Maximum Ducting.. Sparkle Project. Phase two.

Enclosures and acoustic damping to help quiet them.

Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Devonavar

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Maximum Ducting.. Sparkle Project. Phase two.

Post by Bluefront » Tue Dec 25, 2007 5:06 am

This is a test project....not nearly complete, with the HD out of the picture so far. I'm using this project to test out a few ideas, with the modified Sparkle SPI220LE as the center of the design. The case used here is just about the smallest setup you could use with this project....it's an older Dell only 7" wide, and 14.5" tall, and holds an MATX board with a std PS opening on the top. It's deep enough to use a standard optical drive in the top bay. Mounted like this, the Sparkle's fan has enough room for intake airflow (if you actually need the fan). For testing the Sparkle fan is left in place.....later it may be removed entirely or replaced.

The second new device in this design.....the louvered panel directly above the Ninja. This panel is the exact width of the case, and forces any upward airflow to turn toward the Sparkle and the PSU opening. It is custom built from a large wall vent panel. The edge pieces are bent up at a 90 degree angle, which makes the panel very stiff....with no tendency to bend. It just wedges in, resting on some ledges, and held tight when the outer case panel is replaced. Comes out with no tools.

The next new device....the "maximum duct". This is made from a plastic cookie container, cut to be a tight fit running from the bottom of the case, to half-way up the Ninja. It is square, slightly larger than a 120mm fan. It also just sits in the setup without any attachment...comes right out. When the outer case panel is installed, it is held tight by a foam strip which pushes against the outer case panel.

Next the 120mm intake fan....a medium speed 120x20 mm Yate Loon from JabTech. It has the bottom mount flanges cut off, so it pokes through the very bottom of the case. This was necessary for the installation, to clear a bottom mounted PCI card. In order for this to work, the case feet were replaced by taller feet. According to previous calculations/tests....there must be a minimum of 1.5" between the fan blades and the desk-top or the floor.....for maximum airflow, with the least noise. I intend to sit this case over a wood box on wheels, with a built-in filtration system, but it can work as you see it.....with sufficiently tall feet.

How does it work? There is a P4-3.4NW in there.....idles about 65W, and maxes about 155W. Temperatures.....CPU idles about 31C @850rpms (23C ambient). Maxes about 47C @1350rpms. The Nvidia 6200 is right in the airflow path, and runs super-cool. The rest of the MB components are in the airflow path.....some air from the fan escapes out the back of the three-sided duct, directly at the board......and that's without picking up any heat from the Ninja. The heat off the Ninja has a separate path upward through it's fins.

The Sparkle sitting on top of this airflow path....cool as hell with it's perforated DIY cover. It measures around 35C running, no matter where I can get a reading. So far it's working great....I can post more/better pictures if there's any interest in this project. Working great so far...

Image

Image

Image
Last edited by Bluefront on Sun Jan 13, 2008 10:33 am, edited 4 times in total.

sanse
Posts: 399
Joined: Thu Dec 22, 2005 9:48 am
Location: Netherlands

Re: Maximum Ducting....The Sparkle Project.

Post by sanse » Wed Dec 26, 2007 3:00 am

Bluefront wrote:This is a test project....not nearly complete, with the HD out of the picture so far. I'm using this project to test out a few ideas, with the modified Sparkle SPI220LE as the center of the design.
like these vertical ventilated designs. it's so natural. why don't we see more commercial applications of this? btx comes near, but that's usually implemented horizontally.

the memory chips are not cooled in this setup. isn't that worrysome?

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Wed Dec 26, 2007 3:30 am

It's impossible to photograph....but the two memory chips closest to the Ninja are in direct airflow, since there's an opening on the rear of the plastic duct directly facing the two memory sticks. The outer two sticks could be included in the airflow with a few cuts on the duct with a small deflector....in the planning stage. Today I'm building the HD cooling part of this project, which will also include a suspension setup. No more fans...just an extension to the duct.

With this setup, I'm determined to make the installation of everything as easy as possible...it's getting there.

