[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: halogen lights

>>> Do you know the PAR
>>> rating?
>Why don't you humour me and just tell me
>the PAR rating of the lamp if you know

Some hard data to settle (or stir up) the discussion: a typical narrow-beam
60 watt halogen spot (Philips Masterline Par 16 (60PAR16/H/NSP) dumps
208 footcandles in a 12.7" diameter spot at 6 feet, according to the
data sheet at


This translates to about 730 lumens. The color temperature is 2950K and
the spectrum is a perfect Planck curve. Thus we can readily compute the total 
photon flux in the 400-700 nm range, it is 1.28 10^16 photons/sec, or
about 2.13 10^14 photons/sec/watt. A typical high-efficiency tri-phospor
normal fluorescent creates about 7.9 10^14 phot/s/watt in the same 
wavelength range, and a typical plant light (grolux) 5.6 10^14 phot/s/watt. 
So the three bulbs have PAR efficiencies that scale as 1 : 3.7 : 2.6 for 
halogen : tri-phospor : grolux respectively. Of course these ratios are
valid only in the case where the fluorescent tubes are mounted under
"perfect" reflectors, at least as efficient as the built-in halogen
bulb reflector. If a "typical" fluorescent fixture reflector has an 
efficiency of 50% (I think I saw a figure like this in the krib archives)
the ratios would be 1 : 1.85 : 1.3 instead. Not too bad for the halogen,
if one's willing to live with that extremely narrow bean. A wider bean
halogen (50PAR20/H/NFL30) disperses the same amout of light over a 10 
times larger area.

Btw, the power conversion efficiency of the halogen bulb (emited watts
divided by input watts) is only about 0.07.


>>> In plant growing Lumens/watt can
>>> be used to measure system
>>> but has almost no bearing on the PAR
>>> rating of the lamp.
>>That's a strong statement. They are
>very *closely* correlated for most
>>lights, and only really come apart for
>lights whose spectrum has been
>>efficiently tailored to human scotopic
>sensitivity -- for example, CW

More hard data: the correlation coefficient I get from 17 bulb types
(between lumen/watt and PAR/watt) is 0.64, suggesting that hardly any
correlation exists. In fact only 3 out of the 17 bulbs show a systematic
trend (high PAR/watt *and* lumen/watt), the remaining 13 show very 
similar values of PAR/watt (within a factor 25%) but lumen/watt values 
ranging from 20 to 80, a factor 4 (or 300%). The correlation coefficient 
for these 13 bulbs is 0.38, suggesting in fact that no correlation exists 
for most bulb types.

-Ivo Busko