Fundamentals of designing your own overdrive?

[canman]

Hi guys,

First off, thanks for all the help I've received from you all over the last while, it's been very much appreciated.  I'm learning a ton and I've been doing some reading in hopes of better understanding stompbox circuits, and also in hopes of contributing more to this community.

I have lots of reading and learning ahead of me, but I thought I'd at least ask and see where I should focus my reading and studying, in hopes of designing my own overdrive circuit.

So the question is, what are the fundamental "stages" of an overdrive circuit?  From what I've gleaned on here, it seems like it goes something like this:

Input
Resistor to set impedance (still can't wrap my head around impedance, so there's one area to focus my studies)
Input cap (blocks DC, sets frequencies to be affected by circuit?)
Gain Stage Number One (not sure if this is where you introduce the overdrive sound you'll hear at the output?  Add a gain control here?)
Tone Stack (if desired...likely something I'll desire, haha.  Add treble, mid, bass, tone, etc. controls here?)
Recovery Stage (tone stack sucks a bunch of signal, so it needs to be boosted again, right?)
Resistor for output impedance (Is this necessary?)
Output cap (Remove DC again?)
Volume pot (Volume control here, naturally...right?)
Output

So how far off am I here?  Are these stages correct, am I missing anything, am I just making crap up...how close did I get?  Go easy on me guys, I'm not nearly as in tune with electronics as you guys are!  I just want to see if I can put the pieces together from what I've read and learned here, and hopefully try to piece something together that's *kind of* unique

[Transmogrifox]

To simplify further, the fundamental building blocks of a basic overdrive/distortion of any kind usually follows this basic topology:

Pre-emphasis EQ (usually boost the highs/cut lows to "tighten" up the distortion quality)
Amplify
Clip    (amplification and clipping are generally part of the same amplification circuit -- you drive the amplifier to reach beyond its limits but it can't)
Recovery EQ (Usually a roll-off on the highs that were boosted in stage 1, but sometimes a bass-boost or mids-notch)
Tone stack  (sometimes recovery EQ and tone stack are one in the same circuit)

A really good one to study for the classic overdrive pedal is the Tubescreamer.  In fact, RG Keen has been kind enough to explain each part of the circuit in a simple way that most can understand:

http://www.geofex.com/article_folders/TStech/tsxfram.htm

If you start with that article and work your way through each circuit function you will come out the other end understanding how an overdrive works and what makes it sound the way it does.

From there you can start to tweak components to learn what happens to the sound.  The basic TS circuit is such a major building block in most overdrives you will find that many commercially sold "Overdrive" boxes are pretty much modified versions of this circuit.

[GGBB]

Keep in mind that there is quite a bit of variety in the "overdrive" world, but most of what we refer to as overdrive pedals use a combination of gain and diode clipping to create the "overdrive" sound, such as in the TS. Have a look at runoffgroove's Peppermill for a different kind of overdrive - a MOSFET over-driving a JFET - no clipping diodes at all.

[Mark Hammer]

The "stages" depend on what you want/need it to do for you.  Does the user expect the overdrive tone to come strictly from the pedal itself, or is there an assumption it will push the amp into contributing some, or even all of the clipping?

[canman]

Thanks for all the replies, very informative and very appreciated.  

Transmogrifox, with the pre-emphasis EQ, is this shaping done by various capacitors filtering different frequencies?  Or is this more of an EQ circuit with a tone control, etc.?  I really like the concept of the pre-emphasis EQ tightening up the overdrive.  Very cool, thanks!  EDIT:  Does the pre-emphasis EQ often serve as an input cap as well?  Or at least, does the input cap serve as part of the pre-emphasis EQ?

GGBB, I haven't quite explored what kind of clipping I'm looking for, but something simple, I'm sure.  I'm no engineer, just a hobbyist.  I know I've enjoyed MOSFET overdrives (OCD comes to mind) so I'll study up on the OCD and the TS for diode clipping.

Mark...excellent question.  Honestly, I hadn't even considered the tone coming from the standalone pedal or pushing the amp to contribute...I personally enjoy a clean amp, sometimes a slightly broken up amp, so I'd say I'm going for a pedal that delivers a standalone tone, but can push other pedals well or goose an amp a bit if need be.  But the goal isn't to push an amp, like the tubescreamer, for instance.  Does that make sense?

[disto]

Have you seen the 'Design your own distortion' I'm sure it used to be called 'Cook your own distortion'? It covers some of the basics and might help give you an overview of the different stages. There is a link here http://www.generalguitargadgets.com/how-to-build-it/technical-help/articles/design-distortion/.

RG Keen's technology of the tubescreamer and fuzz face are also really good if you wanted to know how they work. I honestly wish there were more examples like this.

[Mark Hammer]

Makes perfect sense.

If a user expects the pedal to produce its own distinctive sound, but append that sound with a contribution from amp and/or speaker breakup, then very often there's not really any "stage" that affords that, simply a passive control to trim back high end.

