What do you think of this preamp design?


I've been on the hunt for a preamp and came across the Audio-gd C3 final version:

http://www.audio-gd.com/Pro/pre/C3End/C3endEN.htm

It comes with schematic of the signal section. Does it seem to be well executed? Any obvious or subtle design flaws? It has had some positive reviews by owners on the web.
dracule1
Ngjockey, if the Audio-gd preamp sounds like any preamp made by PS Audio, I'll pass. I never liked any of PS Audio amps or preamps that I have heard over the years.
Apparently, the neat little thing about this preamp is you can swap different gain modules that were designed to sound like other classic solid state preamps by Mark Levinson, Accuphase, Music Fidelity, etc. As far as I know, Audio-gd does not have any gain modules that suppose to sound like PS Audio.
After a little bit of research, this I/V volume/gain control is based on a "Gilbert cell" and is available in CMOS in something like a AD835, although I have no idea of what is used in the Audio gd, PS Audio and Ayre VGT (KX-R) circuit. The advantage of this circuit is consistency through it's range, and allowing much higher impedance than the "standard" 10K to 250K ohms. Any claims of audible superiority I'll leave to the manufacturers and your own ears. Audio gd extends the concept with it's CAST-like I/O. Krell has a US patent but did not invent the technology. Seems like the most common use for this type of circuit is in phone systems.

We'll be seeing more of this type of control.
The Audio-gd C3 input impedance is 47 kOhms. I asked if it can be changed to 100 kOhms, and they said it was no problem. I wonder why they don't use higher input impedance.
It looks overly complex... A very competent solid state line stage could easily be built with about 1/3 the components. The op-amp in the feedback loop looks to me like an outright design problem...
The opamp is obviously an integrator (DC servo) used to set the quiescent conditions of the "BPM-7110" module, which appears to use current feedback, and a very high output impedance (current drive). The buffer stages are used to interface with non-CAST components. Really not so complicated.

Atmasphere, there's certainly not enough information here to trash this design so thoroughly, and English is obviously not his native tongue. While this is also not the way I go about designing a linestage, it's about as thoroughly "Power Paradigm" as it gets. To me, his stuff looks interesting, unconventional, and definitely reasonably priced compared to many.