Reply from Doug about the advantages of valves in commercial gear he's designed:
QUOTE:
Technical reasons be buggered!
It was purely sales/marketing, to appeal to the wankers who think choobs
sound better than solid-state (which was once true of course, back in
the '70s).
The Rode Classic-II uses a choob operating with about 20dB gain followed
by (shock, horror!) a transistor emitter-follower hung off the anode.
This gave a low-Z drive point for the high-pass filters and output
transformer, and dramatically reduced THD.
Circuit noise is defined by the choob's operating conditions; The
trannie contributes bugger-all noise.
The other one was the Rode NTK.
I'm proud of this beastie; It has an absolutely awesome dynamic range of
(from memory) more than 145dB.
A microvolt or so of noise, but capable of 35V RMS output voltage.
Self-noise equivalent to 12dB SPL, but can handle nearly 160dB SPL
without clipping the electronics.
I used the choob as the front-end of a closed-loop power amplifier (yes,
more bipolars) with a 120V rail.
The condenser mic capsule is directly coupled to the grid - i.e. no
input coupling caps.
A few other circuit tricks gives it extremely good PSRR which, combined
with a pretty clean PSU, means mains-frequency crap is way below the
broadband noise floor (unlike the noise spectra of most choob mics).
This thing's a brute; I connected its output directly to a 32-ohm
headset, and it was LOUD.
Mind you, it's still noisier than the NT1000 (JFET, <6dB SPL equivalent
noise).
After I left, Rode shoehorned my NT1000 electronics into one of their
older mics, which had a more sensitive capsule, and achieved 4dB SPL
noise.
With these seriously low-noise condenser mics, on a clear night you can
hear the curry percolating through the intestines of the family in the
next block.
/QUOTE