Hey!

So I've been trying to build a modular synth on/off for the last year or so, and one of the most simple projects I thought I should start with is a mixer circuit. I found a really simple circuit schematic and shortly after got it working on breadboard, with absolutely no problem at all. That was 8 months ago.



The problem started when I tried to implement it on stripboard (or Veroboard), and failed miserably. The signal seemed to disappear right after the three parallel resistors (R1, R2 and R3) with no trace, which to me would suggest a short somewhere but I couldn't find one.
After a lot of debugging I figured it must be either my soldering job that was terrible (but showed no apparent signs of bad connections through a multimeter) or it was the plate I was using for the module (at the time, alluminum) was creating some sort of weird ground effect (which is a question I do have about modular synths, so if anyone knows if the plate matters in terms of ground interference, I would gladly appreciate some info!).

So I stopped working on it for a while, until I decided to try again but using a wooden plate. I stripped the previous one's components and redid the circuit on breadboard (removing the capacitors from it to see if it would work without them) to check if the components were out of whack, and it worked with no problem again. So I was confident that this time, the stripboard version would be functional.

I was wrong once more. The same exact problem was happening, which tells me this has to be something consistent I'm doing in regards to my stripboard implementation.

​

I pinpointed that the signal disappears exactly in that red area, I thought that the obvious answer would be the potentiometer's broken or wired wrong, but i checked the connections and everything was well soldered and the potentiometer had to be working since it just previously worked on the breadboard.

Although something is weird about that potentiometer (which by the way the circuit calls for the "third" leg to not be grounded to not short the opamp when it's fully closed), since when I checked it with a multimeter it gave me readings of Mega Ohms, which is suspicious given it's a 100K linear pot, but if I checked opposite ends it gave me the 100k value with no issue, although that value did not change with the sweep of the knob.

There is also something else worth noting, that could be completely normal but it still irks me slightly, when the jacks are connected sometimes they short when there is no signal going through them, this creates a short in the R6 resistor sometimes, but it doesn't seem to warrant the signal disappearing on the other side of the op-amp.

Even when I replaced the pot, with a fixed resistor it didn't seem to fix the problem.

I have checked every single connection and supply voltage at the op-amps pins, everything seems to be well connected and in it's place.

I'm gonna leave a couple more images of the circuit (excuse the terrible soldering job), does anyone have an idea of what it could be?



The hole in the middle is something I added to standoff the wooden plate. Couldn't find actual acessible ones in my country so I used a bolt for it.

 

There's no way to tell from the photos what the problem may be.
It does sound like the pot may be the problem though.
I reviewed your schematic and made some modifications to simplify the design.
You will see that I eliminated the 100K pot P4 and replaced it with R7. I also removed R4, R5 and C4.

 

At pin 2 of the TL072 you should see something very close to ground because that is to what pin 3 is connected. In a properly working inverting summing amplifier (which this is supposed to be) there is almost no signal at the inverting input.

You can check that things are wired correctly by unplugging the TL072 from its socket or disconnecting the power supply. At that time a signal, which would be the sum of the three signals being mixed should appear at the pin 2 connection.

If your TL072 is correctly wired, then the second stage should work just fine as it is merely a voltage follower. The DC voltage out of pin 1 should be nearly ground and so should TL072 pins 5, 6, and 7.

It is a simple enough circuit, but it is easy to mis-connect things on stripboard unless you have had a lot of practice. You might want to buzz out your connections with an ohmmeter.

 

I'm confused by the second opamp. Wouldn't the signal normally go to the non-inverting input of a voltage follower?

 

I'm confused by the second opamp. Wouldn't the signal normally go to the non-inverting input of a voltage follower?

Yes.
That amp in post #2 is wired incorrectly.

 

Duh! I should have caught that! Well the TS only had to wait a few minutes for an astute observer to come along. Good observation upand_at_them!

 

Hey, I got one right for a change.

 

I have made hundreds or thousands of circuits on stripboard and they all worked perfectly the first time they were tried.
But I made them neat, compact and tidy.
This one has a mess of long wires (antennas for interference pickup) all over the place and questionable and shorted soldering.

The adjustable gain opamp has a maximum gain of only one (no gain like a piece of wire). The second opamp also has a gain of 1.

With the 100k pot set to maximum then the opamp gain is 100k/100k= 1. The TL072 datasheet shows a typical open loop gain of 10 thousand at 330Hz so if the AC input is 0.1V then the voltage at the negative input is 0.1V/10 thousand= 0.00001VAC which is only 10 microvolts. That is what you tried to measure.

 

At pin 2 of the TL072 you should see something very close to ground because that is to what pin 3 is connected. In a properly working inverting summing amplifier (which this is supposed to be) there is almost no signal at the inverting input.

You can check that things are wired correctly by unplugging the TL072 from its socket or disconnecting the power supply. At that time a signal, which would be the sum of the three signals being mixed should appear at the pin 2 connection.

If your TL072 is correctly wired, then the second stage should work just fine as it is merely a voltage follower. The DC voltage out of pin 1 should be nearly ground and so should TL072 pins 5, 6, and 7.

It is a simple enough circuit, but it is easy to mis-connect things on stripboard unless you have had a lot of practice. You might want to buzz out your connections with an ohmmeter.

