On Tue, 12 Jul 2005 20:19:31 -0400, "Frank"
<Frank@hotmail.com> wrote:

>sorry for top posting but your text is long, but very useful.
>I'll need a 120mm fan, space is plenty for 2 fans and my definition of quiet
>is like having all your pc fans off, boot into bios and all you can barely
>hear are the HD and the PSU. If temps are below 50-55C without fans, I'll be
>glad, but if not, i would have a 120mm fan turn very slowly just to keep it
>below that. But when I need to, I want a fan that will push enough cfm to
>pass accross my rad which is a good 1.5in think. I was planing on using a
>simple 3 postition switch with resistors to control the 1-2 fans. I could
>have Off - Mid speed (7V) - high speed(12V).


That is what I"d do differently... perhaps shooting for Off/
5-6V / 9-10V. With two relatively unimpeded 120mm fans you
should never need run them at 12V, including gaming.


>I also used an aluminium reservoir, (
>http://www.noiseisolator.com/wt-tb.htm ) to help in cooling. I'll pass the
>water from pump-cpu-chipset-vga-reservoir-rad-pump. That way, the biggest
>heat will be absorb by the reservoir and hopefully, I'll manage to run my pc
>at stock speeds with no fans.

That will have the potential to damage the system long term.
The parts with water blocks are NOT the only parts that need
airflow. You are going too far in two opposite extremes.
You should not be trying to turn the fans off, then
alternating with full speed fans. Instead, shoot for
ultra-low RPM fans at idle (turning only one supplimentary
fan off if desired) and still-RPM-reduced fans at full load.
Your radiatior is still going to be the largest heat
shedder, regardless of the reservoir. The reservoir simply
doesn't have the surface arear to water ratio necessary to
be very effective for that purpose.

Putting it another way. Suppose that at stock speeds &
idle your PC only produces 100W of heat. 100W of heat left
stagnant will damage parts. Perhaps you don't actually mean
"no fans"? At an absolute minimum you should have the power
supply fan and one chassis fan always spinning.


>I guess im better off to start off with no fans at all,

That is not necessarily a good idea. Start out with a
properly fan cooled system then only reduce fan speed when
you can confirm it will happen with no thermal thersholds
exceeded on parts you can't temp-monitor as well as those
you can. Remember that CPU and other chips that take temps,
are not the thermal problems. Thermal problems occur in
small surface mount parts and capacitors. You will have to
manually check those, as they are cooled by copper on the
circuit boards, copper that is cooled by fan-forced airflow.


>see how it goes and
>try to use a temporary fan just to see how cool i can get it and see if a
>really need these high cfm.

It might overheat at one extreme and be unncessarily loud at
the other. Systems that don't even have water cooling don't
necessarily face either of these extremes.


>I might be satisfied with a panaflo M but like
>you said, some say its still loud even at low rpm. Ill see if can have a
>shop around hear have me listen to the noise it make just to see. When
>gaming, noise would not really be a problem, but when idle or like watching
>a movie, I need dead silence.
>(sigh) Not easy to have both silence and performance.

It will do you no good to listen at full speed, the point
was to have the greater flexibility in RPM adjustment, but
still not running it at full speed. Take a ready-made (or
homemade) fan controller along with you to test on the
target fan if possible. Even so, this was within the
context of having only one fan. WIth two, you should be
fine with the "L" speed, and if they aren't sufficient the
answer is getting air passages to a lower impedance.

Ultimately, you may be better off with a variable speed fan
controller than trying for a 3 speed via resistors. Either
that, or implement the lowest of the 3 speeds with a
regulator or diodes for the voltage drop instead of
resistor(s).