Can't seem to get OpAmps to work

[SmackaMuta]

If I'm understanding this correctly, this would reduce the voltage to the amplifier to 6V? Is there some sort of buffer circuit I could make to allow the same power supply to run both the preamp and the amp?

Edit:
In reading your post again, I think I misunderstood you. Your saying to remove the audio grounds from the virtual ground and put them to the 0V ground. Wouldn't this remove the virtual ground from the circuit and remove the -6V supply for the negative swing of the opamp?

[Mooly]

The virtual ground now becomes just a divider biasing the opamp to 1/2 VCC... just as it is now, nothing there changes.

The only actual change is moving the all the signal grounds to the zero volt line.

[SmackaMuta]

Ok, that makes sense, but what's the point of dividing the voltage if nothing is tied to the middle? Forgive me if I'm missing the big picture. =)

[Mooly]

The point of dividing the voltage is to bias the opamp correctly.

In a true split supply we would use the centre 0 V line. And the rails would be 6 volts above and 6 volts below that line.

In a single supply we have to bias the opamp in just the same way. If you think about it, the DC voltages are just the same as a split supply. It all depends on where you measure from Ground, zero, plus etc. They are all just arbitrary points.

[SmackaMuta]

That makes sense, but it seems contrary to moving the ground to 0v. Here's what's in my head:

Voltage divider gives us +6, 0, and -6. By grounding to the 0 it allows the opamp to swing both directions because the 'center' is at 0. If we move the ground to -6, we have affectively ignored the 0 and now have +12 and 0, which is what I had before the voltage divider. The opamp would now have no - to swing to and the sound would end up distorted as before.

Is this correct?

[Mooly]

There is something I overlooked actually... but lets work through it

Originally Posted by SmackaMuta

Voltage divider gives us +6, 0, and -6.

Not exactly. This depends where you measure from. If we measure from this divider (and we calling it "ground") then we can measure plus 6 volts above this and minus 6 volts below. The opamp output can swing equally above and below this "ground".
It also means the opamp outputs are at this same voltage. That means that they measure zero volts DC (relative to this point) but the reality is that they are 6 volts above the negative rail.

Because you want to run this off the same PSU and connect the grounds together we have a problem. The problem is that the "ground" of the preamp is the "virtual ground" point, and that is actually at 6 volts dc above the power supply zero line.

So you move the input grounds to the power supply zero.
We do the same for the output grounds.
We now call this new point "ground". All our voltages are now measured from here, and so what was the virtual ground point before, is now a point that sits at 6 volts. This biases the opamp outputs to 6 volts and allows them to swing equally above and below this point.

The thing I forgot is that you must now AC couple the output of the last opamp to block this 6 volts DC. Use a cap of around 10uf (depends on the input impedance of the power amp). Tha cap across the divider also needs to be around say 47uf.

[SmackaMuta]

Ok, solved one problem, now on to the next... =)

The audio portion works great now. The circuit listed in this thread works find and supplies only the bass to my subwoofer amp, the issue I am running into now is a constant static noise that is present even with the opamp removed. By static I mean sort of a white noise like an analog tv tuned to a channel that isn't there.

Any ideas? This noise wasn't present when it was connected only to headphones.

This project has been such an awesome learning experience. I really appreciate the help and knowledge.

[Mooly]

If it does this noise even with the opamp removed then I would be looking at the power amp it feeds.

Perhaps the power amp is unstable with the connection arrangement/leads you have.

Hard to say without seeing it.

Try applying a short to ground at the opamp output (on the outout side of the cap we talked about) and see if the noise is still there.

[SmackaMuta]

Shorting it completely removes the sound, would applying a 1M resistor instead of a full short fix this? Or is this a bigger problem...

Edit:
Went ahead and tried that, didn't help at all...

Another edit:
Not sure if I did this right, but incase this is part of the problem, I put a cap between the opamp output and the amplifier on both the + output and the ground.

[Mooly]

Well the opamp will drive loads down to around 600 ohms so just for interest you could try some lower value resistors instead of 1 meg.

But... whether that works or not, its not the answer.

You mustn't put a cap between the opamp output and the rails or ground. The only cap on the output is the one in series with the output.

If it makes the noise with the op amp removed as you mentioned, then there is nothing connected to the input of the power amplifier, it's just open and floating.

It should all look like this