Mixer has me stumped (with voltages and audio probe info)

[Top Top]

I am building a mixer similar to the GGG mini mixer, but with jfet gain stages for each channel so I can plug guitar or mic straight in. The mixer part after the gain stages is identical to the GGG mixer, with the exception that the caps are before the volume controls instead of after (though I don't think that should make any difference.

Here is my layout, For clarity I omitted the wires that go from the volume controls to the resistor/opamp section:



The only things different on there are that I used TL062 rather than TL072, and I used 4.7uF caps in place of the 2.2uF ones because I didn't have 2.2uF on hand - don't think that should make any difference anyway.


I've gone over and over this and can't figure it out.

All the gain stages work AOK. I am getting a signal after the 100K summing resistors as well (up to pin 2 on the opamp), though I guess I will say, it sounds quieter than I would expect after going through only 100k. Not sure if my judgment is correct on that. It is not as quiet as the non-amped input, but a pretty sizable drop from before them where it is nice and loud. I feel like I have heard a volume drop before from a 100K resistor, and it was almost ignorable. I'd say by ear this drop is around half the volume.

Anyway, Then from there I get NOTHING other than a pretty loud hiss on 1&8 and a hushed hiss on 6.

Here are my voltages on the opamp:

1 = 7.79
2 = 1.73
3 = 4.04
4 = 0
5 = 4.04
6 = 4.65
7 = 1.97
8 = 8.50


** EDIT: one other bit - the inputs of the resistors each have about 1.5V on them if the volume controls are up (0V if they are down), and the volume controls make a scratching noise when you move them, as I think happens when there is DC on a pot... **

I figure that should not be too esoteric to anyone who knows what an opamp would read in this type of very standard setup (I don't know, however)

I tried a couple opamps in the socket as well... all same deal.

I am running from a regulated 8.5v supply, which appears to be working well (which is being supplied from the middle pin of the onboard LM317, which as you can see is connected to that middle rail that the 10K resistors from the JFETs are connected to as well).

The output has 0V DC.

Just totally stumped.

For reference, the GGG schem is  here:
http://www.generalguitargadgets.com/diagrams/mixer_sc.gif

and the layout is here:
http://www.generalguitargadgets.com/diagrams/mixer_lo.gif

[Top Top]

Ok...  so following a hunch I just learned something... maybe? 

I decided to see if I could cut out that scratchiness on the pots by putting caps between the volume pots and the input resistors to the opamp section and... surprise! It works now.

I don't quite understand why this is though.

Anyone who could enlighten me, it would be appreciated... The DC voltage was coming from the opamp section (I know this because when I disconnected the volume pots from the transistor section, the DC voltage on the output of the transistor section disappeared.

But I am still missing something fundamental here... I don't understand why DC on the pots would cause the opamp not to work if it is clear the signal was making it to the input of the opamp. Something about the decoupling caps I must not be understanding.

[R.G.]

...the exception that the caps are before the volume controls instead of after (though I don't think that should make any difference.

It does. Not DC-blocking the signals into the opamp mean that the opamp inverting input is being dragged toward ground by the pots. That shows up in the voltages:

1 = 7.79
2 = 1.73
3 = 4.04
4 = 0
5 = 4.04
6 = 4.65
7 = 1.97
8 = 8.50

Pin is up at nearly the full supply voltage, and that's because pin 2 is very far below pin 3. Opamps amplify the difference between the +  and - pins.  Pin2 is much lower than pin 3, and that forces the output to be higher than pin 3, by about 100,000 times the difference, depending on the opamp. Pin 1 is banged against the + power supply as hard as it can go. No signal goes through. On the second opamp, pin 1 being high similarly bangs pin 7 against the lower power supply (that being ground in this case).

[Top Top]

Ah, I get it. For some reason I wasn't thinking of ground as a concern when it comes to blocking DC, but now I see that the pots were interfering with the operation of the opamp. This actually helps me understand a lot about decoupling caps.

One other thing I found is that I needed to increase my mixing resistors to above 100k because the channels were having weird interactions. I assume this is why the volume pots were 10k in the original design. I just increased those resistors to 200k and it works well now. I used 100k pots because I wanted to use sliders and only have 100k sliders on hand.

Also, it is nice that the jacks I have happen to be switched/normaled, because I can normal them to ground so that the channel does not buzz if nothing is plugged into it but the faders are up.

[Top Top]

Here is an updated layout if anyone would be interested in such a thing:

[R.G.]

Ah, I get it. For some reason I wasn't thinking of ground as a concern when it comes to blocking DC, but now I see that the pots were interfering with the operation of the opamp.

In a classical opamp circuit with a bipolar power supply, the pots are referenced to ground, so are the inputs, and there's no need for blocking caps except for frequency shaping.

Likewise, if you had a solid, low impedance bias voltage for the first circuit, the bottom of the pots could go to the same bias voltage the inputs are referenced to and not cause any DC problems. It's where "ground" is offset from the opamp bias that the DC offset gets you.