LMNC Super Simple Mixer [Strip Board Layout + BOM]

[Caustic]

Layout credit: LMNC (Sam Battle)

Transcription of instructions:

• CGS dc mixer in its simplest form basically. I use this circuit in many things. Including the bruel and kjaer fixed filter bank, who’s video I made this strip-board layout for.
• Remember if you are using a plastic panel, make sure to ground all jack sockets
• Potentiometers are facing up.
• Pin 2 of the TL072 is where all the inputs go into, you can actually have as many inputs as you want (in the bruel and kjaer fixed filter bank video i have 30!). To add more inputs is very simple. Put another 100k resistor to pin 2 of the TL072 and then copy the input jack and socket stuff and BOOM, you have as many inputs as you require.
• You can make this go to a lot more complicated town, adding inverting switches etc, but this is mainly for audio so inverting the signal isn’t too useful. A strip-board layout for a mixer aimed at control voltages will come.

Layout:

Bill of Materials:

SMALL STRIPBOARD 93X55MM (COPPER)

https://www.taydaelectronics.com/small-stripboard-93x55mm-copper.html
$0.66 / £0.50
Quantity: 1

TL072 LOW NOISE J-FET DUAL OP-AMP IC

https://www.taydaelectronics.com/tl072-low-noise-j-fet-dual-op-amp-ic.html
$0.34 / £0.26
Quantity: 1

1K OHM 1/4W 1% METAL FILM RESISTOR ROYAL OHM TOP QUALITY

https://www.taydaelectronics.com/10-x-resistor-100k-ohm-1-4w-1-metal-film-pkg-of-10.html
$0.012 / £0.0092
Quantity: 1

100K OHM 1/4W 1% METAL FILM RESISTOR ROYAL OHM TOP QUALITY

https://www.taydaelectronics.com/10-x-resistor-100k-ohm-1-4w-1-metal-film-pkg-of-10.html
$0.012 / £0.0092
Quantity: 6
Total Price: $0.072 / £0.055

6.35MM 1/4" STEREO PHONE JACK

https://www.taydaelectronics.com/6-35mm-1-4-stereo-phone-jack.html
$0.45 / £0.35
Quantity: 4
Total Price: $1.8 / £1.37

100K OHM LINEAR TAPER POTENTIOMETER SOLDER LUGS ROUND SHAFT DIA: 6.35MM

https://www.taydaelectronics.com/b100k-ohm-linear-taper-potentiometer-round-shaft-solder-lugs-l.html
$0.50 / £0.38
Quantity: 3
Total Price: $1.5 / £1.14

AWG 22 BLACK HOOK-UP WIRE 1FT (30CM) SOLID

https://www.taydaelectronics.com/awg-22-black-hook-up-wire-1ft-30cm-solid.html
$0.10 / £0.076
Quantity: 5
Total Price: $0.50 / £0.38

AWG 22 BLUE HOOK-UP WIRE 1FT (30CM) SOLID

https://www.taydaelectronics.com/awg-22-blue-hook-up-wire-1ft-30cm-solid.html
$0.10 / £0.076
Quantity: 1

AWG 22 RED HOOK-UP WIRE 1FT (30CM) SOLID

https://www.taydaelectronics.com/awg-22-red-hook-up-wire-1ft-30cm-solid.html
$0.10 / £0.076
Quantity: 1

AWG 22 GREEN HOOK-UP WIRE 1FT (30CM) SOLID

https://www.taydaelectronics.com/awg-22-green-hook-up-wire-1ft-30cm-solid.html
$0.10 / £0.076
Quantity: 1

AWG 22 YELLOW HOOK-UP WIRE 1FT (30CM) SOLID

https://www.taydaelectronics.com/awg-22-yellow-hook-up-wire-1ft-30cm-solid.html
$0.10 / £0.076
Quantity: 5
Total Price: $0.50 / £0.38

Total cost of components:

$5.68 / £4.33

[Caustic]

Notes on Schematic:

The TL072 is a chip which has two operational amplifiers, and this mixer takes advantage of a specific type of Op-Amp circuit called a “Summing Amplifier Circuit”.

Tldr;

Summing Op-Amp circuits basically mush all the inputs into a single output voltage that is literally the sum of all the other voltages.

As stated in the instructions, all of your inputs go into pin 2, which is the inverting input of the Op-Amp (the part of the triangle with the - sign).