Of course, some folks may want more flexibility in how they push their amp.  In those instances there will be more complex tonal adjustment, usually either built around an active stage, or else passive, but requiring some added gain to compensate for passive signal loss.

In general, if the intention is to push an amp into submission - as opposed to simply being a little louder for solo-ing purposes - then there will be additional gain on the output stage, or perhaps a "gain-recovery" stage.

Jack Orman has some interesting graphs posted on his AMZ site, comparing several ostensibly "clean" boosters, and the harmonic content they have at their output.  It's clear from the graphs that some lack pretty much any coloration and whatever clipping you hear is essentially from whatever comes after the booster (which could be the amp alone, the speakers, or any devices between the booster and amp).  Conversely, there are some that, while not overtly sounding like an overdrive, DO produce an output with additional coloration resulting from lower-order harmonics.

[smallbearelec]

most of what we refer to as overdrive pedals use a combination of gain and diode clipping to create the "overdrive" sound, such as in the TS.

Yep.

Have you breadboarded anything yet?

http://diy.smallbearelec.com/HowTos/BreadboardICDist/ICDist.htm

[canman]

Lots of really good information here and the links are very explanatory...I'll have quite a bit of studying to do in the next little while before I even try to design something myself...

Most eye opening to me is the interaction between resistors and capacitors, how frequently they create high or low pass filters, and how the value of the capacitor needs to change so that you don't affect the tone (or change the value so you do affect the tone).  Had no idea that interaction was occurring...clearly I need to study more.

I haven't breadboarded anything yet because I haven't felt like I have a firm enough grasp of what's going on to really do anything useful with a breadboard.  But I think that time is approaching soon!  Thanks for the link, I'll be reading that for sure!

EDIT:  OK so I've been looking over this OCD schematic:



After reading through the Build Your Own Distortion page, I was hoping you guys could let me know if I'm correct in understanding a couple things about the OCD.  First, the dual opamp...I never understood why the use of a dual opamp vs a single opamp.  But after looking at the schematic, I think I get it now.  The first half of the opamp is for controlling the gain.  Then you have MOSFET clipping the signal, and then the second half of the opamp is somewhat of a recovery/volume section, with the volume control after it to control how loud it gets.  The tone control doesn't require any kind of a recovery stage because it's just a high pass filter, right?  Or is that a low pass filter..?  Anyways, by using a dual opamp you save on space and don't need to single opamps...obvious now, but I never made sense of it until reading through these links.  And is that a voltage divider I see at pin 3 of the TL082, to bias the opamp?

Also, am I seeing another filter with the 2k2 resistor and the 68n cap in the first gain stage, to control the frequency roll off?

[PBE6]

Dual opamps save you a little bit of money and a little bit of space. You can accomplish the same thing with two single opamps, but using a dual is almost always easier.

The first opamp does indeed provide gain to the signal so it can be clipped by the MOSFETs, which are wired as diodes.

And yes, the second opamp boosts the signal to recover some of the voltage that was lost in the clipping section.

The tone control is a first order low-pass filter (high-cut filter). It also loses some signal, but not nearly as much as the clipping section. There's more than enough left so that it doesn't need any additional boost before the volume knob.

The 2k2 resistor and 68n cap in the ground leg of the first opamp's feedback loop form a high-pass shelving filter.

Good work analyzing the circuit, looks like you are able to identify the building blocks pretty well!

[canman]

Awesome!!  Thanks for the confirmation and compliment, I appreciate it.  Nice to know I'm actually learning something here. 

So the real trick is to breadboard everything and use values that will give you the frequency you're after...I've come across that formula a few times in my readings, I'll have to go back through it and make a spreadsheet with a bunch of values or something.

When you need to use a cap that won't affect the tone, it seems the solution is a larger value cap.  Does this mean you're achieving a frequency that we just can't hear?

[antonis]

When you need to use a cap that won't affect the tone, it seems the solution is a larger value cap.  Does this mean you're achieving a frequency that we just can't hear?

Hmmmm....

It depends on the cap's position in the circuit...
(if it's combined someway with a resistor, it WILL affect the tone...)

Larger cap values decrease frequency (and vice versa) so you can obtain the same result (not audible frequency) with smaller values...

I mean that you can't hear a frequency which is out of a specific filter band without bothering if it's lower or higher than the cut-off frequencies..

And is that a voltage divider I see at pin 3 of the TL082, to bias the opamp?

Quite right, but the actual voltage divider is created at the two 10k resistors junction..
The 470k resistor is biasing op amp's non inverting input to Vcc/2 (and sets - with 1M &10k - input's impedance..)

[knutolai]

tl;dr 
Impedance: http://www.muzique.com/lab/imp.htm
Gain stages/distortion: http://www.muzique.com/lab/boost.htm
Buffering: http://www.muzique.com/lab/buffers.htm

[canman]

Thanks for all the replies and additionally reading material.  Enough to keep my busy for weeks...naturally, I won't wait that long to start trying to design an OD haha...I'm sure I'll have more questions as I read though, thanks for all the help!

[PBE6]

My 2 cents..

