Onkotherium (bass) overdrive

[Fancy Lime]

Hey everyone,

here's a little overdrive circuit I designed. It started off as an attempt at a "tube sound fuzz deluxe" for heavy stoner rock, sludge, doom sounds. It obviously needed an output buffer at the end and a gain stage up front because high gain from CMOS inverters is just too noisy. So then I basically had a Rat with CMOS clipping. But how about using that Op-Amp stage for additional clipping? MOSFETs as clipping diodes was kind of an obvious choice since I wanted very soft clipping to retain the dynamic, so that the CMOS stage still had something to work with. Important thing: The additional diodes (D2 and D3) need to be pointing the opposite direction as the body diode of the BS170 they are connected to, else you just get the body diode clipping, which behaves pretty much the same as a normal silicon diode. The way that it is shown on the schematic you get very soft clipping, ideal for driving additional clipping stages. There are many different forum posts that describe how to do this correctly but for some reason almost all schematics I have ever seen that supposedly use this method have the diodes pointing the wrong way. Very frustrating, that. Now it's more like Robin Tomtlund's Red Fuzz ate a Red Llama. Anyway, here's the schematic:


EDIT: Updated schematic to fix error. R18 was 100k but must be 1k. Thanks to anotherjim for pointing that out!

Design notes:
There are a lot of parts that can be left out depending on taste (marked red in the schematic), I'll discuss the "full version" here.
The first stage, a CA3130EZ in non inverting mode, has a Gain and a Bass control, which are interactive. If you adjust the Bass you'll want to re-adjust the Gain if you want to keep a similar drive. The Bass control lets you go from tight and controlled (a bit like a soft boiled tube screamer) to almost fuzzy boominess. The overdrive from this stage is very soft and fat. The mode switch engages and disengages the clipping diodes. In the setting that uses C5, only the high frequencies are clipped, letting bass through clean like in a Green Russian Big Muff (only more so). The effect is similar to biamping but less flexible of course. Thanks to Alexander (swever) for inspiration for that. C5 might need adjustment to taste. Or a rotary switch with several options?
Next comes the mid cut, nothing special here, other than that it is highly interactive with the following stage, namely two inverters of the CD4069UBE controlled by the Volume pot, which controls how much this one overdrives. This stage's overdrive is much more dynamic than the first on. It sounds fresher and less fat. More blues, less comfort.
Last we have a rat-style treble filter, output buffer and Master volume control. The choice of filter pot and buffer input impedance is such, that decreasing treble also decreases the volume. Normally we want to avoid this but here it is done om purpose because I want to keep the output buffer from adding too much overdrive at high "Volume" (from the CD4069UBE) settings. Usually, high Gain and Volume means that treble needs to be dialed back and the way things are designed here, this leads to some volume reduction on the input of the buffer.

This thing has a monstrous gain range from almost clean to unusably distorted if you crank Gain and Volume. So to get the desired sound, I put Gain and Volume to minimum, then increase gain to the point of slight overdrive. If the mode switch is in the off position, there will be hardly any overdrive from the first stage, so this should be set to the desired mode first. Then I increase volume if I want a rougher, dirtier sound or I increase Gain if I want a fatter, nicer sound. Being able to balance the two is the best thing about this pedal, imho. The Mids control is a bit unnecessary, to me it really sounds best with the mids on full, but some might like a scooped sound so I left it in.

I like it a lot for sludgy, doomy, stoner rock sounds on bass. There is a reason why I called it Onkotherium (= massive beast or swollen beast). Some might want to strap a fuzz in front of it and take it to the desert. Getting Mojave flashbacks...

When I was done designing this, I realized that there are several other more or less similar commercially available pedals on the market. This is not a modified clone of any of them nor a mash-up of several, just convergent evolution due to a number of obvious design choices. I'm sure there are more that I don't know of. Please let me know if you know of something that is so close that the Onkotherium looks like a ripoff. That's not my intention so in that case I'll take it down.

