Opinions About Boost Design

[ElectricDruid]

If this boost is being controlled by a foot pedal, it's unlikely that it uses the full range of the pot anyway, so even designing it to "go to zero" probably won't ensure that it does.

My preferred approach would be to use a single dual op-amp, with the first one set up as a non-inverting boost, and then swap the pots over as Merlin suggested, so that the foot pedal is now controlling the volume (the pot to ground) rather than the gain (the pot in the feedback loop). The second op-amp can then go after the volume control as a buffer to drive whatever follows and give the whole unit a nice low output impedance.

In theory this would be able to give you zero (or as close as pots get, before Antonis steps in to correct me!) if you turn the volume control right down, but since the pot is being waggled by a foot pedal, we probably won't get all the way down. Though you might be able to tweak it to get close, I'd be cautious about adjusting the pot so the pedal pushes it to the end of its track - you're stamping on this with your foot and you could easily shove the wiper straight off the end of the track and bust the pot if you push it too far.

[spacecommandant]

Brian at Madbean responded and drew up a schematic which was similar to what you're all describing and he added an optional passive/active switch at the input..
I also decided to add a tuner output.
I tried it and it works great but I decided to take a shot at the circuit again using 5532 OpAmps instead of TL072.
I changed the inverting stages to non-inverting and added 100r resistors at the +inputs and outputs to help avoid stray capacitance, which is something I've read to do with the 5532.
I'm not positive what the best value would be for C4.
I haven't breadboarded this yet, so I'd love to hear any opinions on this version...

Thanks again for everybody's input!

[antonis]

NE5532 input bias current (worst case scenario) could be as high as 1000nA (1μA) and create a bias offset a high as -1V..!!
(due to NE5532 n-p-n BJTs input..)
This could be OK for IC1_A, IC_B and IC2_A (always dependent on input signal amplitude) but might create issues for IC2_B gain stage..

Despite the above mentioned bias offfset, I'd place a cap between R10 & GND (to roll-off DC gain to unity) and make R16 12k..

P.S.
C4 value can be as high as 1nF (16kHz LPF corner frequency for MAX pot set FCW)
R8 & R9 seem usless here..

edit: Didn't realize 18V supply so plz ignore all about bias offset but DO implement the R10 series cap..

[spacecommandant]

NE5532 input bias current (worst case scenario) could be as high as 1000nA (1μA) and create a bias offset a high as -1V..!!
(due to NE5532 n-p-n BJTs input..)
This could be OK for IC1_A, IC_B and IC2_A (always dependent on input signal amplitude) but might create issues for IC2_B gain stage..

Despite the above mentioned bias offfset, I'd place a cap between R10 & GND (to roll-off DC gain to unity) and make R16 12k..

P.S.
C4 value can be as high as 1nF (16kHz LPF corner frequency for MAX pot set FCW)
R8 & R9 seem usless here..

edit: Didn't realize 18V supply so plz ignore all about bias offset but DO implement the R10 series cap..

Thanks for that.
I should have also mentioned I'll be adding a 1uF tantalum and 100n film from both of the 5532 V+ and V- pins to Vref, also something I read as being a good idea to do.
If I calculated correctly, a 10uF should be a good size for the R10 series cap?
I'm also curious why R16 should be 12k. I trust your judgement, I'm just not knowledgeable enough about circuits yet.
Thanks again for your input!

[merlinb]

5532 is a suboptimal choice for this circuit, stick with the TL07x. What's the bypass switch for?
IC1_B and IC2_A appear to be redundant, you could do everything with just two opamps (one IC).

[ElectricDruid]

As Merlin said, don't use the 5532. Don't add 100r's everywhere. And get rid of two op-amps which aren't doing anything useful. That'll save you more noise than any change to 5532 might be supposed to do.

The circuit needs brutal pruning to get it down to something more realistic. The R7/R8/R9 combination is a giveaway!

[antonis]

I should have also mentioned I'll be adding a 1uF tantalum and 100n film from both of the 5532 V+ and V- pins to Vref, also something I read as being a good idea to do.

The addition of those caps is indeed a good idea but the exact placement is a very bad one..
Vref should be kept as clean as can be..
(C12 takes care of this and we don't want it to bear down on extra burden..)
ICs supply decoupling should be implemented directly between V+ and V- or between V+ and GND (in case of single supply, like here..)

If I calculated correctly, a 10uF should be a good size for the R10 series cap?

