Gainclone as AC regenerator

Anyone ever wire up a gainclone bridged or otherwise with a transformer coupled output to provide AC regeneration? You just get a power transformer and wire it in reverse to step back up to 120V. You could also regulate the GC power supply to provide regulation of the AC output. As long as you power a CD player or something with little power consumption, the whole project remains simple and easy on your electricity bill.
 
- It would let you provide voltage regulation for any piece of gear.
- It would let you provide balanced power " " " " ".
- It would simply provide further line isolation...
- It would let you change the frequency of the output AC to whatever you decided sounded the best. The Mark Levinson N32 internally regenerates AC at 400Hz. PS Audio's regenerators let you choose the frequency or do weird combinations.
 
I think using a high frequency such as 400Hz should reduce ripple starting at the source. Since the electrolytics would not have as far to fill up each time they received a pulse from the bridge, do you think the overshoot on the pulse would be reduced? How about a reduction in motorboating? Also, since the frequency is now 400Hz instead of 60Hz, the motorboating would occur at 800Hz instead of 120Hz, and by moving the point of hum higher, it should be easier to filter out.
 
from http://www.marklevinson.com/products/review.asp?cat=pre&prod=no32&review=4

AC power regeneration refers to the DC supply feeding the voltage gain stages. This supply is derived from an AC source generated within the device itself. This technique is expensive but very effective, according to Madrigal. In the No.32's Controller, a 400Hz oscillator generates the AC source for the preamplifier's audio power supplies, replacing the 50/60Hz reference provided by your local utility. One advantage to operating at 400Hz as opposed to 60Hz is that the transformers operate more efficiently and thus generate less heat. Another advantage is that the filter capacitors are charged more quickly, yielding a smoother DC supply from the filter caps. Equivalent filtering from a 60Hz supply requires larger capacitors, and thus higher ESR (Effective Series Resistance) at high frequencies. Because 400Hz oscillators and power-supply components are commonly used in aircraft electronics, high-quality parts are readily available.


I think they use 400Hz in aircraft to lower the size and weight of transformers.
 
I wonder what is the technical explanation for using an internal power supply of 400Hz. If any, I could say that the electrolitics could be charged more times per second... I'll think on that.

Peter, what trannie did you use?

I used something like 9v/220v/150VA IIRC. It seems the LM is not that happy driving a big inductor; maybe some compensation network is needed. As i felt the transformer to be a silly solution, eventually used the setup to generate 2x15v ac and this is what powers my opamp phono. I tried raising the frequency up to the kHz region and there was not that much difference to the sound. Not sure if it was an improvement at all.

And the mains cable type was still clearly audible, which probably should have been expected.
 
PS Audio gives some info how differend frequencies affect the sound:

It doesn't quite relate to my setup as i don't have transformers at present. I guess the caps don't care either way and the diodes may even generate more noise at higher frequencies.

why do it in 1/3 measure?

Maybe i'm just not mad enough. I have a prototype AVR based oscillator which can easily source out 3 phases with any phase relationship. A 6-channel GC still looks like a bit of an overkill :)