Guitar signal blender/mixer - best practice advice needed

Thread Starter

dbrckovi

Joined Sep 3, 2018
17
Hello!

I am making a guitar pedal which will allow me to blend a clean guitar signal with a distorted/dirty signal from some other pedal.
This is the circuit I came up with:


Pedal blender.png


CIRCUIT DESCRIPTION:
The pedal has 4 signal connectors:
IN - inbound clean signal from guitar
SEND - outbound clean signal which goes to another pedal
RETURN - inbound distorted signal from the other pedal
OUT - combination of distorted and clean signals

The circuit has 3 main sections:
Input section, based around O1
Clean signal mixer section based around O2 and VR1
Dirtiy signal mixer section based around O3 and VR2

Main components:
O1 - unity gain preamp. Makes sure the signal doesn't drop when split between O2 and SEND
VR1 - clean blend potentiometer. Controls the amount of clean signal
O2 - unity gain - sends the desired level of clean signal to output
VR2 - dirty blend potentiomenter. Controls the amount of distorted signal from the other pedal
O3 - unity gain - sends the desired level of dirty signal to output

Other components:
C1, C2, C3, C4 - DC decoupling caps
R3, R6, R9 - biasing resistors which keep the opamps non-inverting inputs near GND level
TRIM1, TRIM2, TRIM3 - trimmers which allow me to fine-tune the gain of each opamp
R8 - raises the input impedance of RETURN input. 10k of VR2 doesn't seem enough when compared to other similar circuits on the net.
R5 - basically a copy of R8. Probably not needed here, but it doesn't hurt.
R4 - protection in case 'SEND' connector is shorted
R7, R10 - prevents the output from O2 to leak out through output of O3 and vice versa
SPDT1 - bypass switch
SW1 - power on/off switch
9V - currently 9v battery, later a standard 9V wall adapter

I built the circuit on breadboard and it's working surprisingly good but I have a few questions before I build it on a PCB:

QUESTIONS:

1.) Schematic has no noise reduction caps yet. Are there any obvious places where you would put them and at what value?
When testing on breadboard, the circuit has a standard 50Hz hum, but based on my previous projects, it will be reduced once I make a PCB and enclose it in a metal box.
Other noise problem which I noticed is that when VR1 or VR2 pots are near the middle position, there is a high pitched hiss.
When pots are either fully "on" or fully "off", the hiss is gone.
Do you have an idea where this noise might be comming from?

2.) Should I add another Opamp in front of SEND, so that signal level which goes to O2 doesn't get affected by input impedance of the other pedal?
I tested it with two other pedals, and I didn't notice any problems, but if there is some 'best practice', I'd like to learn.

This is a hobby/learning project, so I don't expect anyone to spend any serious amount of time on answering this.
I'm just looking for input from someone more experienced, so I can avoid any obvious beginner mistakes.

Thank you!
 

MisterBill2

Joined Jan 23, 2018
18,176
Bypass caps are usually connected from the power supply terminals of the opamp to the system common terminal, "ground".
The hiss is due to the arrangement for gain setting. At the one extreme the gain is unity, at the other extreme there is no feedback and the gain is VERY HIGH. So you have an extreme gain variation. Much better to either run each stage with unity gain, or use the variable resistors as voltage dividers to vary the proportion of output used, and run the op amps with a fixed gain resistor,
OR just disconnect the terminal of the trimmers that connects to the common circuit.
 

LowQCab

Joined Nov 6, 2012
4,023
I don't understand all the complexity ........
By going from the Guitar to an ordinary "Booster-Pedal", ( which can be very useful in any case ),
You now have a "Low-Impedance" Output from the Booster-Pedal that can probably
drive ~4 other Pedal-Inputs simultaneously if You want to.
( more than ~3 Input-Loads may cause the Low-Frequencies to
roll-off to some minor, but unknown extent,
depending upon the internal design of the Boost-Pedal,

but even this is quite unlikely because ANY Input You may be dealing with will be very High-Impedance )

Then all You need is a Switch, and a 1000-Ohm-Pot in a Box,
and a "Y"-Patch-Cord to drive the Effects-Chain, and provide a "Clean" Output at the same time.