If 12Hz corner frequency is OK with you..
In general, input and feedback HPFs corner frequencies tend to be equal but there are many applications where the later corner frequency is set higher than the former one for a steeper slope..

I'm also curious why R16 should be 12k. I trust your judgement, I'm just not knowledgeable enough about circuits yet.

R19 (not R16, typo..) should be 12k to raise Vref for voltage drop across 1M bias resistors counteracting, in case of NE5532 implementation....
(but this should stand for relatively low Vref values in conjunction with signals of high amplitude..)

[antonis]

I think OP should be happy with something like below: 

[FiveseveN]

Almost I think. The purpose of a tuner output is to stay on even when the main output is muted (while tuning, hopefully) and as this is a volume pedal, a muting switch is redundant. Still not clear on what the original switch was supposed to do but it wasn't a footswitch, right?

[spacecommandant]

The switch was added as a toggle to choose between active/passive input direct from a guitar but at this point I've decided to actually ditch it and have it be strictly active.

[spacecommandant]

If I eliminated the switch, what do you think about R5 and C5 values?
EDIT: I just looked again and realize R1 and R5 together act as parallel resistors.... so eliminate R5?
I'm also wondering how the volume pot sweep will react when this is placed after other pedals. As I understand it, 250k and 500k volume pots are generally used in volume pedals with the intention of being plugged directly into the guitar, with 25k pots often used in volume pedals to be used after other pedals.

[antonis]

There is no reason for both R1 & R5 existence..
(either can serve as pull-down resistor for both C1 and C5..)

[spacecommandant]

Thanks

[spacecommandant]

I appreciate everybody's input.
Would the addition of a non-inverting buffer at the beginning of the circuit be an advantage in some situations? I'm just trying to anticipate any issues if this is used at the end of a chain of pedals.

[GibsonGM]

Nah.  You could test the theory out by breadboarding this, then putting any Boss pedal before it - they are buffered when bypassed, so would show what that would 'sound' like.  I think no difference, personally!  No advantage to it, the boost input impedance is already high.

[antonis]

Would the addition of a non-inverting buffer at the beginning of the circuit be an advantage in some situations?

Only for Input signal sources of high output impedance..
(more than 25k for IC1_A and more than 50k for IC1_B, following the 1:10 rule of thumb..)

But, as Sir Mike said above, go ahead..

[spacecommandant]

Sorry for a million questions but I'm trying to wrap my head around this.
Would you see any advantage in changing the volume pot to an A500k?

[antonis]

Higher input impedance for IC1_A but higher noise level, also..

If you really worry about Boost circuit input impedance, place a buffer in front of C1, make Volume pot 10k or so and oversize C1..

But, once more, let it be as it is..

[GibsonGM]

Nothing 'teaches' about impedance and mismatch effects more than playing with your guitar 'toys' like coiled patch cords, fuzz faces & wahs, the wrong pots installed in a guitar...but still coming up with something you like. So then you use your rig in bands or whatever for ages. Maybe later, you learn that it was ALL WRONG, that you had TONE SUCKING - but you LIKED IT!  LOL   What should have sounded horrible somehow sounded GOOD...it can happen.     

I hate TOO high an input impedance. Leave that for piezos.  I often play my LP into a Fender amp thru a non true bypass wah that is "off",  because I WANT it to suck some highs out.  Same for the curly cord.  Sometimes I load tube stages too much to warm things up.  Mess around, do it wrong, do it right...find out what makes your sound what you want it to be   There isn't a 'right' formula for this other than the general rules of 'hi input impedance, low output impedance'...that is usually what will sound good. 

Breadboard your boost, and see if you like the sound.  Put a buffer or buffered pedal out front, listen again   Then decide if you want a buffer in there permanently, or something you can switch in, or none at all.  Chances are good that everything else you use will follow 'the rules' and play nicely with your boost...

[spacecommandant]

Yeah, I'm really happy with the way this sounds... very grateful for the advice.
I also can't hear a difference when I plug the guitar directly in or if there's a buffered pedal in between, so I'm not going to worry about it.
One more question, though (I apologize but this is fascinating learning about this)!
When the Max pot and Volume pot are turned all the way down, the pedal is basically dead quiet... except when I'm using my guitar with fairly hot humbuckers... the signal is faint but noticeable.
My passive Ernie Ball volume pedal doesn't pass an audible signal when it's all the way down.
What would be the approach to attempt to completely eliminate the signal in this circuit when the Volume pot is at 'zero?'
(BTW, I'm turning the pot with my hand and can confirm it's rotated fully)