Bias/DC restorer explained

Got this from one of Bob's 180 watt tube amp auctions on ebay. A tube DIY guy explains in his terms how he duplicated Bob's design:

Hi Bob...thank you for responding to my confusion concerning your bias scheme,,but first I want to say to the bidders:....BID AWAY, BOYS AND GIRLS.I HAVE INDEPENDENTLY DUPLICATED THE ESSENTAL FEATURES AND CAN ASSURE YOU THIS IS THE ONLY WAY TO OPTIMIZE FOR AUDIO QUALITY AND LONGEVITY. SO BID WITH ASSURANCE MORTGAGE YOUR HOME IF YOU CAN GET ONE OR EVEN HAVE A JOB LEFT TO THINK ABOUT IT. Ok, I think I know where my confusion lies----when I say sliding bias I mean changing the G1 negative voltage....when you say sliding bias you refer to a particular TECHNIQUE to change G1 neg. volt. Consider the neg source half or full wave, there will be a smoothing cap followed by a an RC stage ,followed by maybe another RC smoothing stage..Now follow with a resistor voltage divider WITH NO CAPACITOR ..the center connection now feeds the bias pot..One leg of the volt divider is paralleled with a device that changes resistance with input drive, so now there is being fed to the bias pot a filtered neg. DC, that is decoupled from the filter capacitors, that can change without the lag that a capacitor would cause. It must have been pretty darn fast because I could hook the amp to a scope,set idle to 15ma.. suddenly zap a sinewave to it big enough to drive right up to just before clipping, and the sine wave would instantly appear without gradually climbing up with flattened tops. That was good enough for me!!!!! I have bailed out of tube equipment, am now making chip amps from the latest audio grade offering from National Semiconductor.. they sound very "tubelike" to me (and many others) Just in time because I am getting too old to manhandle heavy iron. Nevertheless I still maintain a keen interest in new ideas in tube technology. I would love to see a schematic of these amps..could you send me one to my email? The 6al5 is a high perveance duo diode.I'm trying to imagine how it is implemented..Neal Mar-21-09

Prequel to the above post:

Q: dear Bob....thank you for hanging in there with my audio-amatuer forays into mysteries of tube amps. I don't understand your bias stablization...it doesn't jive with your comments which I quote here "the dc restorer eliminated the need to idle each output tube at 50 watts yielding instead an idle power of 12 watts per tube....It works by keeping the dc componant on each output tube grid the exact value throgh the entire audio signal swing allowing perfect performance up to and even beyond clipping" Well, if by dc componant you refer to the negative bias grid voltage this is exactly how I describe my sliding bias scheme. If your bias scheme doen't shift the negative bias as the amp moves away from idle it will soon run into class B operation and really gross distortion. Your previous response says the bias doesn't slide , but if by slide we mean changing the bias voltage , then it MUST slide to my way of thinking...I'm confused cordially neal Mar-21-09

A: Hi '541, ....It jives perfectly with my comments. A sliding bias scheme has too much delay in responding under dynamic conditions OR it has too much control voltage ripple modulation for my taste. Such a scheme must ALWAYS be a compromise between the two. That's because it will have a detector diode, an integrating capacitor, and a time constant comprised of the final smoothing RC filter. The topology is often similar to a meter drive circuit. On the other hand, a DC restorer operates such that the DC value of the waveform on the grid remains where we want it even in the presence of asymmetrical, haversine signals. And its delay time is essentially zero because there is no filter capacitor to charge and therefore no time constant. Works far better, and is far more elegant than a simple and coarse sliding bias circuit. Bob Carver

Silly wabbit doesn't everyone do their flux capacitor like this?