Sound Quality Vs. Measurements

[a.wayne]

So based on you hypothesis , one big un, better than 2 small un's , I'm assuming Torods ..?

[abraxalito]

Yep, go for the single big 'un. But toroids have maybe 10X the capacitance to mains of the other types, at equal VA ratings. So if the bigger amp requires a toroid, that could be part of the reason the bigger amp doesn't sound so good.

[a.wayne]

Very, very few amplifiers are non-toroids today , even the expensive shiney ones, is this xtra capacitance really an issue...


Pros vs cons, how does it affect the sound ............. ?

[ThorstenL]

Hi,

Originally Posted by abraxalito

I did say RF, explicitly. And RF is not going to fall in the category 'large'. Differential mode leakage is inevitable if you want to, you know, draw power from the mains.

Leakage is unavoidable, however magnetic coupling at high (RF) frequencies is minimal due to eddy current losses in the core, so what remains is capacitive coupling. And that is something that can be readily addressed using traditional transformer design techniques.

Originally Posted by abraxalito

Medical safety has spec for RF leakage currents? If so, do you have a link, I'd like to follow up.

No, it has a spec for leakage current however, that forces certain design approaches that as a side-effect minimise RF leakage.

It is however possible to improve on this too.

One may also employ additional RF filtering, though this is usually only operates above a few 100KHz, leaving a wide open door for some of the most pernicious stuff.

R-Core, LL-Core and Double-C Core transformers allow for example a much more thorough optimisation than EI, M or Torroid transformers.

As for using two transformers, while high HF frequency gunk is not amenable to improvement using two generic transformers, Low RF gunk can be fairly successfully cancelled using suitable arrangements.

Ciao T

[abraxalito]

Well I reckon balanced inputs are largely immune to common-mode induced sound quality issues, provided pin1 is properly terminated on the XLRs. Unbalanced inputs? - don't touch them - on big, toroidal-fuelled poweramps would be my advice. If you only have unbalanced ins, wide-band mains filtering would be the mitigating option. Problem is - with the heavy current demands of a beefy amp it becomes increasingly difficult to filter effectively.

[ThorstenL]

Hi,

Originally Posted by a.wayne

Whats considered "generic" ....?

Most if not all that you find in the catalogs of Digiykey, Newark, Farnell, RS-Components etc.

As a counter-example, Plitron makes transformers (not all, but some specific ranges) that are distinctly non-generic.

Ciao T

[ThorstenL]

Hi,

Originally Posted by abraxalito

Yep, go for the single big 'un. But toroids have maybe 10X the capacitance to mains of the other types, at equal VA ratings.

This is only true for generic transformers. Ones correctly optimised for audio use may have much lower RF coupling than generic EI or related designs.

Ciao T

[jcx]

actaully one "high end" amp I know of has EI mains xfmr in a cylindrical drawn metal case - because for a while toroids were precieved as "hi tech"

nowdays toriods are often cheaper than EI due to the material efficiency, lower shipping weight

[a.wayne]

Originally Posted by ThorstenL

Hi,



This is only true for generic transformers. Ones correctly optimised for audio use may have much lower RF coupling than generic EI or related designs.

Ciao T

Hello T,

So 2 in parallel is better than one big un ..? what is plitron doing differently , bifilar wound ..?

[abraxalito]

Originally Posted by a.wayne

bifilar wound ..?

That would be a way to make matters far, far worse than even a generic trafo