Modular Multichannel Stereo Audio Mixer

A modular mixer where it's easy to add more channel strips.

Public Chat
Similar projects worth following
I want to create an easy to expand stereo sound mixer.

My goal is to learn all the necessary steps and find all problems when creating a whole project. This includes creating the schematics, testing on breadboards, send to create the PCBs, populate the PCBs, 3d printing the enclosure and final real use.

This uses a lot of op amps for sound amplification and buffering of the signal.

The main idea was to create individual PCBs for each input, another one for output and a final one for the power supply. All this connected by a common bus on a stripboard.

In the future if I want to add more inputs is to populate another PCB and add it to the stripboard.

This uses a very basic mixer design.

There is a basic bus (stripboard) with two audio signals (left and right), for the mixing part, then other two lines for the output. Finally there are three power lines: ground -V and +V. Input and output are joined in this prototype, but at the end I explain the reasoning behind. It makes a total of 7 lines. There are some decoupling capacitors in this bus to fix some buzzing and provide consistent power.

Some special attenuation would be needed on the monitoring input.. i guess. It needs some prof of concept.

Each input first passes to a potentiometer to control the input gain, then to DC bias removal and finally to an op amp used a buffer. This finally passes though a resistor to create the proper minimal resistance. Then it just goes to the input line on the bus.

The output goes to a DC offset remover, and then to an opamp in a non inverting configuration. This goes then a potentiometer and finally to the output jack.

And all this is done twice in parallel to allow for stereo.

In the input modules the jack allows a mode in which the right channel is not connected and it just copies the left signal, allowing for mono inputs.

At the output I tried to create a mode that if no right output jack is connected does a mixdown with mixed results. Were it working properly it could double as a 8 mono input mixer, but with per double chanel volume control.

For the power I decided to go USB C PD, 20 v. this gives a lot of headroom for internal voltage. But it introduces some noise. I'm waiting for a LM7815 to limit the voltage but with better noise.. I hope.

Problems encountered while developing:

  • Breadboard was ok, but some extra noise, so I tried on a protoboard.. but I always messed up something so decided to go PCB. Amazing. I should have done it before. Perfect quality, super low price, and only some waiting.
  • Of course first version was wrong. So v1.2 was born... with more care on the lines and the logic.
  • USB C PD noise. A lot. Everywhere. I tried to sprinkle a lot of capacitors around, some low pass... Finally the noise level is not bad, but now I'm waiting for a LM7815 that I hope will make it better.
  • Maybe I over engineered the idea of two stereo audio lanes, but as the minimal order is 5 PCBs. and adding some more is for free... I have a lot of empty PCBs waiting to be populated.
  • The printed case did not account for all the component when the PCB was populated. Finally some pliers made it work. Also I would try a snapping case next time. Although I had to use rubber bands to keep it together but it helps with the grip of the mixer to the table! so happy accident!

Future improvements:

  • The idea to have separated input and an output lines was to be able to create a monitoring configuration. Delaying the union of both lines to after an op amp buffer allows to have a separate output for monitoring and another input for monitoring. The monitoring output doe snot mix the input for monitoring, but the final output does. I've seen on some professional mixing desks and it's very useful. Anyway this configuration is not applied in this prototype. This would be the configuration:
Input | Input | Input | Output | Buffer | Input | Output 
   |_____|____|_____|_____|________ |> ______|______|
  • Maybe add more lanes to allow a button or switch to send individual inputs (or a mix of them) to be sent to an alternative output. This can be useful for recording.
  • It would be nice to have another op amp directly at the input to avoid a possible voltage divider with other connected outputs. This already happened to me using a patch bay. Same at output.
  • It would be nice to have balanced inputs and outputs.
  • Lower power noise. DONE!
  • Maybe there is no need for 100k potentiometers. Actually I don't even know what difference does it make. For sure shorter ones, as the ones I got are super long.
  • Overdrive indicator.
  • Equalization.
  • Now it's using USB PD at 20v. Maybe there is some way to do it at 5v with the same...
Read more »

fcstd - 186.82 kB - 05/29/2023 at 18:49


Zip Archive - 4.37 MB - 05/29/2023 at 18:49


  • 6 × TLC272ACP Amplifier and Linear ICs / Operational Amplifiers
  • 5 × Potentiometer Audio 100k
  • 7 × Pin headers x7 Male
  • 1 × USB C PD module
  • 7 × Pin headers x7 Female

View all 11 components

  • Frequency Response

    David Moreno Montero06/13/2023 at 19:02 0 comments

    I was really bothered about the frequency response, so I decided to take some measurements.

    I used REW (Room EQ Wizard), and got a very flat response for a direct cable.. but a (I think) quite bad for my mixer :(.

    (blue is -48 dBFS, pink Phase)

    I will try to get more adequate Op Amps, and maybe tweak the dc coupling. 

    Also somehow I think I do not output enough volume... and I used 15 volts for that.... so maybe I will have to redesign it all?

    Suggestions are appreciated!

  • Removed USB power noise.

    David Moreno Montero06/02/2023 at 08:30 0 comments

    I received my set of 78xx linear regulators and did some fast test to check if it removes the USB power noise.. and it does!

    So I set together a fast protoboard to try it live with the full mixer, and the noise level is now non existant! Amazing!

    Now I have another problem to solve and is that I'm filtering out low frequencies. Quite probably is the DC bias removal. Need to investigate.

View all 2 project logs

Enjoy this project?



Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates