Hey you guys! How come I seem to be the only one who gets told that I must like the "euphonic" distortion of tube amps.
I read an extremely interesting article the other day on the web. The premise was, that single-ended triode amps sound as good as they do, because of system synergy with single driver speakers(primarily). It seems this guy did some distortion analysis of SETs, and speakers. The SET generated primarily second order harmonic distortion, which was the bulk of the total harmonic distortion measured, very little other. The speakers measured primarily second order harmonic distortion also, with very little other. These distortion characteristics mainly differed in phase relationship.
His hypothesis was that if, by way of chance, the 2nd order distortion of each item appeared at 180 degree phase angles apart, the distortion would be cancelled out, creating the "straight wire with gain" ideal. If they appeared at 0 phase angle(together), they would add, and distortion would double. Since either case was unlikely, his solution was to determine which side of the 90 degree phase angle the intersection was, and phase the speaker wiring to place the phase intersection of distortion closest to the 180 degree mark that you could. This way if it was at 30 degrees and adding, by reversing phase, you could put it at 210 by reversing phase and achieve a high level of distortion cancellation. If you leave it at a phase angle close to 0, then the distortion adds - bad. Since the majority of this occurs in the 2nd order on both items, once a certain percentage of this 2nd order cancellation occurs, the THD is dramatically lowered for both items as a package.
This is his reasoning for why SETs have the remarkable clarity they do, while measuring in distortion %, the way they do.
His finding was that by switching phase connections on your speakers, you could find out which way sounds better(reduces distortion) on your system. You don't have to use measuring equipment, just your ears. The goal is to have the amp distortion cancel the speaker distortion as much as possible with your equipment.
This is not possible with push-pull or SS amps, as their distortion characteristics are comprised of many orders of harmonic distortion, and are not coincident with the distortion characteristics of speakers.
The effect is somewhat lost on multi-driver speakers because of the phase changes induced between drivers by the crossover network.
So, while I cannot personally attest to the veracity of his findings, it sounds like a very interesting explanation for the apparent dichotomy of SETs measuring poorly and sounding great. The amp is not used to drive a measurement meter. It is used to drive a speaker, and as such may exhibit different behaviors when doing so, than some lab freak with a meter may even realize.
Maybe there is no "euphonic" distortion in SET amps, when you connect them to a speaker and listen to music. What a novel concept. I never really liked the sound of a meter anyway.
I read an extremely interesting article the other day on the web. The premise was, that single-ended triode amps sound as good as they do, because of system synergy with single driver speakers(primarily). It seems this guy did some distortion analysis of SETs, and speakers. The SET generated primarily second order harmonic distortion, which was the bulk of the total harmonic distortion measured, very little other. The speakers measured primarily second order harmonic distortion also, with very little other. These distortion characteristics mainly differed in phase relationship.
His hypothesis was that if, by way of chance, the 2nd order distortion of each item appeared at 180 degree phase angles apart, the distortion would be cancelled out, creating the "straight wire with gain" ideal. If they appeared at 0 phase angle(together), they would add, and distortion would double. Since either case was unlikely, his solution was to determine which side of the 90 degree phase angle the intersection was, and phase the speaker wiring to place the phase intersection of distortion closest to the 180 degree mark that you could. This way if it was at 30 degrees and adding, by reversing phase, you could put it at 210 by reversing phase and achieve a high level of distortion cancellation. If you leave it at a phase angle close to 0, then the distortion adds - bad. Since the majority of this occurs in the 2nd order on both items, once a certain percentage of this 2nd order cancellation occurs, the THD is dramatically lowered for both items as a package.
This is his reasoning for why SETs have the remarkable clarity they do, while measuring in distortion %, the way they do.
His finding was that by switching phase connections on your speakers, you could find out which way sounds better(reduces distortion) on your system. You don't have to use measuring equipment, just your ears. The goal is to have the amp distortion cancel the speaker distortion as much as possible with your equipment.
This is not possible with push-pull or SS amps, as their distortion characteristics are comprised of many orders of harmonic distortion, and are not coincident with the distortion characteristics of speakers.
The effect is somewhat lost on multi-driver speakers because of the phase changes induced between drivers by the crossover network.
So, while I cannot personally attest to the veracity of his findings, it sounds like a very interesting explanation for the apparent dichotomy of SETs measuring poorly and sounding great. The amp is not used to drive a measurement meter. It is used to drive a speaker, and as such may exhibit different behaviors when doing so, than some lab freak with a meter may even realize.
Maybe there is no "euphonic" distortion in SET amps, when you connect them to a speaker and listen to music. What a novel concept. I never really liked the sound of a meter anyway.