Popular Mechanics - Plus ca change

[Andrew Gabriel]

Diming halogens reduces the bulb wall temperature, and below a cut-off temperature (which I can't remember, but I have in a book somewhere), the halogen cycle stops. However, it also reduces the filament temperature, which significantly reduces filament evaporation in the first place. To get bulb blackening, you have to find a bulb with a geometry and power rating such that the bulb is below the halogen cycle temperature whilst the filamant still has significant evaporation.

In early halogens, there was a third factor - the max pinch seal temperature, above which the seal would fail where the conductors were passed through it. Designing a bulb to operate above the halogen cycle temperature but below the max pinch seal temperature in varying installation conditions of cooling (or not) gave quite a small operating window for correct operation of a bulb, and dimming it would often take it outside this.

I believe improvements in materials used mean that there's a much larger operating window for current halogen products, and the chance of going into an operating region where blackening can happen is significantly reduced.

[docholliday93]

I remember doing some research on this because we were looking at halogens for use in fibre optic light pens, and coming to very much the same conclus ions. (Looking around the lab today I notice that most of our microscope il luminators use halogen bulbs and allow dimming, so there's another example. ..) I was interested when doing the research to find that the popular impressio n that the bulb life is extended because the tungsten is deposited back on the filament is largely a misapprehension - although the material is deposi ted on the filament it doesn't in general deposit uniformly, and the advant age gained is through being able to use a smaller bulb with higher pressure without bulb blackening. I don't remember any of the papers I looked at mentioning the pinch seal te mperature problem, which suggests that the manufacturers were probably not broadcasting that information so much...

[Theo Markettos]

Do you mean under-running as in under-volt, or the usual triac chopper?

I suspect they didn't have domestic-scale triacs in the 50s, so one possibility is to under-volt via a transformer with several taps. But under-volting is quite inefficient - more heat and less light. So that may be a motivation for switching filaments rather than switching transformer taps.

Theo

[polygonum]

Really, either. In the example you Effectively have a 50W and a 100W. So you can have three different light outputs with more or less the same colour temp/CRI. As soon as you use any form of dimming the colour whizzes towards the red end and CRI plummets. The lumens per watt also drop (as recently discussed).