Having been involved in all the amp trials, I can say that the root problem was not the separate circuits AC feeds. The ARC 160M could not properly drive the YG cleanly on the "good side", and the Luxman, being a stereo amp was on the same circuit as the rest of the system. So was the REF75SE. They just didn't have enough muscle or control for the challenging YG. Further, the m2tech amps were run as monoblocks on the separate circuits and there was zero difference in performance with them, neither flashed their peak LED indicators with up to 440 w on tap, but they also don't demand a large draw from the wall. The ARC REF75SE, also stereo, plugged into the same circuit as the front end, had all the sonic struggles, compromises and limitations heard on the 160Ms run on separate circuits. What was a problem was that with the Ampzillas in place on separate circuits, lifting ground on the LEFT channel on the TVC caused the two B- fuses to blow on the RIGHT channel amp. Now, this happened with only modest ground noise coming through the speaker, so that ground difference caused a surge of current through the right amp's power supply, which did not get commensurately amplified through the audio path. I have to think through what was happening there. Both amps sounded clean up to the fuses blowing. Getting both of the amps on the same circuit as the front end eliminated this phenomenon.
As for the peak LED: This is much quicker to respond to transient conditions than a meter. ~300ms for a good meter; far less, even as little as a millisecond for an LED circuit. Ignore it unless it stays on for several seconds or you hear distortion. The Tron track that Keith plays for bass evaluation has, I suspect, bass content below the woofer response of the YG, which the amp nevertheless sees. At the volume level being played, you are going to see some flickering on that content even if you double or triple the amp power. Flickering should not be obsessed over. Your own amps without peak indicators are instantaneously clipping too. You just don't know it until it gets sonically egregious. Further, the amps were new and "stiff" around noon Friday when I first installed them for Keith. By 5pm or so, they were substantially more limber and flashing the peak indicator much less often. Totally normal. On other music and on analog, you don't see any peak flickering. Keep in mind, I could drive the same SPL in Keith's room with my 101db/w/m Zu Definitions with just ~16 watts. If Keith got Mac MC1201s, he'd only push that incipient clipping point on that track out 3db. This is the penalty of inefficiency. The Ampzillas are plenty strong for his room with the YG.
The Ampzillas sonically control the speaker well and punch through the complex crossover loads. Once limbered up, it is both musical and authoritative, and it's stable below 1 ohm. The amp is conservatively protected, both through active and passive (fuses) shutdown. I also have to note that for that Tron track, when I switched the L/R channels at the amps' inputs, the peak flickers followed the channel feed. The channel specific bass energy is just in the mix or master of the track.
Now, overall, Keith has to live with the problem of not having a true grounded outlet anywhere in the house. It's not what any powerful amp really wants to see, especially one fully balanced. I'll say that the m2tech amps sound smoother and not sterile at my house (they aren't my primary amps) but Keith's AC did not compromise their ability to dynamically control the speaker and energetically power it. In fact, the m2techs drive a bigger soundscape on the YGs than the Ampzillas, thus far. That may change. Overall, the Ampzillas are better, as I expected them to be. However much the lack of true ground might compromise a big amp's ultimate performance, the Ampzillas show none of the insufficiencies of either ARCs or the Luxman, which were simply underpowered relative to the speaker, in Keith's acoustic space. They also aren't distorting, harsh, amusical, and the resolution shortcomings of the original version sound well addressed in the Second Edition. Maybe Keith could find another amp with an nth degree more transparency. It's up to him if he wants to keep looking. I am quite sure that if found, without sacrificing other things the Ampzillas do well, the cost will be much higher.
The Ampzillas have a 2000VA PT, with lots of electrolytic storage. I expect them to audibly improve at a decreasing rate, over the next 1000 - 1200 hours.
Phil
