modular parallel distortion box

[blana]

Hey,

Just started to work on the second version of an interesting distortion box idea and I wanted to share my work so far.

The idea came to me from the desire of having a heavy stoner distortion that is both brutal/extreme and intelligible at the same time in a band/live situation. These two things are opposed when speaking of distortion...

... but mixing two distortions one with each characteristic seemed a good place to start. I've read about other people/boxes doing it but I couldnt find one that was exactly for me, so I also needed a circuit that is easy to MOD and ITERATE with... so I came up with the idea of making it MODULAR.

Cutting the crap, here are some pics:

https://drive.google.com/file/d/0B3HEK-5LrRCyd3FTTnE5TzRFSkE/edit?usp=sharing

https://drive.google.com/file/d/0B3HEK-5LrRCyVXRZSHhJT05EWnM/edit?usp=sharing

Basically, its a mainboard with the in/out, power and pots into which you can stick any circuit that respects the pin layout. With it you can quickly test out circuits without having to worry about the jacks, the power and most of the pots

The first version i made (the one in the pics) has a few design mistakes which I've hopefully corrected in the second version.

Here is the mainboard schematic (v2):

https://drive.google.com/file/d/0B3HEK-5LrRCyRDVRbVNXTXJSa0U/edit?usp=sharing

The modules I tested with were:

1. modified muff distortion circuit (transistors 2 and 3)
2. modified MXR dist+ circuit
3. Horrific distortion with a cascaded muff distortion circuit


The results were very interesting, but the mainboard circuit had a small problem: the distortion control was too subtle due to a design mistake.  Both distortions were too strong even with the lowest gain - and tweaking their independent gains seems too tedious given the fact that the circuit needs to be changed anyway.  No soundclips yet, but I'll make some soon, probably with the second iteration of the mainboard.

Anyway, what I could learn from this first iteration:

- gain and filtering must be tweaked very carefully - esp for the hi gain circuit, so it leaves room for the cleaner one to be heard
- A good sounding mix I found was when low cutting the cleaner one below 720hz and the hi gain one above 720hz. (hi pass on low gain and lowpass on hi gain; yeah, i know its sounds mindf***)
- Another good sounding mix was surprisingly between the two gain circuits, with one of them being sensibly lower gain than the other.
- Cutting below 150hz right at the splitter seems to have a good effect, taking out the mud from the hi gain circuit.
- Cutting above 4000hz right at the splitter seems to have a good effect, taking out the fizz from both circuits.


So, I'm ready to make the layout for the updated schematic (the one in the pic), but before that I thought to share this - maybe someone else has valuable input on this. Any comments suggestions or criticism will be appreciated.

[Mark Hammer]

Take a gander atthe Empress Multidrive, which does what you're attempting to do.  http://www.empresseffects.com/multidrive.html

I say this not to put the kybosh on your project, but because it a) tells you the idea isn't flat out nuts, and b) may provide you with some ideas for what's enough, what you may want more of, or what you don't need.  The other perk is that there is no end of Youtube demo videos to give you an idea of what the finished product could sound like.

[blana]

Hey, thanks for the tip Mark! The Empress was the first thing I found when searching for parallel distortion, so it was indeed good as a starting point, as you said

I wish there was an available schematic for it.

Anyway, the thing is that every video i saw with it left me with the impression that it somehow sounded too "normal" on all of the settings those guys were dialing in...
Of course thats really cool for some styles, but when you are looking for something more personal, thats just not gonna cut it. Thats what got me DYI ing in the first place!

[Mark Hammer]

Oh believe me, with so many hundreds and thousands of distortion circuits out there, the very idea that 3 circuits could capture the range of what one might want to blend is foolhardy.  So if you have other sorts of fuzzes and such thatyou would like to be able to blend, go for it!

