Angry Grouse: An amp-ish overdrive for guitar and bass

[Fancy Lime]

Hi all and a healthy new year!

I have been toying with some ideas on the breadboard again and found something worth sharing. I called it the Angry Grouse because I happened to be looking at old vacation pictures from Grand Canyon, which included some of an "angry grouse" that the rangers warned us about. It turned out to be a rather nice grouse, so I thought the name would be fitting. Anyway:



The schematic is not pretty but optimized for turning into a PCB layout in KiCad, which I intend to do one of these days. You can ignore the PWR_FLAG things, these are just to make KiCad's bug checker happy.

The inclined connoisseur will see some resemblance with the 1990's Marshall line of pedals. The idea of putting a Baxandall tone control inside the clipping section comes from old Orange amps (although they have a James before the clipping stage but same difference, in a way). Turns out that this makes for one hell of a flexible drive pedal, especially at low to medium gain. At high gain, the effectiveness of any boost on the EQ is naturally quite a bit diminished. So this thing really shines a medium gain and if you need high gain, put another drive pedal in front. That is why I consider it an amp-like overdrive, because I would use it and combine it with other drive pedals in much the same way as I would a one-channel dirty amp. But it is not supposed to sound like this tube amp or the other. However, the thing sounds very nice with guitar but almost better yet with bass. I found myself automatically playing a lot of Motörhead riffs while testing this. Go figure.

The gainpot can be lowered to 250k for a bit finer control on the most useful range. Just for good measure, I also included a primitive "cab sim", which is really just three orders of low pass. It is tuned rather darkly, which I like, but others may wish to move the corner frequencies up a smidgen. The way it is, I think it best goes directly into a mixer or headphone amp.

I also tried an NJM4556 and could not tell a difference from the TL072, but R3 should be adjusted to 470k when using a 4556 or 4558. I tried various clipping diode arrangements and must say I am rather taken with the very simple red LEDs. Nice and crunchy and the right amount of headroom for this circuit. Q1 can be any medium to high gain NPN.

I originally planned to release this together with the PCB layout but it takes me forever to do those and Mark (https://www.diystompboxes.com/smfforum/index.php?topic=126135.0) forced my hand I promise, I'll do a layout and sound clips someday but it may take a while. Busy times...

Cheers,
Andy

[iainpunk]

that clipping in tone control idea is genius, i wish i came up with that myself!
do i understand correctly that if you want to put this through an amp, you can leave out the tone shaping after the last opamp?

cheers, Iain

[Mark Hammer]

Intriguing.  So does stage 2 result in more clipping as the tone controls are set for more boost, and less when set for cut?

And just to shine a light on my naivete, is "n33" equal to 330pf?

[Fancy Lime]

that clipping in tone control idea is genius, i wish i came up with that myself!
do i understand correctly that if you want to put this through an amp, you can leave out the tone shaping after the last opamp?

cheers, Iain

Yes, the tone shaping is probably redundant in that case.

Intriguing.  So does stage 2 result in more clipping as the tone controls are set for more boost, and less when set for cut?

And just to shine a light on my naivete, is "n33" equal to 330pf?

Yes, boosting bass or treble distorts them more, cutting cleans the respective frequencies up. So you can tighten up the bass or make it fat and fuzzy. Obviously, this makes the tone controls highly interactive with the gain control. I was a bit worried that the effectiveness of the tone controls would suffer to much in such a design but it turned out that this only becomes an issue at high gain. At medium gain, bass or treble cuts quickly drop the respective band below the clipping threshold with red LEDs, this getting back their effectiveness. Boosting shapes the tone only at lowish gain settings but shapes the character of the clipping at medium gain. The controls are thus a bit unusual but I found them rather more intuitive than I has expected.

And yes, n33=330p, M10=100k. I like the fact that you can express any standard R or C value with just two digits and one letter. It's the sort of neatness I enjoy. Don't tell my wife, she already thinks I'm crazy.

Andy

[bluebunny]

So you sit through ten months of a pandemic and then two great overdrives come along at once!  Thanks Andy. 

BTW, expect Rob along at any minute.  The word "Grouse" will be like a magnet...   

[Fancy Lime]

Rob's been demagnetized, it seems... BTW, I forgot to mention what D1 and D2 are for. Since this thing is designed to be used like a dirty single channel amp, I wanted it to play nice with boosters up front. Kick on the treble booster for a "lead channel", sort of thing. The diodes at the input protect the TL072 opamp from straying out of it's common mode input range, which sounds, sham, "interesting" but not exactly great in all circumstances. If you use an opamp that does not suffer from phase reversal, D1 and D2 are not necessary.

