How Much Difference Does a More Powerful Amp Make?

Metman - that's true, but I would rather have lower power amp that can perfectly drive very low complex loads. 50W/8ohm Coda S5, I mentioned, can output 200W at 2 ohm and according to Stereophile test can even drive 0.47ohm.

Half of the power loss in crossover, that you mentioned, is only 22% of loudness loss.

06-22-11: Raquel
......
First, it is very difficult to control a high-powered circuit without global feedback and virtually every high-powered amplifier therefore uses it. .....

Raquel, I believe that this is *not* a true statement. I think that you are confusing power amplifier gain with power amplifier (output) wattage. They are different. You need global negative feedback when you have many cascaded gain stages (wherein the power amp has a lot of output to input voltage gain) so that you can keep the entire design stable (oscillation free), lower in distortion & lower in overall noise.
To design a high wattage power amp you need to design in the correct transformer size, the correct number of output devices to handle the output current, the correct size & quantity of power supply caps to handle the charge being delivered to the load. This can be done using minimal number of power amplifier gain stages (for example, Pass does this often with just 2 gain stages hence he can avoid global negative feedback & still deliver to the market a X600 beast of an amplifier).

Also, you use the term "high-powered circuit". what, according to you, is a "high-powered circuit"?? you probably mean 'high wattage circuit'? Might be better to use the correct terminology to avoid ambiguity esp. if you are sharing some of your technical views on a subject. Just FYI.

Bombaywalla:

I was trying to keep it simple to make a point. Most amps, powerful or not, have anywhere from three to five gain stages (very few have two), and we all know that using global feedback with such designs makes the amp builder's job a lot easier. I do not agree that feedback ultimately lowers distortion - it changes the types of distortions to those that are different from the harmonics of live instruments (which is why it screws things up so much). Nelson Pass wrote the following:

"We have seen that nonlinear distortion becomes larger and more complex depending on the nonlinear characteristic of the stages, the number of cascaded stages, and the number of spectral elements in the music.

Negative feedback can reduce the total quantity of distortion, but it adds new components on its own, and tempts the designer to use more cascaded gain stages in search of better numbers, accompanied by greater feedback frequency stability issues. The resulting complexity creates distortion which is unlike the simple harmonics associated with musical instruments, and we see that these complex waves can gather to create the occasional tsunami of distortion, peaking at values far above those imagined by the distortion specifications."

Source: www.passdiy.com/pdf/distortion_feedback.pdf

PS - Pass makes a no-global-feedback amp that is even more powerful than the X600 monos: the X-1000, which also has only two gain stages.

Kijanski:

I really respect Ralph Karsten's (Atma-Sphere's) designs, and I have no doubt that his amp with 40 tubes is really good for what it is, but this is precisely the type of amp that will be noisy ("mucky"!) compared to an amp having, say, 4 tubes. 40 tubes?! Or 40 transistors? Have you ever heard an amp with that many tubes or transistors in a really resolving system? More importantly, have you heard an amp with only a small handful of tubes or transistors in a really resolving system? Come on.

06-23-11: Raquel
Bombaywalla:

I was trying to keep it simple to make a point.

ok, thanx for the clarification. I now know where you were coming from.

Most amps, powerful or not, have anywhere from three to five gain stages (very few have two),

OK, if you say so. i will not contest this as I certainly do not know this for a fact. I have not popped the lid of various amps nor studied their schematics to know any better. So, i'll take your statement at face value....

and we all know that using global feedback with such designs makes the amp builder's job a lot easier.

there was never a contest here - we are in full agreement.

I do not agree that feedback ultimately lowers distortion

I think that you should! ;-) any electronics book that discusses feedback will discuss 'feedback effect on distortion' & you will see mathematical equations showing that negative feedback does reduce distortion.
Having said that, in audio where we are talking about human hearing which is very sensitive to distortion components, I tend to agree with Pass' following statement (taken from your post):

Negative feedback can reduce the total quantity of distortion, but it adds new components on its own,...

I believe that you are in agreement with this statement as well.
this might be a better way of stating the distortion, negative feedback relationship.

Raquel - I've never heard either amp that I mentioned but specified S/N=120dB was confirmed by Stereophile measurement and audition of Coda S5. It is dead quiet and very resolving. As for Ralph amps - reviews praise them just for being quiet and very resolving. Putting two tubes in parallel doesn't make it less resolving or noisier than one tube. Part of SixMoons review of Atmasphere MA-1:

"Compared to my BAT VK-75, the Atma-Spheres just offered more of everything: More detail, more extension, more transparency, better soundstaging, greater micro and macro dynamics, more extended and luscious high frequencies"

You assessment of matching output devices is not accurate because it is easier to get average characteristic by connecting many output transistors in parallel. Linearity also improves since each transistor works at smaller current range.
Noise voltage that limits resolution of bipolar transistor is defined by resistance of the Base and resistance of the Emitter that can be reduced by reduction of emitter current (inversely proportional) while Base resistance become dominating reason for the noise. Placing transistors in parallel not only reduces emitter current but also reduces resistance of the Base (in parallel). Placing many bipolar transistors in parallel is well known noise reduction technique in microphone preamps.

As for negative feedback - It is not the complexity that screws up the sound but lack of bandwidth. Limited bandwidth amp delays the signal (shifts phase). When input signal changes fast, feedback loop is momentarily open creating overshoots (or odd harmonics in frequency domain). Our ears are very sensitive to odd harmonics (especially higher order) because loudness clues lie in them. It is called TIM (transient intermodulation) and is known since 70s.

One way is to compromise THD, IMD, DF and make an amp without global feedback but the other is to use just enough feedback to reduce THD to about 0.2-0.5% and then limit input bandwidth (before feedback summing stage) to one that amp had without feedback applied. There are many great amps designed with or without feedback - there is no simple answer. Some of these great amps are very simple some are very complex.

I just realized that Ralph Karsten has the same last name as first name of the Icepower (that I use) designer Karsten Nielsen. Karsten people design good things.