4
\$\begingroup\$

I'm trying to design an audio mixer for my friend, and he only needs three stereo channels with volume for each channel as well as a master volume.

Additionally, he'd like as high an audio quality as possible in a small form factor. I'm looking at putting this in a guitar pedal case and running it off of 9 V power.

From an electrical standpoint, is there anything wrong / anything I can do to improve this design?

Audio Mixer Schematic

UPDATE:

After reading over the answers I can see how many mistakes I made. After fiddling about on a simulator to get to a point that I liked this is what I ended up with. Updated mixer schematic

The main changes that I have made are to

  1. increase the 1k resistors to 500k
  2. change the DC bias to go directly to the non-inverting op-amp input instead of getting mixed into the inverting input
  3. move the decoupling capacitors to be after the volume pots on the input channels
  4. increase the volume pot sizes
  5. increased resistor size to ground on output

Here's the explanation for why I did each one, or at least how I understand it

  1. having such a small resistor going into 'virtual ground' forms an RC high-pass filter with a very high cutoff frequency, removing a lot of the signal. By increasing those resistors that effect is mitigated
  2. This is how a DC bias for an op-amp configured as an inverting amplifier should be done, I just didn't know what I was doing
  3. I'm not really sure why this makes such a difference, but I think it's that having them before the pots creates a high-pass in combination with the volume pot, so it ends up like a variable cutoff high-pass, which I definitely don't want sitting on my input channel
  4. Not sure exactly why this made the difference it did either, but it seemed to lower the cutoff of the high-pass filtering on the output when I was simulating, plus these parts are cheap on DigiKey
  5. this lowers the cutoff frequency of the output high-pass so I'm not removing so much of the signal

Let me know what I've done wrong this time, thanks for all the help!

\$\endgroup\$
7
  • \$\begingroup\$ I know you're having fun. But don't burn your time & money. Just buy something. LIke this: musiciansfriend.com/pro-audio/rolls-mx28-mini-mix-vi/… \$\endgroup\$
    – Kyle B
    Commented Aug 24, 2022 at 7:13
  • \$\begingroup\$ Besides the obvious errors pointed out already, your device is missing specs what it must work with so it is uncertain if it fits the specs and works as intended. With +/- 4.5V on TL074 it may not be able to drive any reasonable amplitude to load on output. Audio can be 2VRMS or 5.65Vpp. \$\endgroup\$
    – Justme
    Commented Aug 24, 2022 at 7:27
  • \$\begingroup\$ Did you calculate the cutoff frequency of c3 rv1 r1 and r4 at both ends of the pot? \$\endgroup\$
    – Colin
    Commented Aug 24, 2022 at 8:26
  • \$\begingroup\$ @KyleB When I click on your link, it brings up a $200 microphone. \$\endgroup\$
    – HandyHowie
    Commented Aug 24, 2022 at 8:42
  • 3
    \$\begingroup\$ Upvoted for a well-presented question including a clear fully-labelled schematic. It makes it easier to discuss (even if there is much to discuss). \$\endgroup\$
    – Graham Nye
    Commented Aug 24, 2022 at 12:41

2 Answers 2

4
\$\begingroup\$

It's a nightmare!

With those large pots and small input resistors you will loose most of your signal between the two. Reduce the value of the pots by a factor of 10 and increase the values of the 1 k input resistors, feedback resistors and input capacitors.

Remove R4 and R11 and replace them with an ac coupling capacitor in series with the inverting input of all 4 op amps.

Connect the non-inverting inputs of all 4 op amps to the 4.5 V reference.

Add a capacitor in series with the output of U1a and U1b.

I haven't checked the time constants (cut-off frequencies).

EDIT

That's much better but still not quite there.

You could try the circuit below in a simulator to get a better understanding of dc levels/biasing etc within the circuit.

I have changed your component values to be reasonably sized and added a capacitor either side of the pot between the op amps to block dc and couple ac.

It would probably work ok without C8 but it's not a good idea to have dc current flowing through the pot, it can make it crackle.

I have specified the three input capacitors as non-polarised (bipolar).

You could play about with the values of R5 and R6 to adjust the gain but bear in mind, that with a 9 V supply, the TL074 will saturate at about plus and minus 3 V either side of mid rail (4.5 V). So it is a matter of adjusting the gain to avoid saturation when the pots are wound to maximum and considering the maximum amplitude of the expected input signal. You may find that you need to increase the supply voltage above +9 V or reduce the gain.

When designing a circuit like this, the 3 main factors to consider are gain, frequency response and input and output resistances (how much signal you expect to loose driving a certain input impedance from a certain output impedance).

Mixer

\$\endgroup\$
4
  • \$\begingroup\$ Thanks for the suggestions! I spent the day on LTSpice working out the issues you brought up, and ended up on a schematic that I think should work better. I fiddled around the the RC values until I ended up with something that didn't seem to affect the signal too much. I've added the updated schematic to the post. \$\endgroup\$ Commented Aug 25, 2022 at 6:33
  • \$\begingroup\$ ok, that makes sense. I have the option of using a different supply voltage, this would allow me to have less gain reduction on the opamps so that they could still operate without saturating, right? How would the schematic change with a 12 or 15V supply? \$\endgroup\$ Commented Aug 25, 2022 at 18:26
  • \$\begingroup\$ @LucasHenry Yes, higher voltage power supply means that you can have a larger signal amplitude before saturation (clipping). So with a higher voltage power supply you could increase the gain a little (increase values of R5 and/or R6) before the onset of saturation. Whether or not you need to this depends on the amplitude of your input signal and how large you need the amplitude of the output signal to be. There's no point in increasing the power rail voltage and gain to then reduce back down with the pots to limit the output amplitude to cater for a receiving device such as a power.... \$\endgroup\$
    – user173271
    Commented Aug 25, 2022 at 19:07
  • \$\begingroup\$ ..... amplifier. Increasing the power supply voltage wouldn't require any changes to the schematic unless you wanted to increase the gain. \$\endgroup\$
    – user173271
    Commented Aug 25, 2022 at 19:09
2
\$\begingroup\$

From an electrical standpoint, is there anything wrong / anything I can do to improve this design?

  1. Trying to bias an op-amp with R4 and R11 as shown is never going to work. Apart from anything else, it will try and force the op-amp output to produce a negative voltage and, because the power rails are limited, this is an impossibility.
  2. It doesn't look like you understand the nature of a virtual ground on an inverting op-amp and the effect R1, R2 R3, R8, R9 and R10 are going to have on the input impedance and bass response.
  3. Ditto for the volume stages following the front-end mixers
  4. The common mode input range for the TL074 does not include the 0 volt rail in fact, to get that op-amp to work correctly, requires a DC offset from the negative supply pin (0 volts) of about 3 volts.

I'm looking at putting this in a guitar pedal case right now and running it off of 9V power.

"Right now" is not going to work. I think you need at least a day to take in what I've said above and modify your circuit to make it operational. I suggest you also invest some time in learning how to use a simulator.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.