Jeff Wisnia wrote:

I wouldn't have believed that there'd be any light in the visible
spectrum at 1/10 voltage, but I just played around with a few small
"12
volt" automotive bulbs and my bench supply and they all put out a red

glow visible in dim light right down to 1 volt; and some even below.

I suppose it's because as the filament is cooler running at reduced
voltage, the resistance drops, so the current falls off more gradually
than voltage, and the power and filament temperature are decidedly
nonlinear functions of voltage.

%mod%