best amplifiers?

[lt cdr data]

what's your opinions and why?

[bear]

Very very easy to answer:

- the one you just bought.
- the one you just made.
- the one you just fixed.
- the one you found in the dump/trash.
- the one that someone you know had that was so good in his/her system, that you just got one too.
- the one you wish you had/could afford.
- the one that everyone who hears/sees your system wished they owned.
- the one you like to listen to most.

See, easy!



_-_-bear

[lumanauw]

There's 2 guy. 1 lives in North pole and 1 from Death Valley. They both asked the same question, what is the best drink in the world. The North pole guy said "A hot chocolate is very yummy". The Death Valley guy said "An ice crushed orange is very yummy". They altered their offering, the North Pole guy said "Any ice cold driink sucks ". The Death Valley guy said "Any hot drink sucks"
So, what is the best drink in the world?

[jacco vermeulen]

http://www.melroseflowers.com/mkic/r...es_cendol.html

[lumanauw]

He..he..Jacco..

This opinion is for myself. I don't speak for anyone else. I could be wrong, but for me it seems that better sounding audio power amp is the ones who doesn't use VBE multiplier. This could be transconductance amplifier (speaker is driven by collector/drain), or quasi complementary ala Kaneda or JLH. Examples : Nelson Pass's Aleph, Musical fidelity. Simple/few stages is also very important to good sonics, when you use global feedback.

[AKSA]

David,

I have found that the Vbe multiplier does have profound effect on the sonics, but it can be optimised with a very high quality cap, preferably a BG NX, of 1.5uF for each mA of VAS current.

Cheers,

Hugh

[lumanauw]

Hi, AKSA,

Yes, 3 stages amp will almost always use VBE multiplier as biasing device. It has been there since I'm not born yet I'm doing a funny thing. I look at the schematic of this 3 stages power amp (with VBE multiplier), and look at this same schematic with 30cm more distance than normal reading distance. I see that the important rails are +rail and -rail. GND point is floating in the middle. Unfortunately, speakers are connected to this GND. But if we consider the main rails are only + and -, it seems quite strange to maintain biasing the final transistors when it is hanging towards gnd (middle point) It seems we are forcing something unnatural trying to bias output transistors at this middle point when the swing are full towards +and-rails.