On Sat, 29 Nov 2014 11:33:01 +0000 (GMT), "Dave Plowman (News)"
wrote:

In article ,
Andrew Gabriel wrote:
I'd recommend this too. Makers tend to lie through their teeth when
giving light levels from the latest whizz bang technology.

Make sure you run a lamp for 100 hours before measuring.

Thought that was to allow a CFL to achieve full level on a cold day? ;-)

I saw the smiley but Andrew was referring to the fact that a low
pressure mercury vapor discharge fluorescent tube starts off at around
a 110% of its 'design lumens' before swiftly dropping down to 100%
during the first thousand hours or so of an 8 to 16 thousand hour
design life when the output will have dropped to something like 80%
before the drop is considered to be end of life.

The light output plotted against time follows an exponential curve
where most of the initial drop to 105% takes place in the first
hundred hours or so. The mechanism for the loss of efficacy in such
lamps, istr, is largely to do with mercury argon 'poisening' of the
fluorescent coating.

White LED lamps which use a blue LED to excite the phosphor coating
must also suffer some loss of efficacy in the flourescent material
used but, with lifetime ratings typically in the 30 to 50 thousand
hour range, the rate at which the efficacy of the phosphor coating
drops must be far lower than in the case of a mercury vapour
fluorescent, presumably on account of the lack of such a 'poisening'
mechanism.


Some lamp technologies are slightly brighter when brand new.

Yeah, fluorescent lamps as described above.

All lamp technologies slowly get dimmer with age, so you will get a
misleading reading if you compare one newish lamp with one near end
of life.

That's true enough. I think the least effected by ageing during their
rated lifetimes are the standard gas filled tungsten filament lamps,
particularly when burning 'cap up' in pendant fittings.

I haven't seen any lifetime light output plots for phosphor coated
blue LED lamps but I assume these are also based on the 80% of design
lumens end of life criterium.

When I did my comparison test using a digital camera, comparing a 12W
LED against an existing 20W CFL, I felt pretty confident that the CFL
had only clocked up 1 or 2 thousand hours so was within a percent or
two of its design lumen output. I took care to allow the lamp to
thoroughly warm up to its final steady state operating temperature (10
minutes or so warm up time). The only "unknown" in this case being how
much brighter than design lumens a factor, if any, applies to such LED
lamps.

Since the old CFL has been put to one side, I can repeat the
comparison say every 6 months or so and look for such an initial lumen
efficacy drop mechanism in the LED lamp (the unused CFL should provide
a constant reference source to ensure meaningful results).

Another thing to bear in mind with modern LED lamps (apart from the
effect of colour temperature on the efficacy of its lumen ouptut
rating) is the tungsten filament lamp type used as the comparison
reference.

For example, you may see them compared to a "60W 810 lumen" tungsten
filament lamp. This is only true for the American 117v 750 hour rated
incandescent GLS lamp which is equivilent to a 75W 1000 hour 240v
tungsten filament GLS lamp, not a commonly available lamp size here in
the UK.

As it happens for once, a difference that makes the efficiency
improvement of LED based lamps over the 220 and 240 volt tungsten
filament GLS lamps used here and in Europe even greater than that
experienced by our Yankee cousins. :-)
--
J B Good