One other thing...there's enough room to install a second fan on the bottom of the case, blowing upward. This could be installed as a single airflow path, or a separated path dedicated to hard drive cooling, with both fans sharing a common filter (or not). Like I said.....testing. :lol:

sanse
Posts: 399
Joined: Thu Dec 22, 2005 9:48 am
Location: Netherlands

Post by sanse » Wed Dec 26, 2007 5:02 am

Bluefront wrote:It's impossible to photograph....but the two memory chips closest to the Ninja are in direct airflow, since there's an opening on the rear of the plastic duct directly facing the two memory sticks.
see it now. it's indeed difficult to see.

i would go for 2 120mm or 2 140mm fans in the bottom.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Wed Dec 26, 2007 9:27 am

FWIW....a 140mm fan is too wide to fit in this setup. The round cut on the bottom of the case is within 1/2" of the outer edge, and you have to leave enough metal for the case to remain square. :lol: And if you only use one hard drive, you don't need two fans....here's the HD cooling solution.

Before I started this HD cooling part of the project, it was just sitting on edge in the front of the case, on a pile of foam. Idled about 41C with the case open. No airflow was on it. The temp went over 44C with the case closed. Now it's running cool after these mods....

Image

Image

Pretty simple really. The HD is sitting on a pile of soft foam, which is inserted into a block of stiffer foam (to hold it's shape). There is an aluminum channel screwed to the drive, using a HD cooler bracket with the fan removed. This leaves about a 1/2" channel open on the top. The assy butts up against the duct, with some soft foam between the two. I cut a 65mm hole in the plastic duct, which diverts some of the duct airflow toward the rear of the drive.....

This dropped the Maxtor idle temp from 41C down to 28C. When the CPU is maxed out, the drive temp drops to about 26C because of the increased fan speed. No more HD temp problems. I think this same technique could be used with two HDs, with the drives at 90 degrees to the duct, and the airflow blowing through them.

Now of course this HD cooling, lessens the airflow through the Ninja, but not much it seems....the fan-speed went up about 40rpms, with the other temperatures remaining the same. (I'm using an NTM3 automatic fan-speed controller) Ambient today is about 22.5C.

As to the modified Sparkle on the top.....I haven't been able to ever hear it ramp up off it's fan idle speed. With the case open, I can see the fan is running....that's all. I'm fairly certain after this latest testing, that I can safely remove the Sparkle fan. That might be the next test. :D

Oh... that last photo gives a better view of the cut in the duct, that blows some air toward the ram chips.

sanse
Posts: 399
Joined: Thu Dec 22, 2005 9:48 am
Location: Netherlands

Post by sanse » Wed Dec 26, 2007 10:07 am

Bluefront wrote:This dropped the Maxtor idle temp from 41C down to 28C. When the CPU is maxed out, the drive temp drops to about 26C because of the increased fan speed. No more HD temp problems. I think this same technique could be used with two HDs, with the drives at 90 degrees to the duct, and the airflow blowing through them.
are you sure the temp of the hd went down so much or did only the temp of the sensor on the back-side of the hd go down that much because of the wind blowing along it? most of the hd is not in the airstream, only the backside with the electronics.

during my experiments with the solo with the 6 fans i discovered, that a sensor in the airflow of a fan can give temps waaay off.

is the hd attached to the duct or only sitting b itself in the foam-structure?

jstyles
Posts: 15
Joined: Thu Mar 10, 2005 11:11 am

Post by jstyles » Wed Dec 26, 2007 11:57 am

I enjoy reading about bluefront's designs and this one reminds me of an idea I have been wanting to test, but no time to implement.

I was thinking of a similar design, but where the fan on the power supply itself was used to pull air through a Mini Ninja and past a hard drive suspended above the PCI slots.

The power supply would need to have a large top mounted fan that was mounted upside down over the Ninja.

I was going to make a box that was the size of the MB and width of the PSU/HS. Cut out a hole in the top that was inline with the Mini Ninja and size of the PSU fan then mount the PSU on top with the fan down. It would require another hole in the bottom for air to flow in. I am not sure how I would mount the HD inline with that hole and over the PCI slots. I was planning on using a MB with IGP and no PCI cards like one based on 7050PV, 690G or G35 that have everything integrated on the MB and therefore would not require addition PCI/Graphics cards.

