Feedback capacitor

[MarcH]

I've heard that the quality of a feedback cap can influence sound quality quite a lot. Since the feedback cap in my circuit is large (47uF) film caps are out of the question.

What should I look for ?

Coupling cap spec like low leakage and low distortion, or low esr/esl spec like decoupling caps ? In a coupling situation a nonpolar low leakage elyt is often preferable to a high spec low esr cap like os-cons IMHO.

[AndrewT]

a standard low voltage electrolytic works here.
It may sound better with a good quality parallel plastic cap.
But, ensure it never has a significant AC voltage across it. i.e. make it at least half an octave below the high pass filter fitted to the input.

[MarcH]

Thank's for a quick reply. So a standard elyt cap, perhaps bypassed by a film cap, is what you recommend. In other words a silmilar setup to a signal coupling cap ? Some people find that specific caps like Elna for audio, Rubycon BG nonpolar etc make a difference. Anyone who has tried out different elyt caps in a feedback position ?

[Conrad Hoffman]

It may be all in my head, but feedback caps are the one place where I think I hear a difference. The key is up above- choose a large enough cap that no significant AC voltage appears across it. It can't have much effect if it's a short at all audio frequencies. Though they're great for bypassing and have very low esr, I don't like OS-Con caps in this location. IMO, NP electrolytic caps are undesirable as well. Go with a very high quality standard aluminum electrolytic, possibly protected from over voltage with a diode chain as Doug Self recommends. IMO, standard case sizes and higher voltages are also better than the ultraminiature ones.

[MarcH]

Thank's. You information makes sense. I'll try a high quality polar low leakage elyt cap.

[martin clark]

If you can make it work with the impedances acceptable for your feedback network, do try a film cap alone here; IME it is much better than electrolytic parts.

Evox Rifa MMKs are available in 22uF, and they used to do a 47uF/63v part...

[Conrad Hoffman]

The worst film cap is way better than the best electrolytic, though for the feedback values I use in power amps, I've never been able to shoehorn in enough film to do the job. Still, if you can do it, that's the way to go.

Question- with a jfet diff amp, why can't one scale up the feedback divider so the necessary cap is much smaller? That would seem to be a good reason to go jfet.

[jcx]

If electrolytic feedback cap is needed, nonpolar construction is known to have less measurable distortion (Cyrll Bateman's articles), also "supersizing" - putting the highpass corner F way below 20 Hz reduces distortion as well as phase shift(Doug Self's site)

[MarcH]

I can source "boutique" caps said to to good for audio signal coupling that would fit in my rather small power amp pcb with a lot of smd components. Either a Black Gate type C (C=coupling) or an Elna silmic, both 47uF 50V (high voltage rating for low tan delta). Neither is particularly expensive. There is a more expensive Black Gate type N (N=Nonpolar) but it is too big. Would you use anything else in this scenario?

[AndrewT]

47uF sounds quite small.
What value of feedback resistor are you planning. I would expect 2k7 or 3k0 to match the 47uF. That's high.