<a href=showthread.php?s=&postid=6399476#post6399476 target=_blank>Originally posted</a> by TropTrea
An interesting thing one of my friends purchased three idential bulbs 10K's at the same time and installled the into his 6 foot fisture. Two bulbs gave a blue light than then the third which gave a more yellow hue. He tried swapping them in different positions to see if was the balast or wiring or reflector that was making the change but the yellow i\tint stayed with the bulb .
Gow this difference was not obvious to the point you would notice it on a tank with a single bulb but it was noticable with the bulbs right next to each other.
Simularly I was replacing a PC bulb on my 45 and noticed it was more purplish than its old mate. I then replaced the mate noticed the difference still there. So I switched the bulb positions and the bulb with the more purple color stayed that way, Note these were suposed to be actinic's but one was blue and the other was blue with a slight purple hue.
This makes me wonder how accurate any tests are when run only a few light bulbs. One engineer that worked for a lightbulb company told me that the basic tests on bulbs are run on a hand few from each batch. As long 90% of the test batch falls within 10% of the advertised specs then the bulbs can be legally shipped.
But what is 10% say for a actinic it should peak at 420 nm. if it is allowec a 10$ frequency shift could that mean it can peak any where from 388 to 463? or doe is mean that if it slisted at 1000 lumins at 420nm it can be anyshere between 800 and 1200 lumins? he never broke down to fetail these specifications.
Dennis
