BOOST.. with a built in impedance buffer

k1rkyd · Post by **k1rkyd** » 31 Mar 2011, 17:57

For some time i've been thinking... I usually have a boost at the end of my chain of effects.
So why not build a buffer (KLON-impedance buffer) with half of a TL072 and use the other half as a BOOST!!
As it's at the end of the chain, it's a great combo to regain the lost treble and have that crystal clean tone if you want it and a boost for solo's etc. So i got the schematics of the two that i liked and put them on a 3PDT with a toggle to add in the buffer if you want to!
Below is a HUGE image of the project i have put together. I hope you enjoy.

The vero board layout isn't verified as i have only just finished. It would be great, if anyone found any mistakes, please let me know.

Enjoy,
K1RKYD

phatt · Post by **phatt** » 01 Apr 2011, 08:48

Hi chum,
Simple rules to use when designing simple pedals;
R1 is a garzillion miles to big. unless you wish to attrach space dust from a distant planet then 500k is fine.

R7/8 need only be 10k otherwise you are just asking for more noise.
Again a simple rule is *1k per volt* So a 9 Volt supply only needs 10k Resistors for that divider.

Also R2 can be as low as 220k without massive loss.
More than 680k is inviting buzz/hum/fizz and all those things that will just make it ever so much harder to control when plugged into a 200 watt monster Amp.

Stop hanging 1 or 22meg resistors off everything and you will eventually learn that the only thing that benifits from such sillyness is *NOISE*.

Also make C4 100uF the bigger the better within reason.
Make C5 = 2u2 or there abouts and up R9 to 10k.
Then you will have a fighting chance at defeating the noise that would otherwise destroy the effort you have gone to to make it all.
Cheers, Phil.

k1rkyd · Post by **k1rkyd** » 01 Apr 2011, 09:19

phatt wrote:Hi chum,
Simple rules to use when designing simple pedals;
R1 is a garzillion miles to big. unless you wish to attrach space dust from a distant planet then 500k is fine.

R7/8 need only be 10k otherwise you are just asking for more noise.
Again a simple rule is *1k per volt* So a 9 Volt supply only needs 10k Resistors for that divider.

Also R2 can be as low as 220k without massive loss.
More than 680k is inviting buzz/hum/fizz and all those things that will just make it ever so much harder to control when plugged into a 200 watt monster Amp.

Stop hanging 1 or 22meg resistors off everything and you will eventually learn that the only thing that benifits from such sillyness is *NOISE*.

Also make C4 100uF the bigger the better within reason.
Make C5 = 2u2 or there abouts and up R9 to 10k.
Then you will have a fighting chance at defeating the noise that would otherwise destroy the effort you have gone to to make it all.
Cheers, Phil.

Hey cheers for the reply!
Well i just copied the MXR Micro amp..
This was the schematic just on the internet. The buffer is from the KLON which was also found on the internet!
I just used these two! Can you please tell me again from the vero layout and Bill Of Materials so i know which schem we're looking.

Thanks buddy,
K1RKYD

k1rkyd · Post by **k1rkyd** » 01 Apr 2011, 10:02

Stupid thing sorry i'm in a world of my own!
Can i ask, how do you know how to make these changes? Why does it have to be them certain components that make the CLEAN BOOST.

Cheers,
K1RKYD

mysticwhiskey · Post by **mysticwhiskey** » 01 Apr 2011, 11:15

A few things: C5 on your vero doesn't look right - one end doesn't connect with any other components. The signal should be routed through this cap on its way to R9 but it bypasses it completely. Same with R7 - one end isn't connected to any other components. Also you've labelled both switches as 'Switch 1'

To help avoid confusion, I would also suggest changing the component names on the schematics to match the ones you've got on the layout.

mictester · Post by **mictester** » 01 Apr 2011, 13:38

k1rkyd wrote:Can i ask, how do you know how to make these changes? Why does it have to be them certain components that make the CLEAN BOOST.