Ok, so I did this exactly, removed the IC and even disconnected the pot to make sure there was no interference with the signal. But the moment, the signal goes through the resistors R1, R2 or R3, it completely disappears, doesn't even manage to get to pin #2 on the IC socket. I've buzzed out every connection with a multimeter in this part of the circuit and it seems fine!

Now as my last resort question, is this how you solder a parallel connection between resistors (R1, R2 and R3) , like it's shown in the schematic?

Because this seems like the only thing that could carry as a problem from the last version on stripboard that I did.

​

 

You do not understand that the inverting input of an inverting opamp has an extremely low signal voltage because it is cancelled by the negative feedback. The opamp will still have plenty of output signal voltage.

The mixing resistors are not in parallel. Instead they are joined at the opamp's inverting pin 2 input.

 

You do not understand that the inverting input of an inverting opamp has an extremely low signal voltage because it is cancelled by the negative feedback. The opamp will still have plenty of output signal voltage.

The mixing resistors are not in parallel. Instead they are joined at the opamp's inverting pin 2 input.

So the signal at the end of the mixing resistors is supposed to have a low signal voltage, even when there is no op amp connected?

Is that why there is no apparent signal at the end of them?

 

With the opamps removed then there is no negative feedback to cancel the signal at the mixing end of the 100k input resistors. Then the level should be almost the same as the input signal level.

What frequency is the signal and what are you using to measure the signal level at the mixing end of the 100k input resistors?

 

With the opamps removed then there is no negative feedback to cancel the signal at the mixing end of the 100k input resistors. Then the level should be almost the same as the input signal level.

What frequency is the signal and what are you using to measure the signal level at the mixing end of the 100k input resistors?

I haven't been outputing a specific frequency signal, since it's supposed to be an audio mixer, I've been playing music through the jacks, and trying to listen to it back on headphones with a disconnected jack, pinning points on the circuit.

And you can hear the music perfectly right before the 100k mixing resistors.

 

It looks to me that one of R1-3 is shorted by the copper track. I don't see breaks under all the resistors.

 

It looks to me that one of R1-3 is shorted by the copper track. I don't see breaks under all the resistors.

If you look on the right side, the first leg of the resistors (R1-R3) have a gap in between them to prevent any shorting. I've checked every resistor in the circuit as well to see if it shorts and none of them do. Unless there's some soldering principle I missed.

 

Headphones are a low impedance, maybe 32 ohms. Then the 100k mixing resistor makes a voltage divider to reduce the level to 32/100k= 0.00032 times (0.32 thousandths) which is almost nothing.

I hope you are not connecting headphones to the output of this mixer circuit. The TL072 is an opamp, not a power amp and its minimum load is 2000 ohms so it cannot drive low impedance headphones.

 

Headphones are a low impedance, maybe 32 ohms. Then the 100k mixing resistor makes a voltage divider to reduce the level to 32/100k= 0.00032 times (0.32 thousandths) which is almost nothing.

I hope you are not connecting headphones to the output of this mixer circuit. The TL072 is an opamp, not a power amp and its minimum load is 2000 ohms so it cannot drive low impedance headphones.

I did connect headphones to the output of the circuit before, that's how I tested the breadboard versions. By connecting two different audio sources at the inputs and using the pots to mix the signals in and out, I could hear the mix at the output and control the master volume of the mix with the 100k pot connected to the 1 and 2 pin of the TL072.

So it appeared it could drive the headphones.

The problem started when I put the same circuit on stripboard. Thus making me suspicious of my idea of the stripboard connections and the differences it can have from the breadboard implementation.

 

As drawn IC1A is fine, IC1B is not fine according to upand_at_them's observation (you should drive the + input rather than the - input) to wit:

With this correction the schematic is fine, that leaves the wiring, even though you buzzed out the wiring, there is still something wrong. Unfortunately, stripboard circuits are difficult to troubleshoot over the internet.

It looks like you used a drill to cut the stripboard to keep from undesirable circuit paths (shorts). Did you buzz these out to make sure they are indeed clear?

Audioguru again wrote: "32/100k= 0.00032 times (0.32 thousandths) which is almost nothing." In spite of past experience, if your headphones are 32 ohms even Superman with his super hearing wouldn't be able to hear anything on one of those through a 100k mixer transistor. You can check your headphone with an ohmmeter.

Just trying to clarify. I guess there exists a small chance that the TL072 is broken. If you have a spare, now would be a good time to try it, but after verifying the power supply connections.

Honestly, we are trying to help you find the problem(s). The fact that you had this working on a breadboard is evidence that you can get this circuit to work.

 

As drawn IC1A is fine, IC1B is not fine (you should drive the + input rather than the - input) to wit:
View attachment 233318
With this correction the schematic is fine, that leaves the wiring, even though you buzzed out the wiring, there is still something wrong. Unfortunately, stripboard circuits are difficult to troubleshoot over the internet.

It looks like you used a drill to cut the stripboard to keep from undesirable circuit paths (shorts). Did you buzz these out to make sure they are indeed clear?

I haven't applied this correction yet (assuming it's the corrected version of sghioto's circuit suggestion) since it worked on the breadboard, but I'll definitely give it a try!

The paths I cut, have been buzzed out and are clear.

 

Many audio mixer circuits use an extra output opamp as an inverter so that the output has the same phase as the inputs.