Have fun with it and as always let me know if you find errors in the schematic or if you think I'm trespassing on your intellectual property.

Cheers,
Andy

[swever]

That's really very interesting. Will study this thoroughly and breadboard it as soon as I find a couple of free hours.

Will other, more common opamp types work in place of CA3130EZ?

[Fancy Lime]

Sure, other op-amps will work pretty much the same way with some tweaking of the surrounding components. I wanted something with a high input impedance so I went with a MOSFET op-amp, but there is more than one way to shave a weasel. The LM308 comes to mind, of course, but that one has a lower input impedance. You could use that and simply add a JFET buffer (like on the output) before it. It will sound a bit different but not necessarily worse. A TL071 or half a TL072 could also be used, input impedance should be high enough even without the extra buffer. If you use a TL072, you could even use the other half to replace the 2N5457 buffer at the end (that should make no audible difference but would give you a lower output impedance which can be useful depending on setup).

Cheers,
Andy

[anotherjim]

Interesting and cool naming.
I'd try other CMOS op-amps. I have TLC272 duals (don't know why they don't get used much). Wouldn't use TL072 myself if it's allowed to clip - I don't like them clipping.

The 100k dropper in the 4069 power rail. May I ask why? If to lower the current, beware it can effect the signal, not necessarily in producing compression. If you probe or listen to pin14, you might hear a frequency doubled signal (not necessarily a nice one). I feel a lower regulated supply for the CMOS works better. Have you tried running it from a 78L05?

[Fancy Lime]

Hi Jim,

oh $#!+, thanks for pointing that out! 100k for R18 is of course way to high. It probably would not even work. It is supposed to be 1k. Stupid copy and paste error. I updated the schematic:



Yes, I have thought about using a 5V regulator (UA7805) or maybe an adjustable LM317. Just have not gotten around to trying it, my local candy shop has slightly inconvenient opening hours. The thing sounds just like I wanted with R18=1k and 9V regulated power supply. If one wanted to make it eat, say 9-18V, a regulator is certainly a good idea for anything that uses a CMOS inverter in non-digital mode.

Thanks,
Andy

[anotherjim]

That explains it

I agree the inverter magic is best when it's low gain and driven by a better amp. To my mind, it's better that the driving amp doesn't clip too easily while the inverter is kept pushing into its soft clip region. That at least means either the driving amp can output rail to rail before clipping or the inverter is powered at a slightly lower voltage or has a bit of gain itself. It only needs x2 to definitely limit before the driving amp does, which might be easier than dropping the inverter supply volts.

Consider what happens to the normally high idle current of the inverter when signal swings through it. Either the P or N channel will turn off and then there is very little idle current. If you are dropping significant voltage via a supply dropper resistor when the inverter is idle, then the supply voltage will be modulated by signal at twice the signal frequency (because P or N are alternately off in a cycle). The chip supply volts rising on each swing will be modulating the signal amplitude causing volume expansion. This opposes the soft clip compression the inverter is usually chosen for. That said, it don't mean its bad. If it sounds right, it is right.

[Fancy Lime]

Hey folks,

finally got around to recording something with this. Just going through some different setting that I like. There are a million more possibilities but I don't really dig all of them. Please excuse the playing quality and the out of tune instrument. The circuit is still on the breadboard and catches some noise from the laptop powersupply. Should mostly go away once its in the cage.

http://www.aronnelson.com/DIYFiles/up/Onkotherium.mp3

Signal chain: Bass (Cort B4 fretless with the active electronics removed)  -> Onkotherium -> focusrite scarlett 2i4 -> Ardour4. No effects, compression or eq-ing in the software.

Cheers,
Andy

[Mark Hammer]

Jeez, I haven't heard Robin's name in ages.  I wonder whatever happened to the Scandinavian crowd: Robin, Stellan Lehrberg, and Tobias Karlsson.

Should there be a DC blocking cap in tandem with R10, or am I just accustomed to seeing one there for pedals intent on shaving bass?