Op-Amps have a virtual ground, which makes things a bit easier to calculate all this. Since we know the node with virtual ground is basically at zero volts, we can use Ohm’s Law to calculate the currents (i) of each voltage input (Va, Vb). So for any given input, you can just divide it’s voltage by the resistor values (Ra, Rb). The current leading to the virtual ground is the sum of all of those currents from each voltage input before it. Since the virtual ground is 0v, the current (i) is still the same leading to Rf. Ohm’s law for Rf is the 0 voltage of the virtual ground minus the output voltage (Vout) [simplifies to -Vout] divided by the value of Rf.

The final expression uses the fact that the two currents (i) on either side of the virtual ground are the same current. We just take both of these other expressions and set them equal to each other. Isolating Vout so we can solve for it gives that final expression at the bottom.

Vout = - ( Rf / Ra * Va + Rf / Rb * Vb )

So in essence, the two resistor ratios directly affect the output. A variable resistor at each input gives you the ability to alter these resistor ratios will affect how much of each input voltages are summed.

In Sam’s circuit, All of the input voltages are fed into the first Op-Amp’s inverting input [looks like a - on the diagram] (Pin 2) with a 100k resistor. The IC is given +Voltage (Pin 8), and -Voltage (Pin 4). The non-inverting [looks like a + on the diagram] inputs of both Op-Amps (Pins 3 and 5) are both tied to ground. The output of the first Op-Amp (Pin 1) is fed back into the inverting input of the second Op-Amp (Pin 6) with a 100K resistor.

Since the resistor values are all the same, it means the base expression for Vout would be:

Vout = - ( 100k / 100k * Va + 100k / 100k * Vb + 100k / 100k * Vc)

Simplified to:
Vout = - ( Va + Vb + Vc)

Which just means the output is the sum of all voltage inputs! This is why its called a “Summing Op-Amp”. Now, the voltage is inverted, but as Sam put it in his description, its audio so you wont be able to tell anyways.

So if you wanted to adjust the mix, say with the following values:
Va = 1x
Vb = 3x

Basically, you want Vb to be 3x “louder”. You have to pick resistance values to get you those ratios i mentioned before. If the values are all the same, its all just summing with no modifications.
Vout = - ( (Va * 1) + (Vb * 3) )

How do you get those co-efficients / multiplier dealies?

You pick or change the resistance of Ra or Rb.

Since Rf is shared by all inputs, you have to work with the same numerator (top number).

Rf / Ra = 1
Rf / Rb = 3

If we make:
Rf = 12k Ohm

12k / 12k = 1
12k / Rb = 3

We just pick a resistance value that when its divided by Rf, it results in 3.

12k / 4k = 3

So if we have resistance values of:
Rf = 12k
Ra = 12k
Rb = 4k

We get an output that features more of our Vb input, since there was less resistance.

If I am wrong about any of this, let me know. I tried my best to double check this.

[Jos]

@Caustic

Op-Amps have a virtual ground, which makes things a bit easier to calculate all this.

This is not entirely correct. Opamps with a feedback loop tend to make the voltage difference between their + and - input go to zero. Is one of the inputs connected to ground this basically means that the other input also can be seen as grounded / at zero volts. It is the feedback loop that creates the virtual ground here. Other configurations exist where this is not the case. So opamps themselves do not have a virtual ground. I’m sorry if this may sound like nit picking, but I hope this makes it easier to understand the circuit.

Just one other remark concerning this:

@Caustic

You pick or change the resistance of Ra or Rb.

This is correct, but one might now think, the lower Rb, the higher its part in the output signal will be, and therefore choose a low resistor value. But there is a snag. Remember Rb is connected to a virtual ground, so a low Rb implies connecting the signal source via a low resistor value to ground. Your signal source may not like that. The remedy is to choose Rb at a reasonable level and increase the values for Ra and Rf such that you get to the same ratio. The input impedance of the circuit is then acceptable for the signal source. For the opamp the absolute values of the resistors Ra, Rb and Rf are not of interest, only their ratios determines the amplification factors for the input signals.

[ChristianBloch]

@Caustic

Now, the voltage is inverted, but as Sam put it in his description, its audio so you wont be able to tell anyways.

I seem to remember from a comment over on Patreon that the second op amp is used to invert it back again just to keep things proper. I may be misremembering. Anyway, here’s mine:

[Caustic]

@Jos Right, thanks for pointing out the distinction. That feedback loop is totally important for that property! And as for the other bit yes, i was trying to show that the absolute values of the resistors dont matter as much as the ratios. They could all be 10k instead of the 100k and the ratios would be the same.