In fact, any Effects-Box, or Effects-Chain, is going to have an adjustable Output-Level,
so all You really need is a Switch in a Box that has 2-Inputs and 1-Output.
So, there's no real advantage to having a "Fader".

------------------------------------------------------------------------------------------------------------

A few slight Circuit changes, changing to some High-End Op-Amps, and
some 1/2-Watt, ~1%, Metal-Film-Resistors will permanently eliminate the Hiss-Problem.

Also recommended ........ Run two 9-Volt Batteries instead of a Resistor-Voltage-Divider,
the Batteries will last longer, and the change could possibly eliminate
some Circuit-behavior problems that are not always so obvious at first glance.

Also Recommended ........ Roll-Off the Frequency-Response of every Op-Amp at roughly 8000Hz.
Unless you're actually making a Commercial-Recording with an Acoustic-Guitar and
want to retain Pick, Sting, and Fret Noises, it's just unwanted, and unappreciated, noise.
.
.
.
 

MisterBill2

Joined Jan 23, 2018
18,176
You don't think that running an op-amp at maximum (open loop) gain will lead to some noise?
Certainly the whole thing is more complex than it needs to be, possibly.
 

Thread Starter

dbrckovi

Joined Sep 3, 2018
17
Thank you for responses.

MisterBill2
If I understood you correctly I should put two caps in parallel with R1 and R2?
Or just one between V+ and V-, or all 3, or some other combinations?
The problem is, I'm seeing all kinds of configurations on the net, but have no idea which to use and why.

Regarding the noise problem... The hiss happens when pots VR1 and VR2 are adjusted, not trimmers.
Trimmers are there because when I originally just shorted OUT of an opamp to it's inverting input (trying to achieve for unity-gain), I got less than 100% of the signal on the output. For example... If i's send a 1V sine wave to the RETURN, I'd get around 600mV sine on the OUT. Trimmers are just hacks which allow me to add some small gain to compensate for this. I'll calibrate them once and never touch them again.

LowQCab
I have a fuzz pedal which doesn't have a built-in 'clean blend' knob, so Idea behind this circuit was to mix it externally in a compact box, which can be easily adjusted or turned on or off.
I had a few less complex designs before this, but all of them had problems like: level of clean signal drops as I lower the level of dirty signal, etc. The problems stopped only when I added a specialized opamp for each section.

I'll try all 3 suggestions. Ty
 

LowQCab

Joined Nov 6, 2012
4,023
"" I have a fuzz pedal which doesn't have a built-in 'clean blend' knob, ........ ""

I wasn't referring to a Dry/Wet Knob, or "Clean-Blend", ( never heard of that one before ),
I was referring to an "Output-Level-Knob",
I've never seen a Stomp-Box without one, it's a standard feature.
.
.
.
 

Audioguru again

Joined Oct 21, 2019
6,673
The "wires and rows of contacts" all over the place on a breadboard are interference antennas that guarantee 50Hz hum.

Your opamps have no part number that have datasheets saying how much hiss noise they produce. Some audio opamps have a limited range of gain and can oscillate if the gain is too low.
 

LowQCab

Joined Nov 6, 2012
4,023
Let's start from the beginning ............

1)
What do You normally use as a Power-Source ?
What would You prefer as a Power-Source ?
An actual 9-Volt-Battery, or a Wall-Wort-Power-Supply ?
If it's a Wall-Wort, what is it's actual "Light-Load" Voltage Output ?
Do You have multiple Wall-Worts, one each, for all your various Boxes ?
Do You want to replace multiple Wall-Worts with a single Power-Supply for everything ?

2)
Do You want "everything in one Box", or,
"Individual-Stomp-Boxes" that can be patched together for different configurations / uses ?

3)
Is this project exclusively for Electric-Guitar, or
will it possibly be used for different applications ?
.
.
.
 

Thread Starter

dbrckovi

Joined Sep 3, 2018
17
LowQCab
I understand what you meant, but in that scenario I need a boost pedal and another box which will combine both chains.
I know your suggestion would work and is much simpler, but I won't learn anything if I just buy the boost pedal.

Any way. My main question was about reducing noise and all of you gave me valuable suggestions. I'll try them.

