DaveC1 wrote:
For what its worth, in my opinion, much of the reason for tube power amps sounding superior to SS amps has to do with the output impedance differences between the two classes of equipment. It seems that most conventional loud speakers produce more mid range IM distortion with lower damping factors than with drive impedances that are higher according to at least one white paper that I've read. This may be one reason that zero global feedback amps like some of the Pass designs sound so good. One of the effects of global feedback is reducing output impedance.
For what its worth, in my opinion, much of the rea... (
show quote)
Just some thoughts and comments.
First, most current solid state amps do not use output transformers so the actual load impedance is controlled by the speaker. On the other hand, tubes, which have a much higher impedance than transistors (because the voltage-to-current ratio is higher) do use an output transformer whose secondary impedance is matched to the loudspeaker.
Secondly, in my opinion, based on extensive real-world measurements, there are 3 basic reasons that tube amps sound different, or warmer or more musical, or...
Number one - when tube amps are over driven, their distortion increases much more slowly than solid state amps. A 100 W. solid state amp may have .05% THD or IM at 100 watts, but increases to .5% (an order of magnitude!) at 115 watts (an actual measurement from a Yamaha solid state amp). This is one reason tube amps are often used in studio mic preamps, where it’s had to control the input level, which may quickly change drastically.
Number two - because tube amps have a slower transient response or slew rate than a solid state amp, and the damping factor is lower, they produce a “softer” response to transients. I recall a listener test of multiple CD players, and the worst players in terms of square wave response were the most preferred by listeners.
Number three - I can’t state that this is true for all amps, but the tube amps I’ve built have the majority of distortion products as even-order distortion (2nd, 4th harmonic, etc), while the solid state amps I’ve tested have predominately odd-order distortion (3rd, 5th harmonic, etc). In western music (from what little I know of it), chords are often made up of even order harmonics - an octave is 2x in frequency, and it may be (just my conjecture) that we are more comfortable with even order distortion. That supposition needs some real investigation.
Finally, on the subject of feedback, the holy grail of a segment of the audiophile community is zero negative feedback (along with SET or single-ended triode power amps, usually using a 2A3 - a triode dating back to the beginning of radio). That will get you very limited power and very high distortion, both of which I find unacceptable. Even with effecient speakers (unless you listen with headphones) you need clean power to accurately articulate transients without noticeable distortion. And if you employ zero feed back, the THD and IM distortion is terrible - 3-5% (!) is not uncommon. The question is not only how much negative feedback to use, but how it is applied. McIntosh (mentioned earlier) put an extra cathode winding on the output transformer. I prefer the classic push-pull ultralinear configuration with the feedback path adjustable and the response shaped by a proprietary network.
Cheers