[blana]

Yeah, that's exactly why i thought that having a way to quickly itrerate with various modules would be very useful (not only for me) in this quest for the perfect dist blend

[Mark Hammer]

A good idea, in principle.  One of the things I think you will need to plan for is the often very different output levels of different circuits.  Yes, each circuit can have  its own volume pot, but if the full volume range of circuit A is equal to settings between 7:00 and 9:00 in circuit B, then blending becomes a tricky matter, and potentially frustrating.

Perhaps the solution is to pre-attenuate the signal on the louder circuits.  So, if A and B both have 100k output level pots, but B is a LOT louder than A, use a 50k pot for B with a 51k fixed resistor in series with the input.  This way, B can never be set above halfway, but you still get to use the full rotation range of the volume pot, for blending purposes.

[blana]

Yeah, this problem hit me also, and i thought of adding trimmers to the modules so i can individually adjust the gain on each one - since i dont have separate volumes for the effects on the mainboard, just a blend pot that pans between the two modules, for simplicity

I started by tuning them with a soundcard oscilloscope using a sine wave at the input and observing both exits. Using the trimmers i adjusted them until they were the same height. Now comes the unexpected part: even if I thought i had them at the same level, the thing was that the lighter distortion was clearly at lower volume. I was like "WTF" at first.

Correct me if I m wrong but What I think happens is that in fact the stronger dist produces more harmonics and fills more of the spectrum, being perceived as louder compared to the lighter one that fills less of the spectrum.

So the next method i used to tune it up was to set the gain somewhere in the lower mid part and set the blend pot at maximum for each of the distortions while turning the trimmers to make them seem equally loud - did that make sense?

[Mark Hammer]

In general, it makes sense.  I think you should aim for whatever gain in each circuit gets the best sound out of them, and then use selective/adjusted attenuation on their outputs to get a reasonable balance.  It doesn't have to be exact same maximum level on each.  What you're aiming for is being able to get a desired blend, and not find yourself being unable to dial in enough of one of the sounds.

And, like the Multidrive and the Boss ROD-10 before it, sticking a 2 or 3-band EQ and master output level after the blend/mixing-node is a good idea.

[ashcat_lt]

Correct me if I m wrong but What I think happens is that in fact the stronger dist produces more harmonics and fills more of the spectrum, being perceived as louder compared to the lighter one that fills less of the spectrum.

I'd be careful with that.  I mean, it's kinda true, but I'm not sure that it's really "filling more of the spectrum" so much that the crest factor - the ratio of peak to average level - is bigger for a less distorted signal.  The RMS level (generally equated to "loudness") of a square wave is exactly the same as its peak level, while that of a sine wave is only about 0.7 times the peak. 

[blana]

... really "filling more of the spectrum" so much that the crest factor - the ratio of peak to average level - is bigger for a less distorted signal.  The RMS level (generally equated to "loudness") of a square wave is exactly the same as its peak level, while that of a sine wave is only about 0.7 times the peak. 

Yeah, basically i imagine the power (or loudness) is proportional to the surface below the waveform - hoping the visual description makes sense.

And, like the Multidrive and the Boss ROD-10 before it, sticking a 2 or 3-band EQ and master output level after the blend/mixing-node is a good idea.

Yeah, EQ-ing would be nice, but i was thinking to do it separately for each module (before the blend) to make the sound richer.

I also found this pedal which uses a very interesting signal splitting system.

http://www.brimstoneaudio.com/crossover-distortion.html
https://www.youtube.com/watch?v=KtW_BawdnPs

[Mark Hammer]

Yeah, EQ-ing would be nice, but i was thinking to do it separately for each module (before the blend) to make the sound richer.

A standard "Tone" control for the individual distortion circuits is fine, for establishing the basic character to be blended.  But my gut says that more than a single more-complex EQ-ing setup is asking for needless distraction.  I can see something like two channels of master EQ for alternate overall sounds, but if one is tweaking each of 3 (or more!) bands for each of several distortions, it is too easy to get lost in details that ultimately may not matter.