Hello everyone,

While working on the latest 6502 project, I found a way to create two square waves for audio purposes. I want to mix them into a single channel. I searched the forum for "audio mixer", you would be surprised how little is there. I went looking online and found this:

https://electronics.stackexchange.com/q ... nput-audio

With a bit more searching, this led me to trying things out in falstad.com, particularly this:

https://www.falstad.com/circuit/circuit ... bBvSoBThAA

Which I recreated with the attached picture. If both square waves are being used at once, I get an amplitude of -1V to +1V, with 0V when they are conflicting. That is the basic "adding" principle. But when I only have one square waved being used (the other is just grounded) I get an amplitude of -0.5V to +0.5V, never at 0V because there is never a conflict.

This isn't too bad, but it means each is only 'half volume' by itself.

I went looking further and found folks using op amps to do a better job:

https://www.electronics-tutorials.ws/opamp/opamp_4.html

But the more I look, the more I see the need for an negative voltage to be applied to the op amp (I'm super new to this, sorry), and often +9V and -9V, which is definitely not my simple SBC 5V target here. I also tried using a comparator in place of the op amp, but that sorely did not work, as they work on a different principle entirely.

Do any of you have recommendations on how to combine 2x 5V square waves better than what I currently have? In a *simple* way?

And, while I'm at it, are there any modern new-stock sound chips that pair well with the 6502? Something that I could buy from Mouser, let's say?

Thank you for any assistance!

Chad

The 'half volume' thing is inevitable if you use a simple mixer. If you add two signals of two different frequencies with equal amplitudes, the power of the combined signal will be double the power of each signal individually. To avoid that you would need to use a variable gain depending on whether one or both signals were active.

I'd suggest something that's active... you can use either FETs or older bipolar transistors. Just configure them as emitter or drain followers. You can take the input signals, drive them into 10K or 20K potentiometers, the output of those will drive the FETs/Transistors and you simply sum the outputs of the followers. There will be a bit of gain loss in the followers, so you can always add another gain stage after it.

I don't have the schematic handy, but back in the 1974, I designed a simple mixer for 4-channels with discreet transistors (2SC945) with trimpots for audio input level. It all fit on a small perf board and ran off a single polarity supply.

_________________
Regards, KM
https://github.com/floobydust

_________________
http://WilsonMinesCo.com/ lots of 6502 resources
The "second front page" is http://wilsonminesco.com/links.html .
What's an additional VIA among friends, anyhow?

This is normal Chad. Even with op-amps buffering each input, the mix is the sum of the input signals. Rather than think of each signal being at "half" volume, think of it as that is their full volume. You will get much more amplitude when both inputs are high or low together. You'll get the sum.

_________________
Bill

Thank you all.



I looked into that, and tried it out on the simulator and seems to work as expected. Definitely a good suggestion, thank you.



That was the deciding factor for me. I see that a 100K bleedoff resistor doesn't let C3 stabilize very quickly, so I changed it to 10K and it seems to work ok in the simulator. Then again, I don't think I can trust the simulator because when I add C2, the thing starts freaking out at random times. Lastly, for sake of whoever will read this in the future, I changed the 4.7uF cap to 10uF and it seems to give me a more squared wave, not a psuedo-triangle-thing.

Thank you very much for this Garth, I appreciate it greatly.



Ok, that is good to know. I'm so new to analog, and even more new to audio mixing, that I don't even know what the goal is sometimes. I just think, "Can't I just stick them together?" And yes, kindof, but the result is as expected: Waves that interfere will cancel, waves that combine will add. Simple, but I guess I didn't understand that simple was intended.

Thank you all very much, I'm going to be using Garth's circuit. More updates on the 6502 project soon to come!

Chad