Yes they were Vcap CuTF, about $200 for a set if i remember right, i actually got them - but cannot tell if they are worth it - because haven't built one without them. But as it is with high end components - the minor improvement is very often not worth the price they cost - if you are not aiming for the very best. Also these were matched +-1% if i remember correct, that is a big plus though.
Ake, when you first power the Paradise up it takes about 70 hours of continuous running for the banks of el'caps to fully form. They only see a small voltage and as they form the DC offset begins to drop. You'll only drop to a couple of mV of dc drift once you have the Paradise in a box, it'll never get better than +/-10mV until it is fully cased up.
My builds used to start off at +/-100mV and settle down to 0.1mV> over the course of a few days.
The OPA 627 is a little better than the standard spec op-amp, but at significant cost. I'm not sure the difference is audible.
My builds used to start off at +/-100mV and settle down to 0.1mV> over the course of a few days.
The OPA 627 is a little better than the standard spec op-amp, but at significant cost. I'm not sure the difference is audible.
In my experience, the biggest output offset fluctuation comes from thermal drift of the input transistors. The servo has a time constant of about 1s, so anything faster will not be regulated away.... Try blowing on the input transistors and watch in awe what the output does 
In a box, the thermal drift is massively reduced. You could also glue a piece of copper foil on top of all input transistors, to reduce thermal airflow and couple them better....
In a box, the thermal drift is massively reduced. You could also glue a piece of copper foil on top of all input transistors, to reduce thermal airflow and couple them better....