of it! The original designers balanced the stock circuit very well, but if you’re modifying it, you may find that if you change one thing on one end you may need to change something clear over on the other end to make the first thing sound good.
Japanese electrical engineering companies were a dime a dozen in the 1980s, and with effects pedals hitting something of a golden age in the 1980s, many well-known instrument manufacturers like Gibson, Tokai and even Martin would contract with other companies to produce pedals under their own brand name. Oftentimes these were just relabeled versions of pedals that appeared under other brand names. The end result was a whole lot of guitar pedals, but very little innovation. (There were a couple of notable exceptions, like Ibanez—you may have heard of their contractor: it’s called Maxon and they invented a little gem called the Tube Screamer.)
Pearl—yes, the drum manufacturer—was no different, except that like Ibanez, they also bet on the right horse: Korg Japan and let them develop a series of effects pedals under the Pearl name, called the Sound Spice series. The result was what might appear on the surface to be a few clones:
If you have a peek at the schematics, though, none of them are straight clones: the engineers at Korg went out of their way to innovate, to design something new for each and every one of them. The compressor had a pre-compression tone control. The analog delay had wet and dry outputs. The octaver had far better note-tracking than any of its contemporaries and is universally considered the best analog octave pedal ever designed. And the overdrive had a parametric mids control—more on that later.
This line of pedals still remains highly regarded among people who know about them, and many of them have achieved cult status. But whether it was from poor marketing or poor management, or maybe just the name “Sound Spice”, Pearl did not experience much success with this line of pedals. They discontinued them in the mid-80s and exited the effects pedal market for good. You can still pick up most of them for pretty good prices.
I traced this one myself from an original unit. When I was doing my research for this project, I found four different schematics out there, including the factory schematic from Pearl, and every one of them had errors. In fact, the schematic from Build Your Own Clone—who went out of their way to say “none of the schematics out there are right except for ours”—has far more errors than any of them, although as far as I can tell, the PCB itself is accurate and the schematic was just misdrawn. They’re good folks over there, so no jab intended—I just found it ironic!
In the end, I found that the Pearl factory schematic was accurate except that they don’t show exactly how the dual 100kC pot is hooked up. But that’s not an error so much as a lack of clarity. Their part values were all correct and all the connections check out.
So, at the risk of saying something that’s been wrongly said before, this project’s schematic is accurate, once and for all.
…Or, pretty much. One quirk about the Pearl line of pedals is that they all share the same power section, and the reference voltage (Vref, AC ground) is created by using a 5.1v zener diode instead of the more standard voltage divider. This provides a constant reference voltage point regardless of how much juice is left in the battery, which is probably helpful for the effects in the series that need a fixed bias point, but is bad news for an overdrive. One side of the waveform will hit its voltage rail quicker than the other side, and the lower the battery voltage, the more imbalanced the waveform will be.
To that end, my project (and schematic) replaces the reference voltage section with two 10k resistors. The Fractal PCB has a stripe on the R28 resistor to show the orientation of the zener diode if you’d like it to be 100% original, but just know that this is a cut-and-paste engineering flaw and not something that gives extra vintage mojo.
Input buffer → gain boost → parametric equalizer → clipping stage → tone roll-off and volume control → output buffer
The OD-05 is a very interesting combination of circuit blocks. It’s composed of three main chunks: a boost, a parametric EQ and a clipping stage. We’ll go over each of these stages individually.
The boost section is somewhat over-engineered for what it does, which is to provide a full-range boost with a gain of 2. It’s a common emitter configuration plus an emitter follower. There’s nothing about this that couldn’t be accomplished with an op amp hooked up as a non-inverting amplifier. But instead, we’ve got two transistors, a bunch of resistors, and a few caps, including a huge 100uF filter cap. But it does the trick!
On to the parametric EQ stage. This allows for an adjustable frequency range from 100 Hz to 4 KHz, and the ability to either boost or cut that frequency by 15dB in either direction. This section is vaguely similar to the Ibanez ST-9 / STL Super Tube Screamer, with the notable addition of the cut along with the boost. (It’s actually identical to the midrange control on Lab Series L5 amps… more on that later.)
With this one, you can get tons of tonal combinations, especially considering that it’s placed before the clipping stage. You can crank the mids for a cocked-wah sound. You can tune the circuit to enhance the resonant peak of your specific guitar and amp. You can even get a scooped-mids sound if you want. The only limitation is that you only get one frequency band, so if you want to cut the highs you can’t also boost the mids. But that’s why you’ve got a tone control on your guitar, right?
