> I considered trying to build some kind of AGC for
it, [...]
Perhaps an incandescent light bulb might help out there. =3D)
Perhaps.
Plese don't tell you you didn't get the reference.
Okay.
But it
requires enough current flowing in the feedback divider to
warm the bulb's filament significantly. I'd rather avoid
dissipating that much power - not to mention that I'm not sure the
op-amp I was using was capable of it.
There are some pretty low current light
bulbs (such as ones that glow
to their specified intenisty at 20ma (with 6V across them).
120 mW is more power than the whole circuit was drawing, I think (this
was quite a while ago).
Yes, but you certainly don't ahve the bulb running at full brightness. I
woudl think arund 50mW would be more like it.
If this is an experimental 1-off circuit, the power consumed is not that
important (probided your bench PSU cna handle the current, I've yet to se
oen that couldn't handle that!). If it's a 'permanent' or production
device, you need some way of stabilisign the amplitude.
-tony