NFB capacitor

[karatestu]

Good evening,

Been thinking about feedback caps in solid state power amplifiers. I only have experience of Naim, Avondale and NVA amplifiers. The feedback arrangement is similar in all of them. I know they have to be a certain value ie 47uf and above. Naim used a polarised electrolytic. Avondale recommended a wet tantalum. NVA uses a polarised aluminum electrolytic. I believe that non polar caps can be used as the is hardly any voltage across these.

What i really want to know is which capacitor parameters are important for this role. I didn't think esr was that important as it is in series with a 470R to 0V. I could be wrong though. So what is important? Low leakage, ESL or what?? I am inclined to fit a film cap but obviously they are physically larger. I do have 56uf 25V Sanyo oscon sp caps that are a perfect physical fit and extremely low esr but they don't have a good rep for amologue use.

Any advice received with thanks, Stu

[Nick]

I use the Nicheon es bipolar ones mainly from Bob Cordell's suggestion. But I do feel it's important to have the same cap in the feedback and input blocking position. In the feedback one leg is grounded, but in the input side it's floating so I found insulating the case, wrapping with copper foil and earthing the foil made a big difference wrt rfi sensitivity. For that reason I dislike the use of physical large caps as they make better aerials.

[karatestu]

Thanks Nick

NVA famously don't have input coupling caps but mine do as removed them from my source components . I have red wima mks2 3.3uf in currently for input coupling not sure they are available in 47uf.

My film caps are evox rifa mmk 47uf 63v and bloody massive. I might look for wima and if not try the oscons but they have a reputation for instability. Will have a look at those nicheon es bipolar.

Stu

[Nick]

Yep, I don't think that NVA use a long tailed pair on the front so its less of an issue with the equal caps being used as the conditions on the input and feedback node are different.

[ed]

Yep, I don't think that NVA use a long tailed pair on the front so its less of an issue with the equal caps being used as the conditions on the input and feedback node are different.

not that it's relevent to the op but are you sure about the ltp?. The 4 or 5 circuits I have all have ltp..but they may not reflect current builds.

[Nick]

No, not certain. What is the current circuit?

[ed]

No, not certain. What is the current circuit?

I don't know. I have old circuits for AP10, AP30,AP70,A60,A70. I'm not sure of the provenance of any of them, it was a long time ago.
Perhaps Steve or Stu could comment on the ltp on their boards.

[Mike H]

I use the Nicheon es bipolar ones mainly from Bob Cordell's suggestion. But I do feel it's important to have the same cap in the feedback and input blocking position. In the feedback one leg is grounded, but in the input side it's floating so I found insulating the case, wrapping with copper foil and earthing the foil made a big difference wrt rfi sensitivity. For that reason I dislike the use of physical large caps as they make better aerials.

[karatestu]

No, not certain. What is the current circuit?

I don't know. I have old circuits for AP10, AP30,AP70,A60,A70. I'm not sure of the provenance of any of them, it was a long time ago.
Perhaps Steve or Stu could comment on the ltp on their boards.

Yes there definitely is a LTP.

[Nick]

Fair enough. I would have thought the different conditions on the two sides of the LTP would lead to a DC offset on the output.