[...], I'm
considering an AGC circuit that doesn't use a bulb, but
instead uses a FET as a variable resistor,
Jim Williams (RIP) did a LOT of
research on this. The bulb
outperformed the FET by a gigantic margin. (just sayin'..)
That's not surprising - where the bulb actually works. But if the
period of the oscillator is long compared to the bulb's warmup/cooldown
time, it won't work. It also won't work if the response time required
from the AGC is fast compared to the bulb's
warmup/cooldown time.
Also, if the bulb's filament is coiled, as many are, its inductance may
prove to be significant at high oscillator frequencies - I don't know
how much inductance the typical bulb's filament has, but I doubt it's a
characteristic usually controlled for in bulb design.
/~\ The ASCII Mouse
\ / Ribbon Campaign
X Against HTML mouse at
rodents-montreal.org
/ \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B