Fundamentals of designing your own overdrive?

[PBE6]

Yes that's connected correctly now.

On a side note, I wonder if the output might need a buffer or a gain makeup stage? It might be fine, I think the gain of your BJT stage is ~47k/1k = 47 = 33 dB, and the tone section will remove something like 15 dB, so you still have 18 dB left. See what happens.

[canman]

You know, I hadn't even thought about modifying the circuit...I just wanted to use this as practice for understanding vero/schematics a bit better and see if I could understand pedal circuits more.  I'll for sure try what you suggested and see what happens!  I have used this one on its own with a Triple Wreck clone, it seems to be OK without the buffer/makeup stage but it could be even better with one. 

I suppose this is as good a circuit as any to breadboard and insert different stages and see how it all works together!

On a side note, I had somewhat of a fun idea the other night but I'm not sure what the best approach would be.  I was thinking of splitting the signal and sending it to two parallel gain stages...one that clips hard, and one that clips soft (I believe the Plimsoul has something similar?) and both have a gain control.  Then both clipped signals would blend together and continue through the rest of the circuitry (tone, makeup stages, whatever else sounds interesting).  That way you could turn off hard or soft clipping, or blend the two for different overdrive flavors.

So a) would this work?  And b) can you split the signal and then blend it back together without any kind of buffer before the gain stages, or could you just amplify it with a recovery stage afterwards? 

[PBE6]

I was thinking of splitting the signal and sending it to two parallel gain stages...one that clips hard, and one that clips soft (I believe the Plimsoul has something similar?) and both have a gain control.  Then both clipped signals would blend together and continue through the rest of the circuitry (tone, makeup stages, whatever else sounds interesting).  That way you could turn off hard or soft clipping, or blend the two for different overdrive flavors.

So a) would this work?  And b) can you split the signal and then blend it back together without any kind of buffer before the gain stages, or could you just amplify it with a recovery stage afterwards? 

a) yes this would work, provided you don't invert one of the signals an odd number of times. If you do, you'll get excessive cancellations when you blend them back together.

b) there are lots of ways to split the signal without a buffer, but whether you need one up font or not really depends on what you're feeding the signals into. High impedance loads like opamps won't affect the upstream signal noticeably, but if you're feeding one of the signals into a relatively low impedance load like a BMP tone stack it will affect things upstream.

[canman]

So if you were to use two non-inverting gain stages in parallel and blended them again afterwards there wouldn't be any excessive cancellation, right?  I suppose the problem with that is you can't get to unity gain.

If I were to use one inverting and one non-inverting, and only inverted once, I'd have cancellation correct?  Would the best approach to use 2 inverting opamps, invert each one once, and then blend them?  Or would both being inverted one time cause cancellation?

I didn't even know you could invert an opamp more than one time..!

[PBE6]

Yes, in an ideal world if you took a signal, duplicated it, inverted one of the copies, then mixed them back together, you would get nothing at the output because they would cancel each other out. In your case, each signal starts the same but then distorts differently, so the cancellation would not be complete, but it would still sound very odd (but probably have an octave-up quality, which is cool!).

One thing to keep in mind is that different distortion methods will give you different final volumes. Hard clipping (Dist+ style) will cut off any signal at the conducting diode voltage, while soft clipping (Tubescreamer style) will give you a voltage that is the sum of the input signal plus the conducting diode voltage. If you use a blend knob, the even blend position will be off-centre. Not a big deal, just something to keep in mind.

As far as splitting options go, two that come to mind are sending the signal to two different opamps and using a phase splitter (essentially just a BJT with identical collector and emitter resistors) whee you tap both the collector and emitter signals (although the collector signal will be inverted, so it will need to be flipped back).

[teemuk]

So a) would this work?  And b) can you split the signal and then blend it back together without any kind of buffer before the gain stages, or could you just amplify it with a recovery stage afterwards?  

"Yes" to both.

But how sensinsible arrangement it is can be pondered: Unless your clipping circuits introduce strikingly different harmonic spectrums (e.g. asymmetric vs. symmetric) or frequency responses blending between two distortion modes becomes more or less sub-optimal arrnagement. Yes, there are more higher order harmonics in hard clipped signal than in a soft clipped signal. ....but that's practically the only difference. This actually makes even following scheme work so marvellously that the concept has been employed in dozens of distortion circuits: Blend in totally clean signal with a ridiculously distorted one (e.g. practically square wave-ish). Signal gains must match so if you clip by amplifying attenuate later in respective ratio.

If you think about the arrangement the blend pot basically does nothing but adds or removes harmonics from the signal as it is blending between two signal contents where only difference is amplitude of harmonics. If you blend between two distortion modes you, IMO, just have a "gain" pot that has it's range limited from the "cleaner" end. You could make that pot go from pristinely clean to maximum distortion setting as well and basically the "intermediate" gain modes would be equally accessible, just within slightly different range of the dial.

Here's one quite famous example of blending between clean and highly distorted signal paths:
http://www.prowessamplifiers.com/schematics/images/klipp_60.jpg
Note how the blend pot substitutes a traditional gain control. If both signal paths are "unity gain" the blend pot only controls amplitude of harmonics and doesn't increase overall signal amplitude level like a traditional "gain" pot would.

[canman]

Interesting analysis...I had to read it a few times to really pick up the meaning, haha.  Basically, it's not really worth it to have parallel gain stages like I'm thinking.  I'm totally OK with that.

BUT...what if they were in series?  Would having soft clipping followed by hard clipping make more of a difference?  In a way I guess this is more or less the same as having two overdrive pedals in a pedal chain, now that I think about it.  Which means it might be worthwhile?

The good news is I just ordered a little breadboard (400 point, hope that's big enough) so I'll be able to start experimenting here shortly!!

[MaxPower]

I have an electronics mag which has a distortion circuit/project which has hard clipping followed by soft clipping. Try it both ways and see which sounds better.