When I use nonpolarized caps, I just make sure that they are (channel-to-channel) oriented the same; I send the signal down the print on the outside, presuming that they are manufactured consistently.

I am presuming presumptuously?

Heh,

Poinz

Some manufacturers specify which way to connect the cap, while others don't. I know one brand that outer wraps the caps randomly, so it is an individual cap situation. Mylars may not be that critical as there sonic signature is usually fairly strong.

Take care.
Steve

Always connect the outside foil to the lowest impedance, not the lowest voltage as many recommend. Voltage has nothing to do with it, it's the lowest *impedance* that you want! That'll most always be the plate of the first stage, versus the control grid of the second stage. The control grid is an extremely high impedance itself, with (usually) a bare minimum of 100K all the way up to a 1 Meg resistor to ground.

OTOH, the impedance of the first driver stage (which is the product of the internal plate resistance in parallel with the load resistor), will most always be much lower. Even a very high plate resistance triode like a 12AX7 will only have about 40K output Z when the cathode resistor is bypassed, unbypassed still less than 70K depending on the cathode resistor value. So, the outside foil goes to that side for the best electrostatic shielding.

Thermionic

Yes, it can make a sonic difference/sound, depending on the quality of the cap and quality of the components.

A low value capacitor, say 0.05uf is generallly more critical in respect to the outer foil vs a higher value cap, say 2.0uf. The outer foil should be to the lower Z for sure if the cap is small.

Good shielding and design will minimize or eliminate noise, interference problems to begin with.

Take care.
Steve

Steve