Cheers,
Andy

[Lino22]

I apologize for a beginner questions
1. How does that bypass/feedback network around C10 work<
2. Again, what are D1 and D2 for please? I know Andy explained it but i still might not get it ... 
I thought they might clip the signal that would make the chip distort, then i realized that a clipped signal is already distorted ...
Or is that because the diode clipped signal sounds nicer than an overloaded opamp? LEDs have quite a big voltage drop so they let a non-boosted signal thought ... right?

[GibsonGM]

The feedback at C10 is tapping off some highs and sending them around to bypass the buffer that follows.
R11/C11 LPF what is left.  What this looks like would be easy to check out in LT Spice. 

The interaction of the 2 filters seems kind of interesting. I can only assume (without doing math) that this was designed to enhance highs/sustain, perhaps cut mid too...part of the 'cab sim' he spoke of(?)

I assume D1 & D2 are there to prevent clipping the opamp if a heavy signal is applied before the circuit...(nicer than overloaded opamp).

Correct me where I'm wrong.     

[antonis]

C10 apparent value is much lower than 3.9nF due to Q1 Emitter bootstrapping..
(3.9nF X (1-A) where A = R12 / (R12 + 0.025/Ic)
Something like 25pF to GND, in the absence of R13/C12 LPF..

D1/D2 prevent TL072 phase reversal (inputs swap their polarity) when common-mode voltage gets within a couple of volts of the negative supply rail (GND for single supply..)
Output may hit the bottom rail and then shoot up to hit the top rail (square-wave clipping, might be good.. ) or just latch up till powered off (DC line waveform, might not be good.. )

P.S.
Andy, are R15 & R16 values 10M indeed..??

[GibsonGM]

I thought it was bootstrapping, but edited my reply because I wasn't sure     Thank you Antonis.

[antonis]

I see your point, Sir..

We're used to bootstrap resistors via capacitors, in which configuration capacitor and signal are set on each resistor edge..
One could falsely consider that C10 bootstraps R10 but actually C10 bootstraps itself..

[GibsonGM]

I see! I am also more used to seeing this with tubes, not transistors, but the principles are the same   

[PRR]

The Q1 stage is a 2-pole filter, Sallen-Key or somebody. 3-pole counting the R-C on the output.

[Fancy Lime]

Yes indeed, C10, together with R10, C11, R11, and Q1 is a second order low pass filter of the Sallen-Key topology with (close to) Butterworth characteristic. Togetwith the additional first order low pass R13 and C12, this makes a very primitive cab simulation. Or maybe more of an "ice-pick frequency remover".

D1 and D2 are indeed there to protect the opamp from a specific mode of distortion, which the TL072 is prone to. When you overload the input, it can, as Antonis described go into phase reversal. Not something I would want in a nice mild overdrive. This does not mean that the opamp does not clip at all. At higher gain settings, the output op the opamp may clip quite a lot. But the output section of the TL072 clips nicely. Only the input overload sounds nasty.

@Antonis: No, R15,16 are not 10M but M10, which is a less common way of writing 100k. But thinking about it, these should probably rather be 10k and C16 should be 100u or so.

[Lino22]

So D1 and D2 could be grounded instead of being connected to VR?

[antonis]

So D1 and D2 could be grounded instead of being connected to VR?

I could say just a big "No!!" (for single supply) but I'd like you to think about it a little bit more..

[FiveseveN]

Here's a hint: what's the voltage at pin 3 of the op amp and why?

[Lino22]

Considering there is no current through R3, we have Vref (a half DC) at pin 3. So the LEDs are set between the same potentials and don't lead any current. Ah, there we go.

[antonis]

Here's a hint: what's the voltage at pin 3 of the op amp and why?

You *&%%& cheater..

P.S.
@Lino22: Should it be more clear if Andy has placed those diodes across pins 3 & 2..??
(after all, op-amps 1st golden rule must be obeyed ..)

[Chillums]

Can either use DIP-8 package (on the top side) or SOIC-8 package (on the bottom side) but not both....actually....hmmmm.  Everything else is pretty standard.  Enjoy!!
















Not sure if these are to scale because when I try and attach a .PDF file and it automatically converts it to .PNG.  Anyone know how to attach other files like .ZIP or .RAR?  Can't attach the GERBER files either....errrrr!!  If anyone wants them drop me a PM with your email and I'll get them to ya as soon as I can.

EDIT - Just realized that I put all the traces on the top layer which will make it difficult to solder if etched at home.  I've added v1.01 with all of the traces flipped to the bottom side.  You will still need the top side as its the VCC power plane.  I suppose you could skip the top layer and run flying wires to all the VCC points but I wouldn't recommend it.