Regarding your suggestion to roll off the frequency at 8000Hz... I'll try it because the circuit does sound slightly brighter, and maybe it will help with the hiss. Did you mean something like a low pass filter?
Is it better to put it in front of the OpAmp?
Low pass filter.png

Behind the OpAmp?
Low pass filter2.png

Or did you mean something else?



Regarding your other questions:

1.) Right now I'm using 9V battery on a breadboard.
The finished box will be powered by a standalone 9V / 300mA wall wart, sold by a music store for "guitar pedals".
It's voltage without load is 9.16V. With 4.7K load it drops to 9.05V
However, my other pedals are powered by another single 500mA adapter through a daisy chain, so I might power my box with it as well.

2.) I wan't everything in one box, but I don't want to redesign the circuit unless it's completely wrong.
In that case, I'd like to understand/learn why it is wrong.

3.) This will be exclusively for electric guitar.
 

LowQCab

Joined Nov 6, 2012
4,023
LM833P is soon to be obsolete, and is rated for a minimum Supply of 10-Volts.
I'll work on some building-block Circuit-Schematics when I can fit it in.
In the meantime, get 25 of these, they're very useful, and in stock a DigiKey ........
MCP6294-E/P-ND

Here's a Electret-Microphone Mixer Schematic to look over and get ideas from .........
.
.
.
Mic Mixer 1 FLAT .png
 

MisterBill2

Joined Jan 23, 2018
18,176
The circuit in post #12 has an error, which is that the common output from sll three channels is accidentally tied straight to the circuit common. OR, maybe that is not a ground symbol, but rather an arrow pointing to adding more channels.
One suggestion is to chose the amplifier IC to use the common pin connection arrangement so that different models of IC can be used.
 

LowQCab

Joined Nov 6, 2012
4,023
The Schematic is Color-Coded,
Green Triangle is Ground,
Pink Triangle (Arrow) is 2.5-Volts,
Black Triangle (Arrow) is the Mix-Buss,
almost any reasonable number of Channels may be added.
.
.
.
 

shortbus

Joined Sep 30, 2009
10,045
Last edited:

LowQCab

Joined Nov 6, 2012
4,023
Here's the final product, ( but that's open to interpretation ).
It contains 3-Sets of Tone-Controls that You can omit if
You don't think you'll use them, but you'll be sorry later if You don't leave them in.

The Guitar-Booster-Amp is quite Hot,
and there's a remote chance that You'll be able to Clip the Amp(s) if You dig-In Super-Hard.
Changing one Feedback-Resistor will easily reduce the ~25X Gain.

This seems like the hard way to do what You want,
but never the less, it will always be a useful tool.
.
.
.
Guitar Mixer FLAT .png
 

Thread Starter

dbrckovi

Joined Sep 3, 2018
17
@ LowQCab

Ty for schematics. I'll use the ideas.

However I have a few questions which seem noise-related:

1.) What is the purpose of the opamp on the 2.5V line (marked blue)?
And why are there two of them in parallel on your original schematic (I guess to average out their noise, but I'm talking out of my a**)?

2.) Why do capacitors in pink have different values and why do they have the values they have?

3.) Why are there so many capacitors between 2.5V line and the ground? And why 100nF? (marked green)

4.) The capacitors and resistors marked in yellow are ~8K low pass filters you were suggesting in post #3?

5.) Why are there two Gain x5 opamps in each channel? Why is there not only one with Gain x25?

Based on my questions so far, can you (or anyone else) recommend some book/blog/resource that would help me better understand concepts in audio circuits?

Thank you,

StompBoxQuestionMarkers.png
 

Audioguru again

Joined Oct 21, 2019
6,673
The input has a 500k resistor to ground plus a 100k opamp input resistor in parallel making 83k which will cause a muffled sound with high frequencies reduced like an old telephone. Electric guitars usually are fed to an old vacuum tube or Jfet with at least a 1M input resistor for enough highs played thru the lousy speakers they use.
 

Attachments

LowQCab

Joined Nov 6, 2012
4,023
1)
The Op-Amp is to create a very Low-Impedance "Supply-Splitter",
this helps to make the overall Power-Supply very "stiff" and stable.
A simple Resistor-Voltage-Divider,
while it "may" work just fine under "most" circumstances,
is a cheesy and cheap way to create an "Artificial-Ground" when using a "Single-Supply",
and it can sometimes create hard to diagnose problems.

