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

Re: hot PCF



>Scott H. wrote:
>"Then general rule on fan placement is, place the fan so that it blows >first
>on the hot end of the bulb and blows along the length of the bulb.  This
>prevents the chance of over-cooling the "cool" end of the bulb and reducing
>efficiency
>[snip]"
>
>Bill Wichers wrote:
>"The end with the wiring is the hot end. Ideally you would want to have the
>fan blowing on the cool end and exiting out the hot end, with the hood
>acting as a duct to carry the air along the tube. In this way the entire
>tube would see a nice, cooling thermal differential."
>
>Flip a coin guys? Which one is the better way?

Ahh, you cought that too, eh? ;-) That's what I get for not reading *all*
the messages in a thread before posting something.

Max efficiency for CF lamps I believe is somewhere in the 90-100F range,
although I'm not certain on that so please correct me if I'm off. Basically
you actually *want* them to run warm for best performance. In my case,
which I believe to be similar to the original posters, the CF lamps are
mounted in relativly unmodifed AllGlass "Deluxe" hoods. "relativly" here
means not modified beyond the addition of a 2 3/8" fan on the right rear
side of the hood (2 3/8" is a *perfect* fit for these hoods, BTW, going
from just below the top to just above the lip on the bottom), and mounting
the CF light stuff inside. My ballast is mounted externally -- all the way
on the floor -- and runs two identically configured hoods.

The fans that I use are NMB micro-boxers, rated for about 20-22 CFM each at
12 vDC. They are pretty quiet by my standards, but certainly not silent and
I suspect many list memebers would find them noisy. I work around big
commercial file servers and routers all day so I have funny fan noise
stanrards (WHAT WAS THAT!? !@^&%! FANS! ;-), but that's another story...
They are durable, aluminum fans that hold up well and can *blow against
some back pressure*. Now we're getting to the interesting part....

A typical hood has two sets of vent slits on the top. One one each end. In
my hood, I have the reflector mounted such that there is a roughly 1/2"
(height of the molded-in standoffs in the hood) gap between the reflector
and the top of the hood. I should mention that by "hood" I mean the strip
light part on top only. I have the end near the fan sealed off. The result
is that the fan blows directly over the "hot" (socket) end of the bulb,
down the length of the bulb, and gets to the cool end. Most of the now hot
air goes out the first vent right there, but some continues on in the gap
above the reflector and exits through the vent back by the fan. This
arrangement keeps my fixtures pretty cool. The trouble is, after a while
dust builds up on the fan, finger guard (important!) on the fan, and the
vents. The light and glass get a little but I clean them during water
changes.

After all the dust and the long air path, and the vents, the fan is blowing
against a fair amount of back pressure. I haven't set up a manometer to
check the exact amount, but you can certainly hear the fan fight it when
you put the hood down so it doesn't just blow out the bottom. I would
estimate maybe 5-8 CFM actually comes out the vents and thus is actually
flowing along the tubes and performing a useful cooling function.

In my configuration, the entire tube runs about the same temp, which is
about 95-110F, depending on room temp. If I switch the direction of the
air flow, I end up with a slightly lower but still very similar overall
temp. Basically I suspect that at the low air flow rates one is likely to
actually achieve in a hood against the back pressure created by the vents,
there isn't enough air movement to create significant differences in the
*overall* tube temp. The base is where I had plastic-melt problems, so I
would try to maximize the heat release at that point.

Hmmm. I guess I pretty much just said it doesn't really matter which way
you blow the fan along the tube.... I do think you'll get slightly better
cooling in the way I describe, but I doubt there will be more than a few
degrees difference either way.

Hope this cleared up the confusion :-)

     -Bill

*****************************
Waveform Technology
UNIX Systems Administrator

*****************************
Waveform Technology
UNIX Systems Administrator