Op amp Boost. Help needed

[Buffalo Tom]

Hi.

Im trying to build this:


Using this power supply:


So by combining those two circuits I came up with this schematic.
But it does not work. Cant get no sound out of it. What am I missing?
Im using a 10K linear pot and TL072 op amp with pin 4 connected to negative 9VDC input jack.

[kusi]

first, don't label your 4,5V as GND. label it Vref Vbias or similar. GND should be the negative pin of the dc jack to avoid confusion.

R4 should go to 4,5V, not to 9V.
R1 should to GND not to 4,5V
R6 should go to 4,5V not to GND
R5 should go to GND not to 4,5V
Pin4 should go to GND

[antonis]

I think you're confused a bit..

Vref should only exist in case of single supply powered op-amp.

If you intend to power your op-amp with bipolar supply (+/- 4.5v) you have to connect 4.5v (red mark) to bipolar supply GND.
(1M bias resistor should remain there to form a DC return path..)

If you intend to power your op-amp with 9V single supply, bipolar circuit isn't needed..
(delete 2 X 1μF caps, place a 10-47 μF cap between + & - pins of 9V and rename GND point as 4.5V & -4.5V as GND..)

P.S.
We then can talk about IN cap size and another cap in series with 1k resistor to rolloff DC gain to unity..

[GibsonGM]

Woah, bud.   All you need to do is build the first (top pic) version of this boost, with the right power supply.  What you've shown for a power supply is 'bipolar' - you don't need that here.  The 'real' way it's done is just a little different.

Go HERE https://www.electrosmash.com/tube-screamer-analysis

Find "Power supply" in a box drawn in green, under "1. Schematic" .   Reproduce that.  Make sure the IC is connected to 9V and ground also (pins 8 & 4).  It should work as originally drawn.  Of course the jack grounds also go to...ground (which is battery "minus", same as at node below C16).


Look at the pic at top of your original post....connect the 4.5V points to Vref ("4V5" there).

Welcome to the forum, Tom!

[merlinb]

You combined them wrong:


This works with a battery or other dedicated floating power supply, but it is not the normal way to build a stompbox since it won't work with a shared power supply.

[amz-fx]

[Buffalo Tom]

Thanks everyone for your replies! Wow this place is really helpful. I will rewire the power as suggested.

Another thing I need to understand is the gain/volume control. My intention is to have unity gain at noon. So when the potentiometer knob points at 12 the pedal should just act as a buffer with no volume change. Then by turning full right I want +20 dB boost and by turning full left I want -20 dB attenuation.

On the AMZ variable gain schematic I guess that the 1k resistor and the 14k pot decides the gain/volume.
How do I balance the resistor and the pot to get what I want?

Thanks again!

[GibsonGM]

I'd try an audio taper pot, 10k or 20k, whatever you can easily find, and wire it up in place of the 14k.   The 'taper' of the pot's resistance will allow more adjustment space in the pot's travel 'down low', giving you more control on the "minus" end.      You could get fancier using tapering resistors, but I wouldn't go to all that trouble, unless you really have a need to.   

You could even put POTS on either side of that 14k to set the RANGE it operates in, if you like...overkill?  That's up to you! 

[amptramp]

The minimum gain with the circuit you have shown is unity with the pot shorted out.  With the pot at maximum resistance, you have a gain of 15.  You cannot get a gain below unity with this circuit.

[ElectricDruid]

Another thing I need to understand is the gain/volume control. My intention is to have unity gain at noon. So when the potentiometer knob points at 12 the pedal should just act as a buffer with no volume change. Then by turning full right I want +20 dB boost and by turning full left I want -20 dB attenuation.

On the AMZ variable gain schematic I guess that the 1k resistor and the 14k pot decides the gain/volume.
How do I balance the resistor and the pot to get what I want?