[Fancy Lime]

Hi Mark,

no, you don't need a DC blocking cap between CMOS inverter stages if they have negative feedback (or any inverters for that matter). They both idle at 1/2 the supply voltage of the CMOS and since they keep output = -input, they adjust for any DC component on the input automatically. Adding a cap here would only serve to control bass, as in Craig Andertons TSF or in the Red Llama.

The Scandinavian stompbox scene is alive and well but seems a bit reclusive. That may just be a cultural thing. Frederik Lyxzén is doing very interesting stuff up in Umeå I think (parasitstudio.se). And here in Helsinki we seem to have enough of a scene to support a specialized guitar-electronics DIY shop (Uraltone) in the times of internet. I'm guessing Dark Glass Electronics emerged from that primordial DIY soup but I don't know who the people behind that are. Judging by the designs of Mad Professor pedals, they too seem solidly rooted in the DIY scene. Have not heard or read anything from Christian aka Hemmo since forever, though.

Cheers,
Andy

Edit for the geography geeks: Yes, I' aware that Finland s not part of Scanidnavia. So what I mean is "the Nordic scene". Or maybe the Fennoscandian? I don't know of any pedal builders from the islands...

[Mark Hammer]

One way to find out more is via the brand interviews on the Effectsdatabase: http://www.effectsdatabase.com/interviews/brands

Look for something that sounds familiar, and find out how they started.

[jonnyeye]

As a fellow bass player, this sounds good enough to build! One question, do you know if any important distortion happens in the second section of the CD4069 (labeled IC2d) or is it just being used as an inverting buffer? If it just inverts, I'd put the treble control where R10 is (using Mark Hammer's SWTC) and delete the output stage, probably paralleling the rest of the 4069 stages to increase current output ability. But that might be removing too much...

[Fancy Lime]

Hi Jonathan,

thanks! Since the second inverter stage has a gain factor of 1 it is in overdrive as soon as the first stage is. That has the effect that the combined knee from linear amplification to clipping is a little softer with the two stages than it would be with just one. That being said, replacing R10 with a 100k pot and add a cap to make it a SWTC would still work. It will sound a bit different than the way I did it but that's not to say it will sound worse. Just bread board it with the SWTC and see if you like it. Let me know how it went. If you don't like it you can still add the Rat-style tone control and buffer. IIRC Mark suggested a linear pot in his original article, I actually like a rev-log characteristic better for this purpose but they can be difficult to get and the difference is not huge.

Cheers,
Andy

[jonnyeye]

Hey Andy, thanks - I haven't yet played with the CD4069 so my intuition for how it works is undeveloped. I'll breadboard it both ways and report back.

[Fancy Lime]

Looking forward to it.

Andy

[Fancy Lime]

Schematic has gone missing. And back it is:

[rankot]

Ever thought of splitting the audio band and feeding the bass in parallel with distorted mid&treble sound?

[Fancy Lime]

Yes, but it's been a while. I don't really like the whole "layered sound" thing with splitting, processing separately and the mixing again. I never managed to make it so that the signal did not loose its "compactness" if that makes any sense. It always sounded a bit airy, something which I don't want in a bass sound. I think this feeling of airiness comes from the different dynamic responses of the different signal paths. Often makes it sound like there is a foreground and a background to the sound. It's much easier to keep things together by simply frequency shaping the signal pre- and post clipping. You also have to be rally careful with phase shifts introduced by RC-filters, or you might get unintentional out-of-phase sounds. In short, I don't quite see the point in signal splitting for bass distortion. It seems just like a half-baked workaround to make a guitar circuit sort-of-work with bass. Much better to redesign the whole thing for bass frequency response from the bottom up. Ont trick I do like to use in diode clipping circuits, though, is to only clip the highs by using a smallish capacitor (C5 in the schematic). Or use very asymmetric clipping by just using one diode instead of two (does not work with the capacitor, though). This retains the dynamics on one half of the swing while still sounding audibly distorted.

Andy