[sdiy] Which opamp pins wiring is best for CV out ?? A , B or C ???

[Richie Burnett]

As you say it doesn't take many feet of cable capacitance to cause a 
wideband op-amp to oscillate at HF, particularly if it was operating near 
unity gain.  Add to this the fact that audio inputs frequently have discrete 
capacitance deliberately placed across them for ESD/EMC compliance and HF 
stability can become a real problem for op-amp line drivers!

Regarding this CV driver output discussion, there is a solution that 
provides the best of both worlds:

http://www.richieburnett.co.uk/temp/cv_out.GIF

Resistor R1 limits the maximum output current under short circuit conditions 
and decouples the output of the op-amp from the capacitance of the cable and 
the load at the end of the cable.  (It's value isn't critical, but it needs 
to be significantly greater than the open-loop output impedance of the 
op-amp at the cross-over frequency so that it effectively decouples the 
output node from the capacitive lag at the load.)

Resistor R2 takes negative feedback from the actual output at DC.  This 
means that any voltage drop across R1 due to load current is inside the 
feedback loop and is therefore compensated out.  (Under load the op-amp 
servos it's output to a higher voltage to make up for the volts dropped 
across R1, and ensure the output voltage is still what you wanted.)

Capacitor C2 takes negative feedback directly from the output of the op-amp 
at high frequencies.  This stops the op-amp ringing or breaking into 
oscillation due to load capacitance causing the feedback signal to diminish 
and start to lag at high frequencies.

Essentially the source of feedback gradually shifts from the output cable to 
the op-amp output as you go up in frequency.  At DC the feedback comes from 
the actual output to compensate for the DC voltage drop across R1.  At HF 
the feedback comes from the op-amp output directly to ensure "in-phase" 
negative feedback and guarantee stability as the op-amp's gain rolls off.

Another way you can look at it is by thinking of R2/C2 as a phase-lead 
network.  C2 imparts a phase-lead into the feedback signal around the break 
frequency where the output capacitance combined with R1 would otherwise 
cause a phase lag to kick in.

I hope this helps,

-Richie,