The airflow would be

<- PSU
/\
Fan
/\
Mini Ninja
/\
HD
/\

I assumed this would give a rather small size case 10"x10"x5" with the PSU mounted on top and everything else except the DVD drive with some airflow.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Wed Dec 26, 2007 12:42 pm

jstyles....for the most part, computer builders are stuck with a few designs, which vary only slightly. But it stands to reason, since we are stuck using mother-boards of certain sizes, not to mention all the other pieces.

I've always held to the view that there is a perfect design out there, waiting to be discovered. But just thinking about such a thing won't do it....you've got to make the time and the effort. I've tried using the PSU as a primary airflow path for the exhaust. It can be done with a bunch of difficult mods....not really the way to go. This Sparkle because of it's size, just sits in the airflow, and hardly blocks any of it.....very different setup from anything I've ever tried.

Concerning the HD temp changes .....they are real. I've measured all the different sides of the HD, not just the Smart temps. The bottom of this Maxtor is where the sensor is located, but the bottom is the hottest part, and much of the aluminum body is exposed right there. Blowing ambient temp air at the bottom is all it takes to drastically reduce the temp of the entire drive.....and not a whole lot of airflow either. In this setup none of the drive or the aluminum channel actually touches the duct. There are strips of foam between the two.

Here's what the bottom of the case looks like. You can see the four temporary feet I built....keeps the fan off the bench while I'm working on it. These feet need to be two inches tall to be a permanent solution.

Image

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Thu Dec 27, 2007 2:59 am

I just read the new SPCR article on the new Arctic Cooling VGA cooler....and it looks to be ideal for this "maximum" duct setup. Imagine this SPCR setup with the fan on the bottom of the case rather than directly on the heatsink.....picture from the SPCR article.

Image

The fins on this heatsink appear to be plenty wide enough to impede the airflow only slightly. Combine this card/cooler setup with a newer processor that draws less current than my 95W P4, and a Ninja of course....there you go. That is provided the Sparkle PSU can handle the current. :lol:

This cooler looks to be slightly wider than the plastic duct I'm using...no problem. It's easy enough to cut new slots so the ends of the AC heatsink would fit. Somebody else is going to have to try this though.....

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sat Dec 29, 2007 10:02 am

Here's the same HD airflow technique, using two drives 1" apart. They are held together by copper sheets, .025" thick. These also cover the bottom, so three edges are heat-sinked by the copper. The copper side facing the duct has a vertical slot cut, matching the previous hole cut in the plastic duct.

The entire assy is sitting on a 5/8" pile of soft foam sheets. There is less airflow between the drives with this new setup, than the previous one-drive setup. I could cut the slot bigger, but I'm going to test this way for a while.

There is plenty of room here...nothing is cramped, and the drives are easily changed. Now I'm on the look-out for a new case to try out this entire setup. This old Dell case looks too plain. :lol:

Image

Image

Image

Jipa
Posts: 167
Joined: Sat Dec 22, 2007 4:06 am
Location: Tampere, Finland
Contact:

Post by Jipa » Mon Dec 31, 2007 2:40 am

Interesting design and, if nothing else, it atleast looks effective :) I'm a bit hysterical on the HDD-temps, so I might add another fan to the bottom of the case (at minimum RPM) so it would push air all around the HDDs.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Mon Dec 31, 2007 5:56 am

You'd be quite surprised how effective the copper enclosure is at cooling. I've used these DIY copper HD coolers before. You can adjust the amount of airflow into the enclosure by enlarging the intake hole, or increasing the main fan airflow. As you can see in these next photos, there is enough room on the bottom for a second fan......but you'd need a different HD suspension system. Or you could mount a front intake fan on the case blowing at the drives. I don't like intakes up front....too noisy for me. These older Maxtor drives always ran hot....newer drives run cooler.

