I have been working with the following circuit: enter image description here

to build a audio mixer where you can take multiple line inputs, cd players, computer, phone, synthesizers, etc... and mix them together to a single stereo "master" out.

The "master" out can be a power amplifier or powered speakers. So far tested powered speakers and works good and I expect this to work with "professional" grade amplifier/speakers. I.e: Mackies, Adams, Bryston etc... I will test with as much gear I can get my hands on...

So what I want to do is split the "master" out to 3 outputs: master, room and record and each with their own volumes.

  1. I'm thinking of taking R8A, R8B (the master POT) and replacing with a regular 22K resistor.
  2. Split the stereo signal after R9 and R10 and add 3 22K POTs there
  3. Finally line to 3 separate "stereo" jacks/channels

I'm going to do some testing. But I don't have enough gear to test it all, I'm thinking it will be ok so long the input impedance of the destination devices is bigger enough then the total out?

Does my plan sound sane or do you think I should add buffers to each of the split signal?


Pots are log.

Sample impedances:

Update 2:

Using TL072 as the op-amp.


1) Since users are pretty good at finding the weak point in any design, I'd recommend just using a buffer for each channel (yes that's 6 total opamps but they're cheap). For example, what if a user wants to temporarily disconnect your high impedance equipment to check the sound using their nice low impedance headphones?

2) High output impedance (compared to a low impedance buffered output) will make your transmission line more susceptible to noise, so if you're using high end equipment you're not doing yourself any favors by skimping on the output stage as long as you choose a decent quality opamp.

My recommendation would be to move your master potentiometer to the output of the summing stage (instead of R9 and R10), and splitting that into three different paths to feed each buffer opamp circuit. I think it would be a good idea to put a medium sized resistor (10k??) between the master potentiometer wiper lead and each buffer input too, that way the buffers don't interact with each other.

  • \$\begingroup\$ Cool! Things to consider. For the headphone I supose, but it's XLR. Not once in my life I ever tried to plug my headphones into the master out of my mixing desk of any kind even using some sort of stereo jack to XLR converter. But I get your point. lol \$\endgroup\$ – user432024 Jul 15 '19 at 18:24
  • \$\begingroup\$ Also I don't want the master pot to affect the room. So if you lower the master you can still hear the room and vice versa. Same with record they should simply be separate outputs. \$\endgroup\$ – user432024 Jul 15 '19 at 18:26
  • \$\begingroup\$ Ha isn't that the point of a master control though? (It would be a good idea to include it anyways and just not touch it right?) Anyways, in order to get that much control without adding extra opamps, you may want to consider using non-inverting configurations so you can add potentiometers to the inputs without impact the gain settings. \$\endgroup\$ – Kent Altobelli Jul 16 '19 at 0:38
  • \$\begingroup\$ He, master and room are usually separate volumes. This is common in DJ Mixer. And I just realized I mean to say master/room and booth. Lol \$\endgroup\$ – user432024 Jul 16 '19 at 2:00
  • \$\begingroup\$ My design is modular I can put quite a bit of inputs for now doing 4. instagram.com/p/BzlfgDKnm0o/?igshid=yn9tz6pzfukg Those aren't final pots nor is the master wired yet. Black/green is power the white is the audio. Top right connector is the master out to jacks or it could be another board which can have what is discussed above. Most bottom connectors is where the input jacks will go. \$\endgroup\$ – user432024 Jul 16 '19 at 2:10

Your plan sounds fine, although if the output pots are audio taper, they act as an 80% attenuator at 50% rotation - you might need to increase R8 for higher gain to compensate.

Three 22 K pots in parallel present a 7.33 K impedance to the 47 ohm buildout resistors, a ratio of over 150:1, so the loading on the outputs will be minimal.

Depending on the input impedances of the downstream devices and the output pot settings, the load impedance seen by R9 and R10 might be lower, but you would have to have all three pots at max rotation plus three very low input impedance devices for the loading to be significant.

Note that your plan will increase the apparent output impedance of the circuit, again as seen by the downstream devices connected to the output pot wipers. For example, if an output pot is set at approx.80% rotation (again, audio taper, working from memory), the output impedance will be approx. 11K. If the target amp has an input impedance of 10 K, this forms a 55% (-6.44 dB) attenuator. The net effect is that you might find the pots adjusted a bit higher than expected for the same volume as a direct connection to the original circuit.

  • \$\begingroup\$ I added some sample devices and their impedances to my question. But I think it will be ok. Plus I also expect that each down stream device will be adjusted to the hearing preferences of the user. So for the "master" channel will always the loudest, while "room" less louder for example. \$\endgroup\$ – user432024 Jul 15 '19 at 15:45

To keep it simple connect three log volume control potentiometers to an audio preamp IC output and wire the sliders of the pots to the hot pins of the output connectors. Use low resistance, say 2 kOhm potentiometers and be sure your preamp can feed 600 Ohm well. Use stereo pots for 2 channel stereo configuration.

If you need level setting switches or long cable feed (5 meters or more) ability this is too simple.

ADD: The cable length limit isn't exact, but unbalanced connections have a tendency to catch noise when cables are long and there's several devices with mains voltage supply connected together. The devices have unwanted secondary connections through their power supplies. TL072 should be able to feed those about 1Vpp or less signals to 600 Ohm. The devices cause minimal loads.

ADD2: Just noticed that you have some serious pro audio stuff. Maybe a redesign with proper balanced signals and surely full audio bandwidth amps is well worth the effort. If you can have $2000,-/pair speakers (and all other devices in the same league) which most hobby musicians can only dream of, do not enter into pauperism just now.

  • \$\begingroup\$ It can work, but I have found long unbalanced audio connections collect noise if it is available. It surely is available if you have several mains supplied devices. they make unwanted loops via their power supplies. It's a property of unbalanced lines. I guess you have about 100 mV signals and high impedance inputs. \$\endgroup\$ – user287001 Jul 15 '19 at 15:50
  • \$\begingroup\$ Also are you saying that after R9 and r10 I should have another op-amp and wire the pots to that? \$\endgroup\$ – user432024 Jul 15 '19 at 15:52
  • \$\begingroup\$ I added some sample devices in my question. \$\endgroup\$ – user432024 Jul 15 '19 at 15:52
  • \$\begingroup\$ @user432024 no, connect the pots to pins 1 and 7. The other ends can be at GND. Every pot sliding output should have its own safety resistor against short circuits and static electricity. \$\endgroup\$ – user287001 Jul 15 '19 at 15:55
  • \$\begingroup\$ The 833 is designed from the ground up to be a preamp. As such, it has almost no output current capability. The lowest load impedance for any listed parameter is 2 K, so I don't think it can drive three 2 K pots in parallel very well. Also, it will drive long cables just fine - the better the shielding, the lower the noise, but that has nothing to do with the chip or circuit configuration. \$\endgroup\$ – AnalogKid Jul 15 '19 at 15:58

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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