🚀 go-pugleaf

In article <6pKdnW_chZqtYmzUnZ2dnUVZ_j1i4p2d@supernews.com>,
	Phil Hobbs <pcdhSpamMeSenseless@electrooptical.net> writes:
> Andrew Gabriel wrote:
>> In article <fjlIl.34167$fe2.29061@newsfe06.ams2>,
>> 	"ian field" <gangprobing.alien@ntlworld.com> writes:
> 
>> In countries with different mains voltage, the ratio may need
>> adjusting, as filament lamp efficiency varies with the design
>> voltage. (The most efficient voltage to design a 100W lamp is
>> about 55V, given a reasonable life requirement, and as you move
>> the design voltage away from 55V in either direction, the lamp
>> efficiency drops.)
>> 
> 
> Interesting.  I can see it dropping due to conduction losses when the 
> voltage gets too low (so that the wire gets too fat) but what's the 
> mechanism for the drop-off at high voltages?

Designing a full range of filament lamps to operate over a
wide range of voltages is a challenge rather like trying to
solve a set of simultaneous equations when there aren't
enough variables, so you end up with compromises, such as
lower efficiency.

To make a 240V filament, you need a very long thin wire.
It has far too much surface area to get to the 2700K operating
temperature without radiating 100W away at a lower temperature.
The way 240V filament lamps are made to work is to double
coil the filament, so that much of the filament is radiating
heat back onto itself, and the effective surface area for
radiating heat away from the filament is significantly
less than the filament's real surface area. This isn't as
effective as using an optimum thickness filament in the first
place, i.e. you have just the right surface area to radiate
100W when it's reached 2700K, and you still have a larger
surface area for filament evaporation and thinning.

And yes, you're right about heat loss by conduction down the
lead-in wires becoming increasingly significant at low
voltages/high currents. With longer fragile filaments, the loss
from the ends is proportionally less, but you may also require
filament supports, which are additional routes to lose heat,
and you lose more by convection to the gas-fill.

-- 
Andrew Gabriel
[email address is not usable -- followup in the newsgroup]