Hello! I’m building an analog audio mixer for my senior design class in uni and need some help with implementing RC filters for each of the inputs. My audio mixer does not seem to combine or sum the signals as I want after including it and instead gives a wonky output; I’ve included a photo. The mixer has five inputs each with a pot for volume control and a TL074 op-amp that functions as the summer. I also put a lowpass RC filter on each channel for filtering any unwanted noise. I’m lost as to why it isn’t functioning correctly and is giving the output in uV so any help would be appreciated! I assumed it was my capacitor values but after going through many values it always gave the same wrong output. I’m also using PSpice for transient analysis on the circuit.
You’ve got your filters connected to a virtual ground. If you think of a filter as a frequency-dependent voltage divider you can see why that won’t work. They need to connect to a high impedance (instead of zero).
Not familiar with PSpice so I’m not sure what I’m looking at. With a 12 V 20 kHz AC signal going in and R17 set to 1 there absolutely should be the same 12 V 20 kHz AC signal at the wiper regardless of what the filter downstream is doing. Is that 10.97 pV (and the 0V before R17) just the DC component? Have you checked your circuit with a DC input?
Once you get the filters working I suspect you’ll find their cutoff frequency is lower than you want, and if you are putting 12 V signals in then the gain is probably much higher than you want.
I have taken a quick look at your schematic, and have a couple of areas you may wish to consider.
The test signal has an amplitude of 12v, and your summing opamp has a ±12 supply, a TL074 is not able to operate at rail-to-rail voltages, I would expect the maximum output to be in the region of ±10.6v, any output greater than this will be capped.
The gain on your summing amp being Rf / Rin = 10k / 1k = 10. With the input test signal at 12v peak to peak, this would imply an output of 120v peak to peak, far greater than the 10.6v mentioned in the last point, I would expect to see a near square wave output.
As AO said in the post above your passive filters are all connected to a virtual ground, the signal will not try to pass through the shut capacitors (c6 7 8 & 9) to ground, when it is already connected to ground, The input impedance to the summing amp will need to be much greater, say a factor of 100, than the reactance of these. As this is academic I am unsure of any constraints you have, but my first choice would be to use active LP filters, an opamp with an RC network.
As an audio mixer is your intention to use a low pass filter to limit the upper range? if so at what frequency was the intention, the Fc can be calculated using 1/ 2 pi RC. With the current values of 1k and 0.1uF, so attenuating signals above 1590Hz.
As an audio mixer I would have expected a HP filter to block any DC offset.
Your current arrangement will invert the input signal, this is not good practice, you could look into inverting active filters on the inputs, or invert the output with an inverting opamp buffer.
I see your green V marker, I assume you are trying to measure the signal here, as this is the input to an inverting opamp, it is a virtual ground, I would expect this to be a near constant 0v, as confirmed by the 10.97uV .
The input impedance of your mixer, in your design is set by the total value of the input pot to ground, in this case, 1k, is this as intended?
When you build this don’t forget the bypass capacitors from the power rails to ground, The op amp datasheet will give you the requirements.
I hope these comments help, but the first issue to solve is the connecting of the current passive LPF to ground, as this is not a LPF.