It's a non-inverting op-amp configuration, so you *can't* get a gain less than unity. That's just a result of the way the gain equation works.

https://www.electronics-notes.com/articles/analogue_circuits/operational-amplifier-op-amp/non-inverting-amplifier.php

The only thing you could do would be to add a potential divider on the output that provides -20dB of attenuation. Then the *overall* effect at unity gain would be 0dB from the amp and -20dB from the divider = -20dB overall. If your 12 o'clock position gives you 20dB of gain from the amp, you'd have unity overall, and then you'd need 40dB at the maximum gain position.

Ok, so -20dB is x0.1. So a potential divider with two resistors of 9K and 1K would work. But 9K isn't a standard value, so you'd go with 9K1 and 1K. That gives -20.1dB, which is definitely close enough.

Moving the pot wiper all the way to the right gives a zero value for the pot resistance (since it's shorted out by the wire to the wiper, and the wiper is basically connected to the op-amp output. That gives x1 gain for the op-amp. So how would we do +40dB?

40dB is x100 (I looked this up: http://www.muzique.com/schem/gain.htm)
Our gain equation is Rf/R2+1, which in our case is Rpot/1K + 1. So if we had a 100K pot at maximum, we'd have a gain of 100K/1K+1 = x101. Again, this is very close. And a 100K pot is an easily available item, whereas a 14K pot is rarer than unicorn poo.

So that gives us some values to get us into the right ballpark. There are some reasons I might not do it exactly like that, but I'll let others step in at this point. Antonis wanted to tell you about avoiding problems with DC gain and adding a DC blocking cap on the input, which are both excellent ideas... Antonis?

HTH,
Tom

[PRR]

> turning full right I want +20 dB boost and by turning full left I want -20 dB attenuation.

As said: this form does not work this way.

You "can" leave the opamp at gain of, say, 10. Put an Audio taper pot (as potentiometer) in front. "Noon" on the pot will be 5:1 or 10:1 attenuation. Then 1/10 * 10 is unity gain, full-up is gain of 10 or +20dB, full down is infinite cut but a little end-resistor can trim to 1/100 * 10 or gain of 1/10, -20dB.

The problem is that there's always a gain of 10 boosting self-hiss, and the pot value is a compromise between source loading and added hiss. It works good on already pre-amplified stuff of known high level: Hi-Fi systems. In guitar chains it may be less satisfactory.

67 years ago a guy named Peter devised a +/-20dB tone control using linear taper pots.
http://www.learnabout-electronics.org/Downloads/NegativeFeedbackTone.pdf
You steal his plan but omit the tone-select caps. His Fig 5 shows the basic bass-only network. The tube and resistor above are "an op-amp" for practical purpose; TL072 is a heap cheaper (but not around in 1952).

The drawback is that the input impedance is low and goes very low for large boost. You *need* a buffer in front. In tubes, this was An Expense. A TL072 has a second half so is a natural.

Here is the basic plan *without* the DC networks needed to put the opamps in their happy-zones. Yes, while the extremes are pretty-near +/-20dB, the quarter-points are not exactly 10dB. There's not a simple hack for this. Live with it. In most applications we call this "An Advantage": finer resolution around zero dB, but plenty of gain/loss on tap at the ends.



The "1.1k" values are overly fussy. Pot values are usually +/-20% so 10% fixed resistors are not meaningful; I just did it for the ideal math. 1K will be fine unless you *need* full 20dB gain.

[Buffalo Tom]

Here is the basic plan *without* the DC networks needed to put the opamps in their happy-zones. Yes, while the extremes are pretty-near +/-20dB, the quarter-points are not exactly 10dB. There's not a simple hack for this. Live with it. In most applications we call this "An Advantage": finer resolution around zero dB, but plenty of gain/loss on tap at the ends.

The "1.1k" values are overly fussy. Pot values are usually +/-20% so 10% fixed resistors are not meaningful; I just did it for the ideal math. 1K will be fine unless you *need* full 20dB gain.