[Caustic]

@ChristianBloch Ah, thats why! Look at that dusty boy.

[alex]

although most of it went over my head i very much appreciate the explanation! helps a self taught person like me slowly learn more and more.

[Caustic]

I’m self teaching too! Part of the reason I am posting this is to be corrected if i am wrong, so i can self-reference it, and apply what I learned n stuff.

The basic idea is that a mixer is basically a summing amplifier. A summing amplifier uses an op-amp circuit which mushes the input voltages into an output voltage that’s modified by the ratio of resistors. If it helps, think of how it would be if:

Rf = 1
Ra = 2
Va = 4

Rf / Ra = 1/2
So, that ratio is essentially halving the Va voltage. Making the 4 volts 2 volts.

Then, if the “b” parts were:

Rb = 3
Vb = 4

Rf / Rb = 1/3
Vb is divided by a three, which is 1.33 volts.

Then we add them together.
Va + Vb = Vout
2 + 1.33 = 3.33

The way to think of it is how these ratios interact with the input voltages. If all the resistors are the same value, the ratios would just be like 1/1 or 1. So when they are all the same, you are just adding all of them together.

1 * Va + 1 * Vb = Vout
Va + Vb = Vout

Hope that helps!

[Doolang]

I made a stereo 8 channel version of this. Love it.

[Riggidyboo]

I built the mixer the other day to use with the Performance VCO’S. Didn’t work at first, but found a connection problem under one of the pins on the TL072. My brother works at a sign company, so I was able to get the panel for free and cut to dimensions from their scrap pile.

[MarkKelsall]

That’s exactly the sort of thing I need for my 4 simple oscillators. I figure that an oscillator is a good basis for starting out. What’s stopped me getting further is lack of a mixer, plus I have a bunch of spare op-amps sitting around for a small valve amp I own. Found tomorrows project. That’s what I love about this group, just plug a search in, and up pop the results… someone has always already done it.

[d42kn355]

My take on it with switches for each channel.
(idk if that works… does that work? lol)
just moved a couple jumpers to have the Eurorack Power connector onboard.

[sebastian]

I found this layout here and made a pcb some time ago:

I have not yet built it.
Can someone explain to me, what the difference to Sam’s mixer is?

[fredrik]

Your schematic has voltage followers on all inputs, followed by an inverting summing node. Upside is that the inputs are more isolated from each other than the other design, downside is that you have an inverting signal path which may be a problem, depending on where you’re putting the mixer. There’s also no output resistor protecting the output opamp (also not sure what R4 is doing there (*)).

The LMNC/CGS one has the inverting summing node directly on the input, which is then followed by an inverting buffer so you get the same polarity out. It also has a 1k resistor on the output, which protects the output opamp against shorts (1k is a bit high, 1/3 - 1/2 of that is enough). The unbuffered inputs could in theory led to some crosstalk if you have weird sources, but that’s probably not an issue in practice.

EDIT: *) guess R4 might an attempt to compensate for input bias current, but that’s extremely low for TL07x (typically 65 pA) which iiuc corresponds to potential nanovolt errors in this circuit…

[fredrik]

Also, 10k for the input pots is pretty low. Feel free to go higher; 100k is a more common choice (you want high input impedance) but the TL07x has extremly high input impedance (10¹² ohms) so you can go higher too.

[sebastian]

Thank you a lot! I think I will go with LMNC design then! It’s even more convenient, because I still have 100k pots around!
Also: For my first PCBs I completely forgot to put enough distance between pots! So they are not good to use!

Still a bit sad, because I like my drawing

[d42kn355]

love love love your euro LMNC style pcb… when you get it all settled. please remember i vaaaantt. lol

[fredrik]

To clarify, you can use 100k potentiometers with your schematics, so that’s not in itself a reason to throw it away. Not being able to fit them on the PCB might be a bit of a problem, though (unless you can connect them via wires). I’d also add, say, a 330 ohm current limiting resistor on the output to get a bit more protection.

(fwiw, true purists would probably go for a mix of these schematics and also buy the most expensive opamps they can find, but electronics design is pretty much always a trade-off )

(well, maybe not the most expensive things they can find, things like PA99 will set you back a thousand bucks per chip)

[Caustic]

Maybe try connecting the pots to the pcb to test that no other failing in the design first. That way you aren’t wasting the prototype! They look rad, and wouldn’t be out of place next to a kozmo module!

[Jos]

The outputs of the TL07xx series are short circuit protected. I remembered it from and just looked it up in the datasheet. So there is no need for a resistor.