I think the first choice to make is to choose how your dirt pedal is going to generate dirt. The usual suspects include diode clipping (either to ground or in a feedback loop) and overdriven devices (like BJTs, JFETs, MOSFETs and opamps, all in various configurations), but whatever the choose this will be the heart of the circuit.

Next I would choose the EQ/tonality of the pedal. Is it going to be mid-forward or mid-scooped? Dark and woolly or bright and tight? Or as neutral as possible? Depending on what job the pedal will be doing, you can decide what EQ building blocks are necessary and where to put them (if any).  The overall EQ scheme will make a huge difference to the sound, Mark Hammer has some very useful advice on this subject.

Third I would think about "special effects", like clean blends, octave up effects, compression, etc.. Most pedals don't need any of this stuff to sound great, but variety is the spice of life.

Once the major decisions are made, you can start putting the building blocks together. And once you have a rough draft together you can look for impedance issues, synergies and simplifications. After a few review/revise sessions, you should have a dandy original circuit!

[canman]

I suppose the hardest and probably most exciting part would be deciding how you want to get the overdriven sound.  I haven't given it much thought yet, but I for sure want something warm and natural sounding (along with every other guitarist out there  ).  I know MOSFETs and JFETs are supposed to be the most tube-like, so I'll probably go down that path in some way or another.  

For sure going for an OD that has the capability to be transparent, but with a 3-band EQ to you can use it to shape your tone if you want.  I really dig the tone stack on the Wampler Triple Wreck...really useful and natural sounding. 

A blend knob would be pretty cool too...easiest way to incorporate that would be a buffer of some kind, maybe?  But that's probably as far as I'd want to take any special effects, haha.  Blend knobs are just sweet!

[Quackzed]

this might be helpful...
http://www.diystompboxes.com/smfforum/index.php?topic=78111.0
it's a thread that goes through a big muff distortion pedal one part at a time , explaining what is doing what as you go along...

[aron]

This is an awesome site!!!!!!!!

http://www.gmarts.org/index.php?go=217

[canman]

Just spent some quality time on that site...I'll be referring back to that many times in the near future!

EDIT:  I had a quick little thought, with regards to breadboarding your own circuit (or any circuit, for that matter).  Say for instance that I wasn't sure what frequency I wanted to initially flow through the circuit.  I see a lot of high-pass and low-pass filters at the input of the circuits I've reviewed thus far, sometimes one of the two, sometimes both a high- and low-pass filter.  Would I be able to hook up a power supply and use an audio probe to test and see if I like the frequencies that are allowed to pass through?  That is, do I need to connect the rest of the circuit (opamps, tone stack, recovery stages, etc.) or can I simply breadboard the high- and/or low-pass filter and probe the output?

I imagine that might be a good way to build up an overdrive...put pieces together one little chunk at a time, probe, and adjust until satisfied.  Tedious, I'm sure, but VERY educational, assuming that's an option.

EDIT ROUND 2:  I know you guys for the most part don't like Tayda.  BUT...are their breadboards any good?  For $2.50 I figure it can't hurt to try.  I'll be ordering one with my next round of pedal stuff, whenever that will be.  Any size recommendations?

[Thecomedian]

FET stuff is more tube like because of some property regarding how they work. I've seen a video lying around explaining that FETs are highly non-linear, and usually what happens is you "zoom in" with the amplification.

https://www.youtube.com/watch?v=tDQ0aBdUAHE&feature=player_detailpage#t=492

Basically, you want FETs driven out of the tiny linear area into a larger change, and they easily start to distort.

I threw together a mockup of a 2n3819 in spice with 1u caps all around and a 4k drain/1k source resistance setup and 1meg on gate to ground, and getting up to 1 volt AC with a 15 volt DC, and its starting to do a nice soft clip on one side, while the other side is getting a harder clip. So, I guess you could bias this first stage so that it has a gain of around 1-5 and produces that soft clip on one side while leaving the other side unclipped, in order to go a second stage to clip that other side, achieving soft clipping on both ends. Need a pre-amp stage to boost the incoming voltage up a bit from a weaker guitar signal, possibly.

I've been fiddling with how to emulate voltage sag for BJT-based circuits to get soft clipping from them.

There's only one fundamental rule to follow for overdrive: make the signal lose linearity. The rest is all up to what people enjoy sonically. Some people like harsh fizzy sound, some like that kind of guttural sound that is similar to a muscle car with those special exhausts for making it go brrroom.

I don't think there's any wrong way to produce a circuit, since people will love and hate the same circuit because of how it sounds, when they've got so many variables to work with.

Person A has a type of guitar and amp and thinks this circuit sounds ugly.
Person B has a different guitar and amp and thinks this circuit sounds amazing.
Person C has the same stuff as person A and thinks it sounds amazing.

I feel like its a bit tilting at windmills trying to come up with "the right" sounding circuit. It just needs to sound "right" to the person its designed for.


I just noticed you said people here don't like Tayda? I get a lot of stuff from Tayda and they have amazing pricing, great for anyone starting out with not a lot of money or perhaps not sure yet if DIY electronics is the hobby for them.