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

Re: replacing CF bulbs



Wayne is a smart guy, so I hate to disagree with him.  I'm not sure
that Wayne Jones and I disagree on anything much concerning replacment
of PCs, but our recent posts seem to suggest otherwise.  I think we
agree on the basics but would give slightly different counsel to folks.
 So let me go through a few of the comments and see if I can't clear up
the agreements ;-)

> Wayne wrote:
> 
> > There is no way to predict when an inividual lamp will have to be
> > replaced.
> 
> Scott wrote:
> 
> True, but the quality of the name brands is quite high and most bulbs
> will reach the rated life span if used on cycles of no less than 3
> hours.  They will last (keep burning) longer if the cycles are
> typically longer,m say 10-12 hours "on."
> 
> Wayne writes:
> 
> There is a wide variation in lamps even form the same manufacturer.
> Manufacturers will tell you the expected life span or lumen
> maintainance
> of a lamp but they will not specify it. A suprisingly large
> percentage
> of lamps fail altogether within a few hundred hours and a suprisingly
> large percentage far exceed their rated life. The only way to be sure
> of
> replacing lamps in a timely manner is to get a Lux meter.

I must agree that some die earlier than others, but generally, the
quality is pretty good and the rated life will hold for about 80% of
the bulbs -- but I'll have to double check that standard.

A lux meter, if it is weighted for PAR rather then human sensitivies,
will certainly tell you just how much useful light is coming from a
bulb.  If you know just what your tankful of plants need you can tell
just when a bulb is ready for use and when to replace it to maintain
precisely x amount of light for your plants.  You'll need a very
special lux meter for you the bulbs in your UV lamp ;-).  But if you
use a PC for a year, you won't have much of a problem.


> Mostly the rated life of a lamp is not all that important. It is the
> lumen maintenance that is of most concern. Nobody cares if a lamp
> lasts
> for 40,000 hours if it only produces half it's initial output at that
> time. I would tend to not want to use a lamp if I could not find out
> it
> what the expected lumen maintainance would be. Major manufacturers
> typically provide this info with every lamp and believe me it makes a
> difference in the frequency of lamp changes and the cost of operating
> a
> system. PCs are not cheap and knowing the expected lumen depreciation
> before you buy should be something of great interest. Lumen
> maintenance
> is a critical piece of info in sizing a lighting system and
> determining
> the freqency of lamp changes.


Well, I agree.  Check with the manufcturer if you can, rated life
doesn't always mean you will have 80% of (post burn-in) output for that
duration.  But don't forget to place a value on your own time and
energy.
> 
> Now that I think about it, I think I would size a light system based
> on
> the lumen output at 40% of rated life as that is a piece of data that
> is
> often provided by lamp manufacturers.  So, a 55 watt PC gets changed
> after 5200 hours and a 30,000 hour T8 gets changed after 12,000
> hours.

Most folks going for a Boothian or Barroque high-light tank will have
to base total wattage on how many bulbs they can fit on their tank. 
PCs give a few more options in many cases.  People using the slow and
Walstady method might have even more options.

Never having too little light means always having too much light, and
vice versa.  A middle road will bring you somewhere in the, well, in
the middle.  You have slower growth on one side (never too much light)
and higher risks for algae on the other (always too much light).  I've
always subscribed to the middle road:  try to get a system a little
bigger and then replace bulbs less often.  But that's not the right way
or best way, its just one of ways you can trade off risks and benefits.

I suggested using half the rated life if you are nervous, or whole life
if you're trustworthy.  If you go by plant performance, you might run
your bulbs for several years. :-)  I think many folks have found that
this sort of duration has not been problematic.  Use a PC for a year at
least, maybe two.  Some folks use them much longer without adverse
results -- especially if you rotate several bulbs rather than change
them all at once (you must have more than one bulb lighting your tank
to take advantage of this, of course  :-)

> Since the PC loses more light than the T8 it should start off with a
> somewhat higher initial output that the T8.

It doesn't lose more light, other things being equal, it loses it
faster, it has a steeper curve. It will be replaced (or fail) more
quickly, but if you pick a middle road, it will be too much light half
the time and too little half the time, just like a T8.  The only
difference will be how often you restart the curve.

A PC is a high current bulb (like a VHO).  That makes its filaments
want to vaporize faster, coat the the inside of the tube faster, yield
to metal fatigue sooner.  T8s are cooler, lower current, bulbs
originally designed to make lighting maintenance cheaper in office
buildings -- BTW, the ones with green metal endcaps are "lower mercury"
meeting new environmental standards -- they are also a bit dimmer than
their "high mercury" older siblings.
> 
> Scott wrote:
> 
> All flouresecents lose about 10-20% of their acdtual lumen output
> after
> their intitial burn-in period of several hundred to about a 1,000
> hours.
> 
> I write:
> 
> This is too sweeping a statement. Some do some don't.

If they don't it's only because 1) they have already been burned-in or
2) they are rated for lower lumens.  Maybe a T8 has an even shallower
curve.

Lower current bulbs, other things being equal, should have a shallower
curve than high current bulbs.  There is more below on all bulbs
dimming as they are used.

T8s will dim more slowly than PCs, last longer and generally let you
fit less light into a given space.

> Scott wrote:
> 
> All flouresents are brighter for the first few hundred to 1,000 hours
> - -- after this steep drop-off in brightness, the drop-off is much
> slower, but steady and persistent until the bulb fails.  Most makers
> rate the lumens for the output *after* the initial burn in.  Some
> even
> rate it lower than that and then tell you the bulb keeps 90% (or
> whatever high percentage) of its output all the way until failure.
> 
> I write:
> 
> The lumen  rate of lumen drop off lessens with age. The first 100
> hours
> is the worst. Lumen depreciation is much less significant in the last
> half of a lamps life. All lumen maintainance curves I have seen look
> like this.

So I think we agree, they all lose and the more they lose the slower
they lose.  And this applies to T8s, T5s, Twhatevers.  And life until
flash-out is shorter and dimming rate is speedieer for higher current
bulbs.  A bulb type that lasts much longer will have a shallower curve
because the curve is strecthed out over a longer time.

> 
> Scott wrote:
> 
> If you buy from trustworthy and knowsledgeable vendors that come
> highly
> recommended by a wide variety of experts, you can follow their advice
> and forget about all these details.
> 
> I write:
> 
> I have never seen so many people scammed as in the the aquarium
> retail
> business over "special" lamps. It is better to figure these things
> out
> for oneself and then seek out venders who don't lie.
> 

Well I think more folks have been scammed on air pumps.  But really,
again, no disagreement here.  However, most folks, can't won't, or
don't know how to gather all the right info -- or, more often, don't
have the time and inclination.  [Which reminds me of the dissertation
committee back at the UCSB philosophy department:  Won't read it, Won't
understand it, and Won't Care.]  And that's okay, you can still do fine
if you you trust the right people.

As for vendors, I was thinking more of AH Supply as a trustworthy
vendor that many experts recommend.  If you don't want to fuss too
much, you can call Kim Bryant and get very good advice.  There are
others that are trustworthy and helpful.  

Scott H.

__________________________________________________
Do You Yahoo!?
Yahoo! Greetings - send holiday greetings for Easter, Passover
http://greetings.yahoo.com/