You complained about noise - Phil has just suggested simple remedies that will make your project hiss less. The changes he suggests are just simple, obvious, basic electronics. To keep the booster clean, you just need to keep the impedances right and keep the signal levels well below the supply rail voltage, so the op-amp doen't "run out of voltage"

Never fall into the trap of thinking that the designer of a commercial pedal actually knows much about what they're trying to design - many of the "designers" just cut-and-paste older designs and fiddle the components to suit what they have in stock! For example: I remember seeing a whole lot of pedals coming from Electro Harmonix for a while with 1M8 pull-down resistors in them because they must have bought a million of them, and wanted to use them up!

earthtonesaudio · Post by **earthtonesaudio** » 01 Apr 2011, 14:15

You could put an SPDT on the microamp and be able to turn it into the Klon buffer, more or less.

merlinb · Post by **merlinb** » 01 Apr 2011, 19:01

k1rkyd wrote: Can i ask, how do you know how to make these changes? Why does it have to be them certain components that make the CLEAN BOOST.

There are two fundamental rules for low noise design:
1: Resistors in series with the signal should be made as small as possible.
2: Resistors that shunt the signal should be as large as possible.

I think phatt may be a little confused, so I'll go over what he said in more detail for you.

"R1 is a garzillion miles to big. unless you wish to attrach space dust from a distant planet then 500k is fine."
R1 is an anti-pop resistor that lets the input cap charge. It's value is not that important as long as it's big enough not to load the pickup. Typically 1Meg would be a minimum value, and 10M is commonly used. 22Meg is indeed huge, but it doesn't need to be changed.

"R7/8 need only be 10k otherwise you are just asking for more noise."
These resistors set the Vbias, and they don't affect noise because any noise they produce is dumped to ground via C4. Admittedly they don't need to be as big as 100k, but they don't especially need to be changed.

"Also R2 can be as low as 220k without massive loss."
R2 is used to set the input impedance. Pickups need to see at least 470k before they start to show signs of tone sucking, so that sets the minimum value. 1M is most common, and 220k is, I think, much too small. 10Meg is indeed enormous, although isn't really a problem in this case since virtually no current flows in it (FET opamps have virtually no input current).

"Make C5 = 2u2 or there abouts and up R9 to 10k."
C5 should be made large enough to pass all audio frequencies, so that the output impedance of the pedal remains low down to below mains (hum) frequency. Making it small is liable to increase noise pickup outside the pedal. 2u2 is also weird value... 15uF is fine.

R9 protects the opamp from cable capacitance, but because it is in series with the signal it should be made small as possible. 100R is typical, and 470R is also fine. 10k will make noise worse. R10 is the anti-pop resistor for the output, and could easily be increased to 100k; no need to load the opamp more than necessary.

Also, I don't know whether you thought of this, but a booster pedal like this one is ALSO a buffer (one that can amplify too).

k1rkyd · Post by **k1rkyd** » 01 Apr 2011, 19:54

It's funny you should say that because i was thinking about the inverting half joining to the output!
But it is the same as the impedance buffer??

merlinb · Post by **merlinb** » 01 Apr 2011, 20:10

k1rkyd wrote:It's funny you should say that because i was thinking about the inverting half joining to the output!
But it is the same as the impedance buffer??

Pretty much any opamp set up in non-inverting mode is/can be a buffer. In this case both circuits are basically the same, except the MXR has a gain pot added.

phatt · Post by **phatt** » 02 Apr 2011, 15:31

Hi,
Sorry for confusion, I was reffering to the *Boost circuit ONLY* ,, not the layout or buffer circuit below.

noise is often a big issue but If I try to explain it all I'll likely add confusion.
So rather than blab on here is how I would do such a trick.

I doubt you will have any noise issues with this, Winky.

