Creating silent 2.55ghz P4 system... new duct ideas
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
Creating silent 2.55ghz P4 system... new duct ideas
(continued from discussion in this thread.)
My new idea was to buy two ready-made fan ducts here, and to install them as demonstrated in this simple diagram I scratched together:
Provided I find a place to machine the holes, I think this may provide highly effective forced-air cooling with very slow-running fans.
Any feedback or other ideas?
Leo
My new idea was to buy two ready-made fan ducts here, and to install them as demonstrated in this simple diagram I scratched together:
Provided I find a place to machine the holes, I think this may provide highly effective forced-air cooling with very slow-running fans.
Any feedback or other ideas?
Leo
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
For my own systems (including the ref P3 ~12-15 dBA system), I only use an intake fan if I have more than 1 HDD and can't keep the drive temps below 50C. Then I position them in a rack on the bottom front of the case, just behind an intake fan that I decouple mount and run as slowly as possible. The HDDs are suspended in elastic of course, and sometimes, side-sinked. The system cooling does benefit somewhat, but mostly it's for the drives, which usually drop ~5C or more just from this minute amount of airflow.
There's nothing wrong with having only exhaust fans. The air that is blown out has to be replaced -- it naturally gets sucked in through whatever lowest resistance openings there are in the case. Fan theory says that 2 identical fans in push-pull mode through an unobstructed (low impedance) space don't create any more airflow than one. If there is some impedance, then 2 push-pull fans, because they create higher pressure, may help keep the airflow up. But if you're looking for maximum airflow, you really want to keep all the airflow in one direction -- either in or out -- because this nearly doubles the CFM for each additional identical fan (if there is no or low impedance, anyway). With or without impedance, 2 fans blowing in one direction almost always provides more airflow than 2 in push-pull mode.
If you want to control the inflow and the path of the cool air through the case, block all openings other than the intake vents, which should be as smooth-edged and unimpeded as possible. With very low airflow, the former doesn't matter much. YOu could try blowing some colored smoke (even -- horrors -- cigarette smoke! in a pinch) to see how the air flows through the case, if you have a transparent side case to view it all.
Another thing to note is that while any impedance (objects that obstruct) within say an inch of either side of the fan blades causes turbulence noise, the intake side is MUCH more affected. Just try it with your hands. Putting your palm with outstreached fingers anywhere less than 1 inch of the intake side of the fan dramatically increases turbulence noise. Doing the same on the other side only just barely affects it, basically it's not significant till you get within about 1/4" inch.
What does this mean?
Well the Alpha HS fan recommended position is to have it sucking in through the shroud, coupled right on top of the HS. This is a guaranteed way of maximizing turbulence noise. If you simply flip it over, you should notice a ~2-3 dB improvement in noise. Even better is to mount it on a 1" funnel on the back panel, blowing out, with a tube or duct fitted tightly around the fan & the HS. (The funnel is unnecessary if the back panel vent is completely unobstructed.) The duct I have used is flexible vinyl duct, which is basically a coil of stiff wire with a vinyl skin on it that accordians in length when pulled. Used to exhaust bathroom fans, available in 4-5-6-7-8" diameters, totally easy to work with, just a fews bucks for many many feet. Best when stretched to minimize wrinkles that can cause turbulence.
Sorry for the l-o-o-o-o-ng post. Got carried away...
There's nothing wrong with having only exhaust fans. The air that is blown out has to be replaced -- it naturally gets sucked in through whatever lowest resistance openings there are in the case. Fan theory says that 2 identical fans in push-pull mode through an unobstructed (low impedance) space don't create any more airflow than one. If there is some impedance, then 2 push-pull fans, because they create higher pressure, may help keep the airflow up. But if you're looking for maximum airflow, you really want to keep all the airflow in one direction -- either in or out -- because this nearly doubles the CFM for each additional identical fan (if there is no or low impedance, anyway). With or without impedance, 2 fans blowing in one direction almost always provides more airflow than 2 in push-pull mode.
If you want to control the inflow and the path of the cool air through the case, block all openings other than the intake vents, which should be as smooth-edged and unimpeded as possible. With very low airflow, the former doesn't matter much. YOu could try blowing some colored smoke (even -- horrors -- cigarette smoke! in a pinch) to see how the air flows through the case, if you have a transparent side case to view it all.
Another thing to note is that while any impedance (objects that obstruct) within say an inch of either side of the fan blades causes turbulence noise, the intake side is MUCH more affected. Just try it with your hands. Putting your palm with outstreached fingers anywhere less than 1 inch of the intake side of the fan dramatically increases turbulence noise. Doing the same on the other side only just barely affects it, basically it's not significant till you get within about 1/4" inch.
What does this mean?
