This overdrive is almost perfect (I swear I'll be done soon)

[Transistor-Transistor]

(R11 is supposed to be 390R not 47)
Productive struggle my friends.
I have way too many JRC4558's for some reason so I decided to use one for the buffer and tone control. The old buffer wasn't high enough impedance.
I was playing around with the active tilt EQ and I found that if you change one of the caps you get a nice mid notch so I added a switch to toggle that on and off.
This sounds great the diodes clip the bass really nicely.
only one problemo
It can't be cleaned up. Theres not enough distortion in this circuit to justify the amount left at the lowest possible gain. It's not exactly a priority but I would like added versatility to the gain side of things. Maybe somehow add more distortion and keep the minimum the same or keep the same amount of max distortion and have it clean up nicely. Again, not a priority but I think that's what would kick this up from good to great, so if someone wants to help me with that that'd be awesome.
(note, higher pot values don't really do much)

also interesting:
this thing acts so different with different guitars its insane. High output pickups fuzz this out in a good way and it just sounds like a nice beefy overdrive with standard single coils. I had fun playing around with that last night.

So yay it sounds good, theres no more noise (I may want to add better filtering to the power supply though), and I'm genuinely excited about this design.

Atticus M.

[ElectricDruid]

Not cleaning up properly sounds a bit like a bias problem to me. If it was biased nicely in the centre, it'll clean up because reducing the input level will bring it out of clipping. If the bias is way off to one side or the other, reducing the level will reduce the clipping, but it's still likely to clip on one side, whcih sounds much the same as clipping both sides to our ears. So you don't really get the reduction you're expecting.

Another potential source of the problem is the tilt EQ. If the EQ has a decent signal level going into it (how big is the signal coming off those clipping diodes?) and then you add, what?, +10dB of boost...have you got enough headroom? So you might even have a clean signal that's barely pushing the diodes, but that is then boosted to clip the final op-amp tone stage.

This is why "gain-staging" a design is such a big deal. It completely determines the way a design responds to increasing levels (and that includes the difference between single coils and humbuckers or active pickups). If that response isn't what you want it to be, you'll notice it.

HTH

[Transistor-Transistor]

I can assure you that the EQ is completely clean.

Probably a bias problem. How do I make sure it's right? Transistor biasing has always plagued me 

[ElectricDruid]

I can assure you that the EQ is completely clean.

Ok, jolly good.

Probably a bias problem. How do I make sure it's right? Transistor biasing has always plagued me 

Tweak the sh!t out of it?!? While listening to the output, obviously...

Or you could do what are technically known as "hard sums", but you'll have to ask other people here about that

[m4268588]

https://guitarscience.net/calcs/cecbe.htm
https://guitarscience.net/calcs/ce.htm

[Transistor-Transistor]

https://guitarscience.net/calcs/cecbe.htm
https://guitarscience.net/calcs/ce.htm

Thats perfect thank you

I'll also do some tweaking as well 

[antonis]

Do you know anything about impedance (mis)matching..??

e.g. Q2 Collector resistor of 100k is trying to drive Q3 stage input impedance (which is much lower than 100k) and Q3 output impedance (about 10k) is trying to drive EQ's impedance (which might be much lower, dependent of pot setting..)

So, at first, either follow 1/10 rule of thumb or put a couple of buffers (either BJT Emitter followers or op-amp voltage followes) and then deal with "proper" bias..

[Transistor-Transistor]

Do you know anything about impedance (mis)matching..??

e.g. Q2 Collector resistor of 100k is trying to drive Q3 stage input impedance (which is much lower than 100k) and Q3 output impedance (about 10k) is trying to drive EQ's impedance (which might be much lower, dependent of pot setting..)

So, at first, either follow 1/10 rule of thumb or put a couple of buffers (either BJT Emitter followers or op-amp voltage followes) and then deal with "proper" bias..

I haven't a clue as to what impedance mismatching is. What's the 1/10 rule?

[Transistor-Transistor]

If I made the circuit with the correct impedances would it sound different? (Noob question I know)

[antonis]

If I made the circuit with the correct impedances would it sound different? (Noob question I know)

Definitely, yes..
(ex-noob answer..)

e.g. by considering Q2 stage alone (infinite load impedance) we expect a stage gain from R7 / [1k + (0.026/I_Collector)] up to R7 / (0.026/I_Collector)..

In your circuit, we have to replace R7 with the parallel equivalent value of R7//R8//R9//(h_FE(Q3)xR11. which is obviously MUCH lower than R7 itself..
(resulting into MUCH lower gain, hence lower distortion..)

Same stands with R10 and EQ impedance..

[Transistor-Transistor]

If I made the circuit with the correct impedances would it sound different? (Noob question I know)

Definitely, yes..
(ex-noob answer..)

Thats what I feared. I love the sound it has right now and as much as I'd like to follow rules I don't know if it'll be worth it y'know.

[antonis]

Thats what I feared. I love the sound it has right now and as much as I'd like to follow rules I don't know if it'll be worth it y'know.

If you like it, go on..!!
(but it should be polite, at least, to clearly state at the beginning that you're not amendable to deal with rules/formulas and all those %^*&*(&*^%$ nasty things..)

[antonis]

OK.. Let's proceed more seriously now..
(no explanations, formulas or questions – just obey..!!)

Make R2/R3 10k/10k and add a 22μF on their junction..
Make R4 100k..
(if you look for higher input impedance, you can split R4 and boostrap it via a cap ..)
Make C2 100nF (in case of R4 not bootstrapped..)
Add a 1k to 10k k resistor between C3 and Q2 Base,,
(its value will set Q2 stage gain in conjunction with R5)..
Make R5 470k – 560k..
Make C6 470pF..
Make C7 47 to 100 nF..
Add a 100k resistor from Q2 Base to GND..
Make R6 47R – 56R..
Make Q2 Emitter pot 100R..
Make R7 10k..
Make R1 22R and C1 220μF

P.S.1
DON'T ASK WHY..!! 

