In the previous post, I described the circuit that I’m using to emulate a potentiometer using a couple of opto-isolator (or opto-couplers?). There is one small problem with the circuit: the input impedance has to be quite low.
If you just plug a guitar in the circuit, which generally has a high impedance output, it will have a hard time keeping its signal straight. Mostly, you will loose some of the high frequencies present in the guitar signal. Even more so if the cable between the guitar and the circuit is long. For a real and very detailed explanation, take a few minutes to read http://www.soundonsound.com/sos/jan03/articles/impedanceworkshop.asp
The best way that I know of to correct the impedance mismatch is to use a line buffer. So I decided to include one right into my VC-1. This way, the user (sorry… guitar player…) can plug his/hers instrument into the PU-2 and the signal will be brought down to the proper impedance. There is another benefit to this: the impedance buffer/VC combination will keep the impedance fairly low all the way to the output. This way, the rest of the guitar signal run will receive a lower impedance signal, which is always a good thing.
The circuit I’m using:
The TL072P Op Amp is just used for pin placement. I’m actually using a NE5332A Op Amp because it has a very low noise circuit and is only marginally more expensive than the TL072. The 2M resistor are Metal Film for even lower noise. The input impedance is 1M, which is plenty for any guitar. The ouput impdance is only a few ohms. The gain, which is controlled by the connection between pin 1 and 2 (and 6 and 7 for the other Op Amp) is set at 1, so the signal at the output should be at the same volume as the input.
What is the point of using 2 opamps in parallel (noise reduction)?
I remember the 5534 family does not perform very well on low voltage. Also 5 V supply is a littel low – esp. as the 5532 is not a rail-to-rail amp.
Have You ever calculated how much (thermal) noise a 2M resistor will generate ? Well, OK You actually have 1M/parallel Guitar-output-impedance) relevant for noise in Your circuit. But still this is a lot more than the 5532s (an TL072s) input noise voltage, plus the 5532 also has input noise current (worse than its input noise voltage and worse than the 072) – this OPAMP is good for low-impedance circuits (like Mic Preamps).
For hi-Impedance circuits better use FET opamps.
There is all sorts of (sometimes bizarre! http://www.barberelectronics.com/parallelopamps.htm) info on parallel opamps out there. A lot of it points to noise reduction.
But the real reference was taken here: http://www.muzique.com/lab/superbuff.htm
“Have You ever calculated how much…”
I did not calculate much on this circuit. I used the good old method of trusting my ear, and my oscilloscope. The circuit used for that little buffer didn’t affect the sound negatively, and the oscilloscope was used to determine if any “digital artifact” was bleeding through from the control side. Keep in mind that the gain is unity. I’m not pushing anything here. The buffer is only used as an impedance matcher for the input to the volume control. Any buffer could be used, as long as its output is low. And I know that a lot of people have developed very good ones! For voltage: 5 volts is what i had on hand (or in circuit…). I would package this with a 9 V battery instead and use a 7805 to produce 5 volts for the volume control (I have one Beta setup like this). The opamp is also mounted on a socket: switching it with another pin compatible chip is an exercise left to the user… but definitely possible.
My goal for this sub-project was to present a low input signal to the volume control. Its circuit wants low impedance… I plan to print the circuit boards independently.
If it sounds good to your years: this is all that matters.