|
|
Post by mkjackn on Oct 27, 2021 20:40:41 GMT
Hello, Im trying to design a 4 channel audio mixer but am having some problems finding information about key bits. Using pots its possible to attenuate each input channel but how do I ensure the output sum is not bigger that 5v to avoid damaging other parts of the synth? Do I need to use an op amp buffer on each input or will the fact I'm running 4 signals into an op amp (mcp602) to mix them serve the same purpose? I know this is all a bit vague but my analogue electronics is not great an I'm having trouble adapting theoretical designs to the AE world.. Any help much appreciated!!  |
|
|
|
Post by robertlanger on Oct 28, 2021 8:28:17 GMT
I will give a basic schematic here later today!
|
|
|
|
Post by mkjackn on Oct 28, 2021 9:06:24 GMT
I will give a basic schematic here later today! Thanks Robert that would be amazing! |
|
|
|
Post by robertlanger on Oct 28, 2021 9:07:55 GMT
Here you find the core of the 4-4 MIXER module, showing the basic principles. R10/R11 make a virtual ground of 2.5 volts, which is also the "ground" for the pots; C4/R9 make the normalization of the audio sum to the virtual ground level, meaning the summed signal oscillates around the 2.5 volts level. For audio, the first opamp would be sufficient; the second one does only an aditional phase reverse (because the first one is also an inverting mixer) to keep the phase identical to the inputs (which does not an audible difference usually). HTH! ![]()  |
|
|
|
Post by mkjackn on Oct 28, 2021 9:45:32 GMT
Here you find the core of the 4-4 MIXER module, showing the basic principles. R10/R11 make a virtual ground of 2.5 volts, which is also the "ground" for the pots; C4/R9 make the normalization of the audio sum to the virtual ground level, meaning the summed signal oscillates around the 2.5 volts level. For audio, the first opamp would be sufficient; the second one does only an aditional phase reverse (because the first one is also an inverting mixer) to keep the phase identical to the inputs (which does not an audible difference usually). HTH! ![]() <button disabled="" class="c-attachment-insert--linked o-btn--sm">Attachment Deleted</button> Thanks Robert this has answered a lot of the questions that I had! |
|
|
|
Post by keurslagerkurt on Oct 29, 2021 13:53:06 GMT
Here you find the core of the 4-4 MIXER module, showing the basic principles. R10/R11 make a virtual ground of 2.5 volts, which is also the "ground" for the pots; C4/R9 make the normalization of the audio sum to the virtual ground level, meaning the summed signal oscillates around the 2.5 volts level. For audio, the first opamp would be sufficient; the second one does only an aditional phase reverse (because the first one is also an inverting mixer) to keep the phase identical to the inputs (which does not an audible difference usually). HTH! ![]() <button disabled="" class="c-attachment-insert--linked o-btn--sm">Attachment Deleted</button> Thanks for giving a glimpse in your modules! I was wondering if there are specific reasons for using the LMV358 as opamp? I always reside to the MCP60(0)2 in my AE designs. |
|
|
|
Post by leethargo on Jul 10, 2024 7:46:39 GMT
Here you find the core of the 4-4 MIXER module, showing the basic principles. R10/R11 make a virtual ground of 2.5 volts, which is also the "ground" for the pots; C4/R9 make the normalization of the audio sum to the virtual ground level, meaning the summed signal oscillates around the 2.5 volts level. For audio, the first opamp would be sufficient; the second one does only an aditional phase reverse (because the first one is also an inverting mixer) to keep the phase identical to the inputs (which does not an audible difference usually). HTH! ![]() View AttachmentThis is great inspiration for my clone of the "Nearness" stereo mixer. Thanks for the schematic and explanations. I now know the function of most components, but the capacitor C3 to ground (from virtual ground of 2.5V) is still puzzling for me. Is this a way to reduce noise in the output? Most resistor values chosen here are 100K, with the notable exception of R10 and R11 with 4K7 (for virtual ground). Would it not make sense to pick a higher resistance for this part as well, so that less current is "wasted"? Or would that mess with stability/noise in some way? |
|
|
|
Post by keurslagerkurt on Jul 10, 2024 8:23:09 GMT
leethargo The voltage divider has lower resistor values, because otherwise the 2.5V reference can become unstable. You can see it like this: the 'outside circuit' that is connected to the voltage divider (ie: the inputs with high resistors) form a parallel path to 0V or 5V or something else, depending on what you plug in. So depending on what you plug in, the equivalent resistance of this divider actually changes. So in order to keep these changes small, the resistors of these parallel paths have to be large compared to the voltage divider resistor (rule of thumb is: at least 10x larger). People will often refer to this as 'making the voltage divider "stiff"'. Another way to solve this, would be to either buffer the 2.5V reference, or use a dedicated 2.5V reference IC/Zener Diode/... Still, if you calculate the 'wasted current' from the divider, it is still very small (approximate 5V / 10000 Ohm = 5 µA) As of the capacitor, i think this is indeed for noise, to keep the 2.5V more stable (not completely sure). But it makes sense, as you might for example have 4 large voltage spikes coming in the mixer (eg when four drum sounds play at exactly the same time), so this might destabilize the 2.5V temporarily, but the capacitor acts as a kind of supersmall 'battery' counteracting the spikes. Not 100% sure about this though!
|
|
|
|
Post by leethargo on Jul 10, 2024 15:18:36 GMT
Thanks for the explanation, that makes sense to me intuitively.
I still don't have good enough coverage of the fundamentals to reason about these things from first principles. But at least after some weeks of exposure, trying to decompose schematics into functional blocks has become somewhat easier :-)
|
|
