From: Phil Hobbs
Subject: Re: Halogen bulb drive
Date: Thu, 03 Oct 2002 17:43:12 -0400
Organization: IBM T. J. Watson Research Center
References: <4LPj9.email@example.com> <firstname.lastname@example.org> <3D9CB0F7.email@example.com>
NNTP-Posting-Date: 3 Oct 2002 21:43:14 GMT
X-Mailer: Mozilla 4.61 [en] (OS/2; U)
Charles Edmondson wrote:
> So, some genius put some iodine in the bulb, used quartz to handle the
> high temperatures, and cranked up the temperature of the filament. The
> evaporating tungsten combined with the iodine, so it didn't concense on
> the envelope. The high temperature of the filament broke down the
> tungsten-iodine molecules, re-depositing the tungsten back on the
> filament! So you got long life, more light, better efficiency, and all
> sort of other goodies.
> The problem is not all the tungsten is re-deposited on the filament.
> The support structures are also pretty hot, so some of the tungsten is
> deposited there, so eventually you still get filament failures.
Right. The quartz envelope is there to handle the very high gas
pressure inside the bulb, which retards the diffusion of tungsten iodide
vapour away from the filament, so that most of it gets redeposited right
away, near where it came from. This high gas pressure is why tungsten
halogen bulbs have *much* longer life than an ordinary bulb run at the
same filament temperature. The envelope has to be thick to stand the
pressure, and small enough for its temperature to stay above the
sublimation point of tungsten iodide. The allowable gas pressure is
limited by safety and thermal losses due to convection.
The tungsten iodide recycling mechanism is just to keep the (very small)
envelope clean, as you say. Unfortunately, there's no known way to make
the metal redeposit itself selectively on the hot spots of the filament,
where the evaporation rate is highest, so the hot spots eventually neck
down and the bulb fails.
This happens fastest when the filament is thin, as someone else pointed
Still, it's a pretty neat idea.