Well the Alpha HS fan recommended position is to have it sucking in through the shroud, coupled right on top of the HS. This is a guaranteed way of maximizing turbulence noise. If you simply flip it over, you should notice a ~2-3 dB improvement in noise. Even better is to mount it on a 1" funnel on the back panel, blowing out, with a tube or duct fitted tightly around the fan & the HS. (The funnel is unnecessary if the back panel vent is completely unobstructed.) The duct I have used is flexible vinyl duct, which is basically a coil of stiff wire with a vinyl skin on it that accordians in length when pulled. Used to exhaust bathroom fans, available in 4-5-6-7-8" diameters, totally easy to work with, just a fews bucks for many many feet. Best when stretched to minimize wrinkles that can cause turbulence.
Sorry for the l-o-o-o-o-ng post. Got carried away...
Last edited by MikeC on Fri Sep 20, 2002 12:27 pm, edited 1 time in total.
>>>Well the Alpha HS fan recommended position is to have it sucking in through the shroud, coupled right on top of the HS. This is a guaranteed way of maximizing turbulence noise. If you simply flip it over, you should notice a ~2-3 dB improvement in noise. Even better is to mount it on a 1" funnel on the back panel, blowing out, with a tube or duct fitted tightly around the fan & the HS. (The funnel is unnecessary if the back panel vent is completely unobstructed.) The duct I have used is flexible vinyl duct, which is basically a coil of stiff wire with a vinyl skin on it that accordians in length when pulled. Used to exhaust bathroom fans, available in 4-5-6-7-8" diameters, totally easy to work with, just a fews bucks for many many feet. Best when stretched to minimize wrinkles that can cause turbulence.
----------------------------
ermmm.. please could u post a pic of that? I cant understand it very well...
So... if i use an alpha (for my p4) should i use the fan blowing or sucking? my idea is to put the fan sucking and put a duct on the rear panel... from the alpha to the rear i mean.
thx
----------------------------
ermmm.. please could u post a pic of that? I cant understand it very well...
So... if i use an alpha (for my p4) should i use the fan blowing or sucking? my idea is to put the fan sucking and put a duct on the rear panel... from the alpha to the rear i mean.
thx
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
Not sure which part you don't understand... Maybe the duct? Here it is:ermmm.. please could u post a pic of that? I cant understand it very well...
If it is the spacing between the fan & back panel you don't understand, imagine that you take a fan and rip out its center so that all you have left is the square frame. Simply put this -- like a spacer -- between the fan and the mounting panel. I'll see if I can find a photo...
ok ok .. its ok
I understand.
Just some troubles with english
omhh.. that idea.. the fan frame.. hehe its nice. The problem is.. I can NOT find that stupid vinyl duct in my city! I can only find some aluminium ducts , grrrr!
If i can find it i'll post some pics of my system
ur nice man , thx
gekkani
I understand.
Just some troubles with english
omhh.. that idea.. the fan frame.. hehe its nice. The problem is.. I can NOT find that stupid vinyl duct in my city! I can only find some aluminium ducts , grrrr!
If i can find it i'll post some pics of my system
ur nice man , thx
gekkani
There is also a problem with putting a fan on top of the HS due to vibration caused by it, as I pointed out in the original thread.
I agree with Mike about an intake fan. You don't need it if its sole purpose is getting air in. The negative pressure suffices provided you have holes in appropriate positions. If you can mod your case, it may be better to have holes on the bottom of your case, also.
Although you don't normally need an intake fan, the only exception may be the case where you need it for cooling HS that requires air to be blown onto it.
I'll upload a picture of my duct made of a Ziploc freezer bag later.
I agree with Mike about an intake fan. You don't need it if its sole purpose is getting air in. The negative pressure suffices provided you have holes in appropriate positions. If you can mod your case, it may be better to have holes on the bottom of your case, also.
Although you don't normally need an intake fan, the only exception may be the case where you need it for cooling HS that requires air to be blown onto it.
I'll upload a picture of my duct made of a Ziploc freezer bag later.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
You can also construct a duct using thick poster carton "paper", as Leo did for his ultimate undervolt project. (Scroll down to about the middle of the page)
Last edited by MikeC on Fri Sep 20, 2002 2:15 pm, edited 1 time in total.
-
- Site Admin
- Posts: 12285
- Joined: Sun Aug 11, 2002 3:26 pm
- Location: Vancouver, BC, Canada
- Contact:
One bad thing about all accordian-type stretchy ducts is that the surface is not smooth unless streched taut (very hard to do). The "corrugated" aspect of the surface does cause more air turbulence. Whether you can hear it, and whether it stops too much of the airflow are the key questions. If you use a smooth surfaced duct, it does make for less turbulent airflow. Of course, if you're running the fan at very low speed, this might not be an issue at all.
Are you sure they don't? They still do for me. Perhaps the linking got screwed when they reorganized the site. Mike?
http://www.silentpcreview.com/modules.p ... =44&page=1
You can also check out my original Tripod site too: http://rkinder1.tripod.com/silentpc/[/url]
http://www.silentpcreview.com/modules.p ... =44&page=1
You can also check out my original Tripod site too: http://rkinder1.tripod.com/silentpc/[/url]
just for your information, check out this link from anandtech...coolermaster apparently has worked with intel to develop a cpu ducting system (though its on the side). just an idea. http://www.anandtech.com/showdoc.html?i=1768