P.S.2
Tom, now it's your turn about active EQ configuration / items values..

[Transistor-Transistor]

What if the lowest value pot that I own is 1k...?

(I could put like a 120R resistor across pins 1 and 3 on a 1k pot to make it ≈100 right?)

[antonis]

Ahaaaa..!!!

So, you now DO have questions, no..??

Of course you can do it..
(pot travel/taper altering is currently out of discussion..)

[Transistor-Transistor]

OK.. Let's proceed more seriously now..
(no explanations, formulas or questions – just obey..!!)

Make R2/R3 10k/10k and add a 22μF on their junction..
Make R4 100k..
(if you look for higher input impedance, you can split R4 and boostrap it via a cap ..)
Make C2 100nF (in case of R4 not bootstrapped..)
Add a 1k to 10k k resistor between C3 and Q2 Base,,
(its value will set Q2 stage gain in conjunction with R5)..
Make R5 470k – 560k..
Make C6 470pF..
Make C7 47 to 100 nF..
Add a 100k resistor from Q2 Base to GND..
Make R6 47R – 56R..
Make Q2 Emitter pot 100R..
Make R7 10k..
Make R1 22R and C1 220μF

P.S.1
DON'T ASK WHY..!! 

P.S.2
Tom, now it's your turn about active EQ configuration / items values..

It sounds about the same, although now it doesn't feel like there's as much sustain:(
To be completely honest with the emitter pot on max gain it sounds like how I'd expect to sound around 3 o-clock in the perfect world

[Transistor-Transistor]

Ok ok so I know you said don't ask questions
But-but I have questions

If I were to replace the emitter gain control with a pot in the place of the resistor between C3 and Q2 would that be okay? A pot around ≈50k gets to a decent clean sound.

Also

If I wanted to raise the gain would it be better to raise the 560k resistor in the feedback path or lower the 100k tied to the base?

Also

Thanks for dealing with me

[m4268588]

If you need gain, can you choose to amplify with U1A?
In addition, Ge.diode in NFB reduces the gain by its leak.
If you want to maintain the symmetry of "tilt", you can do this but the change when "tilt" is rotated will be less. However, Symmetry can only if Q3 is not clipped.  When Q3 is clipped, the output impedance is undefined.

It can be improved by inserting a buffer. (Additional resistors are no longer needed)

http://vulcanofx.blogspot.com/p/tilt-eq-calculator.html
If you want the "tilt" to the way it is, leave it as it is without changing anything. C8 = 47n remains the same.

[antonis]

If I were to replace the emitter gain control with a pot in the place of the resistor between C3 and Q2 would that be okay? A pot around ≈50k gets to a decent clean sound.

If I wanted to raise the gain would it be better to raise the 560k resistor in the feedback path or lower the 100k tied to the base?

Answers to both questions above rely on particular stage gain understanding..

Q2 stage is considered inverting CE amp configuration with negative feedback(s)..
1. Any Emitter resistor (Emitter degeneration) lowers gain..
2. Gain is set by R5/Base series resistor ONLY if open-loop gain (R7/R_Emitter) is MUCH HIGER than R5/R_Base..
(and R5 much bigger than R7..)

So, place a low value Emitter resistor (150R or so) with by-pass cap (like C4), retain 560k feedback resistor and make Base series resistor (the one between C3 & Q2 Base) 50k pot (wired as variabale resistor)..

You'll be able to set gain between X11 and high enough(*)
(*) particular gain number calculation requires formulas involving open-loop gain, closed-loop gain, feedback fraction and U1A output impedance..

If you're not satisfied from that distortion level, proceed to Q2 input "extra" gain, like proposed by m4268588..
(just sersch for something like this..)

As for that 100k resistor from Q2 Base to GND, it's there to "stabilize" bias conditions and form LKPF with C3..
(no major contribution to gain..)

P.S.
Also, replace Ge feedback diodes with Schottkies, for leakage current purpose, like also said by m4268588..

[Transistor-Transistor]

If you need gain, can you choose to amplify with U1A?
In addition, Ge.diode in NFB reduces the gain by its leak.

I forgot to mention that I switched to Si diodes long ago 


I probably will use U1A as a little input boost.

If I were to replace the emitter gain control with a pot in the place of the resistor between C3 and Q2 would that be okay? A pot around ≈50k gets to a decent clean sound.

If I wanted to raise the gain would it be better to raise the 560k resistor in the feedback path or lower the 100k tied to the base?

Answers to both questions above rely on particular stage gain understanding..

Q2 stage is considered inverting CE amp configuration with negative feedback(s)..
1. Any Emitter resistor (Emitter degeneration) lowers gain..
2. Gain is set by R5/Base series resistor ONLY if open-loop gain (R7/R_Emitter) is MUCH HIGER than R5/R_Base..
(and R5 much bigger than R7..)

So, place a low value Emitter resistor (150R or so) with by-pass cap (like C4), retain 560k feedback resistor and make Base series resistor (the one between C3 & Q2 Base) 50k pot (wired as variabale resistor)..

You'll be able to set gain between X11 and high enough(*)
(*) particular gain number calculation requires formulas involving open-loop gain, closed-loop gain, feedback fraction and U1A output impedance..

If you're not satisfied from that distortion level, proceed to Q2 input "extra" gain, like proposed by m4268588..
(just sersch for something like this..)

As for that 100k resistor from Q2 Base to GND, it's there to "stabilize" bias conditions and form LKPF with C3..
(no major contribution to gain..)

Oh ok that makes sense.