Good Coupling and Decoupling Capacitors

[wrenchone]

To Chocoholic - I would never think of using electrolytic caps without some bias in the right direction. Otherwise, I would expect nonlinear distortion. I actually have a couple of positions on my preamp output boards for electrolytic caps in parallel with the film caps currently there. They would be a last resort in case I find I have to add more output load than the current 3uF films can handle (a question of LF corner frequency, not power handling). These caps would have a quiescent bias of about 15V.
The place I worry most about reverse bias on electrolytics is the DC blocking resistor commonly used in series with the feedback network for a non-inverting amplifier. You get a random amount of offset voltage across the cap (could be either polarity), and some negative voltage during negative output swings. I would prefer to use an Oscon in this application if I used a polarized cap at all. There is an alternate feedback connection that allows use of a fairly modest film cap. This has been discussed in another thread, so I won't flog it again here.

 

[ChocoHolic]

Hi Wrench!
...glad to hear your comments.

I did not read that other thread, I would expect that they were
discussing some sort "T-network" for feedback (three resitors one cap), right? Always had good results with that....

Bye
Markus

 

[wrenchone]

That's the one - a high value of resistor for unity gain at DC, then a voltage divider coupled to the inverting node with a film cap. I first saw mention of it in Audio magazine in the late 70's as part of a design description of a Nakamichi amp. I 've use it since.