If you need a better schematic I'll draw it up,, kinda run out of time tonight.
Phil.

merlinb · Post by **merlinb** » 11 Apr 2011, 15:14

phatt wrote:Hi,
I doubt you will have any noise issues with this, Winky.
If you need a better schematic I'll draw it up,, kinda run out of time tonight.

That circuit has a lot of unnecessary noise added. What's with the 10k-10k voltage dividers after each opamp? Also, the first opamp isn't actually doing anything; you could feed the input straight to the second opamp.

phatt · Post by **phatt** » 11 Apr 2011, 15:44

merlinb wrote:
phatt wrote:Hi,
I doubt you will have any noise issues with this, Winky.
If you need a better schematic I'll draw it up,, kinda run out of time tonight.
That circuit has a lot of unnecessary noise added. What's with the 10k-10k voltage dividers after each opamp? Also, the first opamp isn't actually doing anything; you could feed the input straight to the second opamp.

Well you go build 2 units,, one your way then build mine.
clue, consider what a buffer does?

You have to lower imp before you up the gain otherwise you just create a noise Amplifier stage.

Hi Z inputs with high gain is a very common mistake.

Phil.

merlinb · Post by **merlinb** » 11 Apr 2011, 16:13

phatt wrote: clue, consider what a buffer does?
You have to lower imp before you up the gain otherwise you just create a noise Amplifier stage.

The first opamp is just a buffer; it has a high input impedance. The second opamp also has a high input impedance (just the same as the first one!), so you can inject the signal straight into the second opamp. In this case the first opamp does nothing (except add some noise from the opamp itself, which is negligible). ALso, you have some 10k-10k dividers which just attenuate the signal while adding their own noise.

Hi Z inputs with high gain is a very common mistake.

I think you are confusing this with an inverting amplifier. Trying to get high input impedance with an inverting amplifier will add lots of noise because the resistors are in series with the signal. With a non-inverting configuration like you have, there doesn't need to by any resistance in series with the signal, it goes straight to the + input of the opamp, and the feedback resistance can be quite small so it doesn't add much noise.

earthtonesaudio · Post by **earthtonesaudio** » 11 Apr 2011, 16:54

This is another way of doing what you want, with fewer components and perhaps better performance:

And again, this time with a buffered Vref, some other tweaks:

Note the max gain of 11 is not very much, but generally it will be clean. If you want higher gain I would suggest forgoing the DC coupled feedback path.

Also I agree with merlinb on his analysis of the other circuit.

phatt · Post by **phatt** » 12 Apr 2011, 08:30

Quite obviously some of you folks just don't do a lot of *Real World circuit
testing*?

Just build the classic TS9 circuit and delete the humble BJT buffer and run the
signal directly to the opamp. Replace that 10k input with a 1meg R,,, Have fun with
the resulting noise. Winky.

Even though I have done this many times I'll be a mug and test it all,, yet again.
Just came back from the shed and worked it both ways.
I then simulated the results for you.

1/ The circuit with no buffer is as noisy as hell.

2/ Added the buffer as shown and the noise is no issue at all.

So Go figure?

Sadly you will have to do the hard yards if you want to hear it for yourself as
simulations don't do actual sound very well. LOL

Sorry if I'm not much good with math’s (It's only a hobby for me) but I have worn out 3 bread boards over 30 years muckin about with all these supposedly wonderful text book circuits that just plain Don't work.

Some text book stuff is smack on but you gotta read between the lines for the hidden
tricks.

Rather than me trying to Ezzplain it all, go look at *Mesa Vtwin* schemo.

Now delete all those buffers that the casual observer might see as pointless.
Get back to me with your noise results. Winky.
You will have one hell of a noisy circuit without those buffers after the tone
stack. (Note the use of 10k/10k Voltage dividers)

From memory there was some good comments on noise issues in *Art of Electronics*.
Even after lots of reading such books it took me many years to grasp the hidden
secrets of getting around noise problems.

