"Pszemol" wrote in :

"atomweaver" wrote in message
...
I cannot confirm the PAR or the equivalent lighting output, but the
50,000 hours service time is fairly accurate for LEDs.

I would guess it will depend on the type of the LED,
and on the condition it is used at (overcurrent, temperature).
We had already a precedence where some moon-lights were
worn prematurelly (less than a year) due to the manufacturer
supplying them with cheapo, unreagulated power supply driving
them at or above their maximal tolerable current values.
Yes, they were very bright, but they did not last long...


Sure. Poor engineering can _always_ overcome the most robust components
:-)

LEDs will fail in a "full off" manner, meaning unlike the spectral
decay you experience with MH or HO bulbs, you'll get the full power
of the LEDs rught up until the day they die. This is both good and
bad. Good, in that you'll get very consistent light, and an
incredibly long service lifetime. Bad, in that some day when the
lights _do_ fail, you'll have to scurry around for an interim
lighting solution, whilst waiting for your replacement/ reconditioned
LED (note, 50,000 hours at 16 hours lighting per day, means you'll
have this problem once every 8.5 years, on average.)

There are many different LED technologies and we do not know
what type of LED are used in this fixture. I have seen already
white LED made with the similar principle white fluorescent
tubes are made: native LED spectrum is UV and the LED lenses
are covered with white phosphors, re-emiting white light when
they receive UV rays from the LED. I could only expect the same
kind of decay of light spectrum and similar longevity with
this kind of LEDs.

You caught me! I assumed from the claim of 50,000 hours that this was a
visible specrum semiconductor light array, assembled without the use of
lenses (which, as you say, will degrade just like bulbs). Such would be
the only tech I'm aware of which could make the claim of 50K hours of
steady output. Still, its an assumption I shouldn't have made...

Assuming a white LED is made on a different principle of mixing
base colors to get white effect - with at least three different
color LED chips inside one "bulb" I could not imagine how can
you get properly spread light spectrum to imitate sun-light...
You would rather get light with three or four (depending on
number of led chips colors per bulb) peaks in the spectrum and
not spread spectrum like incadescent or good HQI/MH fixtures give...


I'd expect something more like 15-60 different selected wavelengths of
output, and thus a more complete spectrum.

Claims of low heat output from hese systems are accurate.
The load demand on a chiller may be significantly reduced.

Do you have any access to the efficiency of LED light data ?
Do they really produce more light per Watt of electricity used than MH
?

I know from direct experience how efficient they are in terms of an
energy source for pushing industrial chemical reactions. Spectrum of
output is much narrower than a standard bulb, but if the catalyst you
want to have run your reaction happens to be efficient at absorbing
light in the wavelength of output of the LED, then Yes, the LED source
is much more efficient, watt for watt. Same goes for visible light,
although the light fixtures are vastly too expensive at this point to be
reasonable for replacing typical flourescent fixtures for human use
(althoguh I don't doubt that they'll be the standard at some point in
the future, if they can get their costs down).
Being an industrial chemist, I don't know nearly enough about the
active absorption wavelengths used photosynthesis to make that same
assesssment of these lights for use in reef aquariums. The vendor claim
of 178 PAR vs. 133 for a 250W MH source is a starting point, but that
kind of measurement could be an inaccurate means of assessing an LED
source.

DZ