Thanks! This seems like a great approach!
Im not sure how to connect the non invert inputs on U15 and U16.
They are both labeled 0. Should they be connected together and used as "guitar input"? 

[antonis]

Im not sure how to connect the non invert inputs on U15 and U16.
They are both labeled 0. Should they be connected together and used as "guitar input"? 

No..!! 

First op-amp is used for "buffering" purpose (it offers power gain) where its Output is used as GND (due to its negligible impedance..) for R291, pot & R292..

Note: As it is, is a DC amp - meaning it can't be used for AC signal without DC rolled-off to unity..
(it equally amplifies AC+DC signals..)

About non- inverting inputs labeled 0 : Probaly your scheme refers on real ground (not virtual..) for a bi-polar power supply..(didn't read Paul definitions..)
(or, maybe, I didn't get your query clear enough..) 

[Buffalo Tom]

Here is the new schematic. And it works! I was not sure what to do with the non invert input B.. But when I connected it to +4.5V (VR) everything started to work. This is my first pedal. Am I missing any parts? Can the circuit be improved? Next thing is to get the switching part done. Plan is to use the buffer on all the time and then switch in/out the boost part. Thanks everyone for your help so far.

[antonis]

Here is the new schematic. And it works! ...... Am I missing any parts?

As far as it works, there shouldn't be any missing parts..

About improvement:
If you could be more specific about what you don't like on your circuit, we could make some suggestions.. 
(e.g. for supply noise/ripple, a low pass RC filter.. - for IC instabilities, a 100nF ceramic cap from very close pin 8 to GND - for high frequency/high gain oscillation, a "some"(depending on Gain pot value) decades pF cap shunting IC pin 7 & Gain pot lug 3 - for anti-pop only(*) purpose of C4, 1M R5 value...)
(*) 1M should also be OK, for an OUT HPF of 1.6Hz cut-off frequency..

[Buffalo Tom]

About improvement:
If you could be more specific about what you don't like on your circuit, we could make some suggestions.. 

Having trouble switching in/out the boost part in this schematic. Here is what I came up with. It kind of work but not really.. There is switch "pop" How can I improve the switching?

[antonis]

Can't get your swtching idea..

When switch is set to NO, buffered signal is boosted..

When switch set to NC, buffered signal is mutted so it can't be used for any other (obvious) purpose..
(it's shorted to DC GND via C3 & AC GND via C2 due to PS very low internal impedance..)
You additionaly drive IC1A to current limiting mode..

If you want to switch buffered signal to OUTPUT (or to another circuit input..) you have to use a DPDT (at least..) for completely by-passing IC1A OUT..
(from pin 1 straight to OUTPUT..)

[Buffalo Tom]

Can't get your swtching idea..

When switch is set to NO, buffered signal is boosted..

When switch set to NC, buffered signal is mutted so it can't be used for any other (obvious) purpose..
(it's shorted to DC GND via C3 & AC GND via C2 due to PS very low internal impedance..)
You additionaly drive IC1A to current limiting mode..

If you want to switch buffered signal to OUTPUT (or to another circuit input..) you have to use a DPDT (at least..) for completely by-passing IC1A OUT..(from pin 1 straight to OUTPUT..)

Thanks.. I have updated the schematic above! But it works just as bad....  Im using an illuminated pushbutton switch its a DPDT but plan was to use the other poles for the illumination. So I really want to bypass the boost part with just 1PDT if possible... But It seems not.

[antonis]

Your update isn't satisfactory enough.. 

You now by-pass booster but you place on buffer OUT a series combination of R8+C4...

Best is to wire SW_NC to C4 positive leg..
(and pray for switching interval time to be small enough for C4 positive leg left "floating"..)

[Buffalo Tom]

Best is to wire SW_NC to C4 positive leg..
(and pray for switching interval time to be small enough for C4 positive leg left "floating"..)

Thanks but did not work. No sound at all with SW_NC to C4 positive leg..