All that one line *Fundamental Rules* stuff comes from the HiFi school of thought
where everything is an *Ideal* circuit.
Sadly Guitar circuits are worlds away from the ideal fantasyland of that flat 20/20
bandwidth when using DC coupled Low Z circuitry. (So perfect, beyond belief)

That *Perfection* is part of the reason why SS circuits often sound so crappy for guitar.

It's not so much a SS problem but more about circuit design which has more to do with those things called Resistors/Capacitors than the mojo of matched Valves or exotic opamps with zero noise figures.

Even if your opamps where made by God,, if the circuit design is crappy,,,
well you just have to wear the noise/hiss as well.

@ *k1rkyd*
A Word of advice;
Go get yourself a breadboard and build yourself a few circuits before you commit to
stuff. In the long run you will get there faster and you will also KNOW WHY it
works. (or the more likely outcome,,,does not work??? bugga)

Side note;
Notice how the buffer dramatically improves the low freq response even though I have not altered the components. You don't need to use buffers but if you want max performance and lower noise,,,use buffers.

Another small subtile point the use of series R does help wipe off extreme high freq which in most cases will be more pleasing to the ear.
The closer you get to pure DC coupled HiFi type circuitry the worse it becomes for guitar. to much high freq just destroys great OD guitar sound.

@ *earthtonesaudio*
Sorry I don't get it?

I assume you have used the second part of an 8 pin opamp to improve the bias voltage but the circuit is still potentially noise prone. Where is the benefit in that?

Anyway I've done my bit. I've retested what I already knew to be so.
What others do is up to them.
I have no doubt there are better circuits than mine but I've not as yet found them.

Cheers to all, Phil.

benobo · Post by **benobo** » 12 Apr 2011, 08:52

phatt wrote:Quite obviously some of you folks just don't do a lot of *Real World circuit
testing*?

Just build the classic TS9 circuit and delete the humble BJT buffer and run the
signal directly to the opamp. Replace that 10k input with a 1meg R,,, Have fun with
the resulting noise. Winky.

Even though I have done this many times I'll be a mug and test it all,, yet again.
Just came back from the shed and worked it both ways.
I then simulated the results for you.

1/ The circuit with no buffer is as noisy as hell.

2/ Added the buffer as shown and the noise is no issue at all.

So Go figure?

Sadly you will have to do the hard yards if you want to hear it for yourself as
simulations don't do actual sound very well. LOL

Sorry if I'm not much good with math’s (It's only a hobby for me) but I have worn out 3 bread boards over 30 years muckin about with all these supposedly wonderful text book circuits that just plain Don't work.

Some text book stuff is smack on but you gotta read between the lines for the hidden
tricks.

Rather than me trying to Ezzplain it all, go look at *Mesa Vtwin* schemo.

Now delete all those buffers that the casual observer might see as pointless.
Get back to me with your noise results. Winky.
You will have one hell of a noisy circuit without those buffers after the tone
stack. (Note the use of 10k/10k Voltage dividers)

From memory there was some good comments on noise issues in *Art of Electronics*.
Even after lots of reading such books it took me many years to grasp the hidden
secrets of getting around noise problems.

All that one line *Fundamental Rules* stuff comes from the HiFi school of thought
where everything is an *Ideal* circuit.
Sadly Guitar circuits are worlds away from the ideal fantasyland of that flat 20/20
bandwidth when using DC coupled Low Z circuitry. (So perfect, beyond belief)

That *Perfection* is part of the reason why SS circuits often sound so crappy for guitar.

It's not so much a SS problem but more about circuit design which has more to do with those things called Resistors/Capacitors than the mojo of matched Valves or exotic opamps with zero noise figures.

Even if your opamps where made by God,, if the circuit design is crappy,,,
well you just have to wear the noise/hiss as well.

