Ernst Dinkla wrote:
Could there be also some inherent balance in that lamp where the thinner filament burns with a higher output and the tungsten condensation on the lamp envelop dampens that output again?
Sorry, I don't know the details of QH lamp physics in enough detail to really know. I do wonder though, that if a portion of the filament is getting hot enough to be a measurable contributor to the visible output, that it would also be very close to burning out.
Till heat brings the tungsten back to the filament.
The reason that QH lamps burn out eventually, is that the tungsten is not evenly re-distributed when it condenses back. So eventually the filament develops thin spots. I would guess this depends on the filament geometry to a large degree. Typically this is a coil or coiled coil, so I guess this doesn't help.
You mentioned that the lamp is glued into its socket, is the window where the light is coming from quite small? The glue keeping the lamp isolated to bring it on temperature fast or the other way around creating a heat sink?
Hard to say. The lamp itself is much like any small "grain of wheat" lamp, a rounded end cylinder. The dome of the lamp is exposed, while the body is glued into the support. See <http://www.argyllcms.com/i1pro_lamp.jpg>. The Spectrolino is quite similar, although about 2/3 the size. Graeme Gill.