Hi!

I want to build this audio mixer circuit. This willl be used inside a digital organ build with stm32 mcu. The DAC will output the organ sound and the drums sound.

I would like some feedback from you on how i could improve the design and on what i'm doing wrong.

 

Just a couple of things that I can see.

1) Where the +15V and -15V enter the board, there should be 10 μF / 25V aluminum electrolytic capacitors.

2) At every IC, there should be two 0.1 μF ceramic disc capacitors, between pin-4 and GND and pin-8 and GND.

These are simply common circuit design practice to eliminate LF and HF noise on the supply lines.

 

Could wire the input stages as buffer amps eliminating 3 resistors.

 

What is the purpose of the three inverting, gain-of-one op amps at the input, as their input resistors are all 47kΩ, the same as the summing op amp input resistors.
Typically those amps would only be needed if you wanted a higher input resistance.
All they would seem to do here is add the op amp intrinsic noise to the signals (with a noise gain of two).

Why not just connect the inputs directly to the summing amp?

Edit: sghioto beat me my a second.

 

You could eliminate U1A, U1B, U2A, R1, R2, R3, R4, R5 and R6 with no change in operation.

 

Could wire the input stages as buffer amps eliminating resistors R1-R6.
View attachment 329047

Needs a resistor between pin 3 and ground.

 

Why not just connect the inputs directly to the summing amp?

As suggested by crutschow.

 

Needs a resistor between pin 3 and ground.

Done, thanks.

 

Why not just connect the inputs directly to the summing amp?

A problem with circuit #7 (with either an inverting or non-inverting amplifier) is that it makes assumptions about the three source output impedances that might not be true. This could lead to very different gains of the signals through this system.

Also, I'm not a big fan of the high-impedance summer approach. But if you are going to go that route, I'd add three voltage followers at the input as in circuit #3.

For better summing action, you could make the three input buffers inverters, so they could drive a *true* inverting summing amplifier for no phase reversal through the system. HEY, that sounds like circuit #1 - !

ak

 

Didn't notice that sghioto changed the inverting summer to a non-inverting summer.
I agree that the inverting summer is better.

Phase-reversal is not an issue for this circuit.

 

Phase inversion - ?

 

Phase inversion - ?

What?
It's the term you used--

so they could drive a *true* inverting summing amplifier for no phase reversal

Why do you now question it?

 

Thanks a lot! Based on your suggestions i kept the original design and added the filtering caps.

Also, if i want to atenuate the signal on organ and drums, putting R3 and R4 as 22k would result in a gain of 0.46. Is that correct?

 

I would put pots on R3, R4, and R5 so that you can adjust the gain on each channel.

 

What?

I think we have a failure to communicate. When you said " Phase-reversal is not an issue for this circuit. ", I thought you were correcting my use of the term.

 

Thanks everyone!

Not the best looking pcb, made with nailpolish, but does the job.

 

Next time, try Sharpie marking pen. It acts as resist also.

 

I think we have a failure to communicate. When you said " Phase-reversal is not an issue for this circuit. ", I thought you were correcting my use of the term.

No.
I used phase-inversion instead of phase-reversal, without thinking there was any difference in the two phrases.

 

The reason for having the input amplifiers is certainly to add isolation between the different signal sources, in addition to allowing the adjustment of amplitudes. If you have ever listened to "music" that included a noise-gate drum synthesis signal, it would be clear that there can be interaction that is unfortunate.

 

Phase-reversal is not an issue for this circuit, U2B is also reversed, so the output is in phase again.