@ *k1rkyd*
A Word of advice;
Go get yourself a breadboard and build yourself a few circuits before you commit to
stuff. In the long run you will get there faster and you will also KNOW WHY it
works. (or the more likely outcome,,,does not work??? bugga)

Side note;
Notice how the buffer dramatically improves the low freq response even though I have not altered the components. You don't need to use buffers but if you want max performance and lower noise,,,use buffers.

Another small subtile point the use of series R does help wipe off extreme high freq which in most cases will be more pleasing to the ear.
The closer you get to pure DC coupled HiFi type circuitry the worse it becomes for guitar. to much high freq just destroys great OD guitar sound.

@ *earthtonesaudio*
Sorry I don't get it?

I assume you have used the second part of an 8 pin opamp to improve the bias voltage but the circuit is still potentially noise prone. Where is the benefit in that?

Anyway I've done my bit. I've retested what I already knew to be so.
What others do is up to them.
I have no doubt there are better circuits than mine but I've not as yet found them.

Cheers to all, Phil.

Nuff said !

earthtonesaudio · Post by **earthtonesaudio** » 12 Apr 2011, 15:02

The entire electrical engineering discipline has just been schooled by a guy with 30 years of breadboarding and no "maths".

Bravo sir!

[/sarcasm]

merlinb · Post by **merlinb** » 12 Apr 2011, 19:48

phatt wrote:Quite obviously some of you folks just don't do a lot of *Real World circuit
testing*?

Side note;
Notice how the buffer dramatically improves the low freq response even though I have not altered the components. Y
Cheers to all, Phil.

Phil, you seem to be deeply misunderstanding some circuit principles.

Firstly, the bass response of the 'buffered' circuit is better because you have a different time constant at the output; the buffer is not causing this. So you have altered the components- your circuit has a 22k in series with the output while the other circuit has only 100R. If you had used a suitably large cap in both circuits then the bass response would be the same, as you can see in this figure where I have copied your circuits but used a 10uF output cap:

: Freq_Response.jpg (147.8 KiB) Viewed 1897 times

As for noise, a fundamental rule that you should know is that attenuating the signal ALWAYS worsens the signal-noise ratio. Your circuit attenuates the signal not once but twice, and by quite a large amount! This has a massive noise penalty. The buffer does nothing to improve this. As you can see in this figure, your circuit is about four times (12dB) more noisy than the standard version.

: Noise_response.jpg (149.44 KiB) Viewed 1897 times

Buffers can help you make quieter circuits, but only because they allow you to use small resistances without loading down the source. But in this case the source gets the same load with either circuit; your circuit just adds some attenuation which does add noise.

phatt · Post by **phatt** » 14 Apr 2011, 13:31

Thanks Merlin, whoops I fluffed that 10uf cap.

I do appreciate your input.

I see the point and I get the noise graph but maybe something else is going on.
My problem is I don't have the skill to Explain a lot of it.

But heck I'll try

Your graph is showing noise *Inside the circuit* now unless I'm mistaken this does not account for external issues, both before and after. Maybe if I use the word *Sensitivity* might be a better approach.

Most of these circuits plugged into a Low tape/line input you will hardly notice any background noise but even the cheapest of Guitar Amps these days have Hi Z, high gain inputs and by implimenting even more gain via an external unit the sensitivity of the whole circuit is raised to a point where the noise becomes a living nightmare.

Here is where I'm coming from.
Take your best low noise noninverting circuit and hang a hi Z circuit on the front. (a hi Z tone stack comes to mind).
Now start uping the gain to make up for the loss and my guess is the dreaded hiss will become evident.

The only way I found to circumvent this problem was to impliment a buffer to isolate the Hi Z and return it to a much lower Z and THEN apply some gain. I've found this approach works quite well.
Heck it may not be ideal but it made a huge difference to the dreaded fizz and buzz of a hi Z, high gain circuit.
So I'm just sharing what I found by experiment.
Believe me If someone can show how this can be done with one single stage I'm all ears?

When time permits I'll try and setup some more breadboard testing and see if I can find this elusive circuit.
Cheers, Phil.