2)
These Capacitors are within the manufacturers recommended range for
insuring the stability, and performance, of this particular Voltage-Regulator.
The Input- Capacitor ( 100nf or 0.1uf same thing ),
would be generally assumed to be a Ceramic-Capacitor because of it's small value ( 1uf or less ).
Ceramic-Capacitors have much better High-Frequency-Performance,
as opposed to Electrolytic-Capacitors,
which can sometimes have limited High-Frequency-Performance.
This insures that the Regulator will have very quick response to any demand changes,
and also blocks any RFI ( Radio-Frequency-Interference ).

3)
This is just generally a "thing" of mine .........
These 100nf Ceramic-Capacitors can cover-up problems caused by
poor Circuit-Layout,
excessively long Wires between components,
RFI-noise getting into the Circuit,
etc., etc.,
and help to prevent "Cross-Talk" between Components,
and other odd performance problems that
normally are just over-looked, or un-noticed,
until something doesn't work as expected.
So, as a general rule, I like to put them everywhere that seems like it could be beneficial
since there's no down-side to them.
The only place where they are absolutely REQUIRED is one 100nf Cap connected
as closely as practical to the Power and Ground Pins on each of the Op-Amps.
This rule applies to ALL Integrated-Circuit-Chips, always.

4)
Yes, they also have a secondary function of preventing
any possible "High-Frequency-Oscillations" from being developed by
any of the Amplifiers under unknown, or adverse conditions.

The High-Frequency "roll-off" is very gradual at only ~6-db per Octave.
This means that,
at ~8khz, the Output will be reduced by ~3db,
and at ~16khz the Output will be reduced by ~6db,
and at ~32khz the Output will be reduced by ~12db,
and since there are 2-Stages, these reductions will be doubled,
this is standard practice in many Guitar related Circuits,
You probably can not detect it,
and a Guitar-Amp can barely reproduce these Frequencies in any case.
In addition to this, you are likely to loose more useful "Mid-High" Frequencies
by using a long, and/or, cheap Cord between your Guitar and Amp.
For this reason, and because of the tendency for Guitar-Cords to
pick up lots of RFI-noise, Hum and Buzz,
I strongly recommend installing a Battery-Powered-Pre-Amp inside your Guitar,
or go with a Wireless-Transceiver set-up.
This will make Audioguru's comments a moot point,
since there will be no highly reactive, High-Impedance, Pickup-Coil
connected though a very long, poorly shielded, Cable.
Also, the problems that he warned against are mostly applicable to
certain Hot "Single-Coil" Pickup designs,
a "Humbucker"-Pickup-Coil is not likely to have any trouble with Amplifier-Input-Impedance.
The higher the overall Input-Impedance gets, the more NOISE You will get.

5)
This Amplifier-Circuit doesn't have a problematic amount of Gain, ( at only 25X total Gain ),
but, when feasible,
there can be certain advantages to "splitting the Gain" between 2 or more Amplifier-Stages.

Reducing the Gain from 25 to 5 X 5 tends to reduce Amplifier-Noise, and increase Fidelity,
( neither of which is actually a huge concern with an Electric-Guitar type of application,
but it's generally a good practice to reduce the Gain when practical ).

Also, for the crude Tone-Controls to work,
they must be "Buffered", on both their Inputs and Outputs.

Also, having 2 Amplifier-Stages makes it possible to double the
effectiveness of the built-in Filters.

Also, since the most advantageous Amplifier-configuration turns out to be
an Inverting-Amplifier-configuration,
using 2 of them, in series, makes the Output-Signal the same polarity as the Input-Signal,
although, in this application, the Input vs Output Polarity probably won't
make any difference to the operation of
any of the features of the Box, or anything connected to it.

Keep in mind that,
if You have the time, expertise, and the proper test equipment,
to test and test, and re-verify repeatedly, and then test some more .........
You could probably build this setup with less than half of the parts,
but it wouldn't be nearly as stable or bullet-proof, or quiet,
still, it might "work" reasonably well.
But, You are not building this Box to be as cheap as possible for mass-production.
Since I want this project to work for You, I've put in as many "safeguards" as I know how,
this naturally makes it appear to be more complex
than what is strictly required for basic functioning.
.
.
.
 
Top