The feedback clipping stage is the exact same topography as a Tube Screamer, with a different gain ratio and diode configuration, but the same low-end rolloff frequency and roughly equivalent overall gain. But the tone section comes before the clipping, which means you can play tricks with the clipping stage and get a lot of really interesting sounds that wouldn’t be possible by just shaping the EQ somewhere down the line.
The one interesting thing about this stage is the use of resistors in series with the clipping diodes. This has the effect of raising the clipping threshold slightly and softening it. I haven’t seen this before in any other commercial pedals, but it sounds good.
After the clipping stage comes a low-pass filter created by a 10k resistor (R25) plus a 15n capacitor (C12) to ground. This creates the other end of the natural mid-hump, with a roll-off frequency of 1062 Hz. It then goes into a standard volume control and output buffer.
This circuit is ripe for modification, with a lot of potential pieces you could tweak and a lot of really fantastic combinations that could come out of it. One thing to say ahead of time, though, is that this is not a Tube Screamer where you can tweak just about any part and get something good out of it. The Korg engineers balanced the stock circuit very well, but if you’re modifying it, you may find that if you change one thing on one end you may need to change something clear over on the other end to make the first thing sound good.
The specific modification recommendations can be found in the documentation above, but here are a few high-level things to keep in mind.
The input capacitor (C1) is 33n. If you tweak the gain/EQ of the drive section, you will probably need to reduce this input capacitor down to 22n to keep everything balanced. The input capacitor works well with the feedback diode + resistor combination, and with the gain ratio of the Drive pot to the low-pass filter off of the op-amp’s feedback loop, but if you change either of those, you may find that you like the the results better if you change the input cap.
The Q (width) of the midrange frequency can be changed by adjusting the ratio of C6 and C7—raise or lower them both by the same percentage. I haven’t found anything I liked better than the stock circuit here, though.
The clipping diodes are also very sensitive to changes. The traditional “diode lift” mode that’s so common with Tube Screamers doesn’t work very well here, and you’ll get some splatty artifacts in the clipping. Likewise, LEDs with their ~1.8v forward voltage also don’t work very well. The sweet spot is around 0.6v (one 1N914) to 1.2v (two 1N914s in series) and anything in between. You may get different results if you use different op amps, though, so it’s worth experimenting.
Speaking of op amps: the original uses two JRC4558s. These are good. The Lab Series amps use these as well for the midrange portion of the circuit. But since we’re still dealing with the clean signal at this point, you might find that you like a higher-fidelity op amp like a TL072 or OPA2134 for IC1.
I’m a big fan of Lab Series L5 amps, another obscure-yet-legendary line of gear that never got much attention until it was too late. (Sound familiar?) I created a pedal version of the preamp and in the process I became very familiar with the circuit.
When I traced the OD-05, I felt like I’d seen the boost/cut functionality before, so I opened up the Lab Series L5 schematic and sure enough, the OD-05’s two-knob frequency cut/boost is identical in topography to the Lab’s midrange section with some very minor value tweaks.
Did Pearl copy the Lab Series L5? The L5, designed by Moog Music, first came out in 1977. This parametric EQ circuit probably wasn’t invented specifically for the Lab Series amps—although it’s likely that Moog had a hand in developing it because they were so innovative with analog filters. I’d be fine with saying that there was a common source that both the OD-05 and the L5 drew from, and leaving it at that.
But… if you keep digging, on the fringes of the Sound Spice series is the rare TH-20 “Thriller”, another semi-parametric filter which I don’t think ever made its way outside Japan. I can’t find a schematic for this pedal, but check out some of the copy—here’s the Multipeak knob’s function:
Multipeak: Boosts the frequency of the six points of the high frequency range. Turning clockwise raises the peak level.
This sounds suspiciously like the Lab Series “Multifilter” control:
The multifilter is essentially a six-band EQ with fixed settings, where you control only how much of the signal through it gets mixed back with the main signal, sort of like a flanger’s comb filter with the sweep set to zero. The frequency centers are at 1000Hz, 1370Hz, 1900Hz, 2630Hz, 3630Hz and 5000 Hz.
So did Korg copy the Lab Series amps? Yes. And that’s awesome. Korg’s brilliance in designing the line of pedals for Pearl was that they knew exactly what to sources to borrow from, and they were able to use what they borrowed in fabulous ways.