These next photos show a few more changes. You can see the HD suspension. The bottom of the HD assy sits on the black center section. This is fairly soft, and compresses about 1/4".

I made a bottom wood stand and now there is enough clearance to mount a 120x25mm fan directly on the bottom of the case, without cutting the flanges. The fan blades are now 2 1/4" from the floor. I'm now using a 1600rpm S-Flex fan....nice. The fan spins about 650rpms at an idle, and 1150 maxed out. (CPU at 31-44C). At these low rpms, the HDs run warmer than the previous YL fan setup. Now at about 38C...ambient today of 23C.

And I covered the plastic duct with Presto felt sheets, inside and out. This definitely reduces the fan/duct noise....well worth the trouble. It is important to seal the edges of the duct (you can see the foam strips I use). You want all the airflow to be effective.

And I added a angled deflector panel above the HD assy.....which helps the airflow heading toward the exhaust.

The next step....a different lower stand, slightly taller, with a built-in air filter setup, which slides out from the front for maintenance.

Image

Image

Image

Image

Image

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Tue Jan 01, 2008 7:50 am

Here's the next mod to the HD cooling. I cut four 3/4" holes in the rear of the copper plate, centered between the drives. On the outside I mounted a set of ventilated heatsinks......cut from a hard drive cooler bracket.

Then on the top, I covered about 1/3 of the front of the opening between the drives. This forces more of the airflow toward the rear of the assy, and out through the blue heatsink.

Works pretty well. With no other changes, the two drives dropped about 2C. They now idle about 36C. Ambient today is about 23C....the new Scythe fan is idling about 740rpms.

Image

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Thu Jan 03, 2008 5:44 am

Maximum quieting requires maximum modding....heh. I spent some time tweaking the airflow relationship between the main duct, and the side HD cooling duct. Obviously the more airflow diverted to the HD cooler, the less airflow is available for the CPU and other areas.

And I also want to keep the Scythe fan running as slowly as possible. The CPU temp at an idle is un-important....it will be low enough. But the drive temp at an idle is important to me. At an idle the fan speed is the lowest, using the NoiseMagic NTM3 controller....no getting around that. At an idle speed of about 5V, the Scythe runs about 640rpms, so that's what I'm dealing with. Modifying the amount of airflow into the HD cooler is the easiest solution here....and accomplished by an airflow deflector which is located in the main duct, and forces some airflow sideways into the HD cooler. This next set of photos shows this final deflector (I tried a bunch of different sizes). The two Maxtor drives now idle about 31-33C, while the CPU idles about 31-32C. (ambient of 23C). At max CPU usage the CPU now runs about 47C@1240rpms, while the HD temps drop slightly. I'm well satisfied with these numbers, which have been verified with several temp measuring devices. :D

Image

Image

Image

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Wed Jan 09, 2008 4:49 am

This setup is progressing nicely....I just completed a slightly taller lower stand which has a removable filter. Link to thread. In addition to filtering the single intake, it also quiets the intake....a second benefit to filtering done right. :D

Image

The next on the list.....trying out a few different fans for the setup. Maybe a Scythe Slip-Stream. And I'm still looking for a different case into which I would transfer this entire design. The Dell case I'm using here is ok.....but could stand some improvement. Actually when you don't have to worry about the rear fan opening, the front bezel design, or the HD mounting....picking out a new case should be easy.

jhhoffma
Posts: 2131
Joined: Mon Apr 25, 2005 10:00 am
Location: Grand Rapids, MI

Post by jhhoffma » Wed Jan 09, 2008 6:23 am

Blue,

One of these days, I'm driving out to MO just to look at your garage... :lol:

PretzelB
Posts: 513
Joined: Tue Feb 11, 2003 6:53 am
Location: Frisco, TX

Post by PretzelB » Wed Jan 09, 2008 9:13 am

jhhoffma wrote:Blue,

One of these days, I'm driving out to MO just to look at your garage... :lol:
I think that may be the only way to figure out what's going on!! I don't know why but I struggle with the pictures trying to figure out how it all works. For example, I can't tell where the air comes from for the drives.

The stand/filter is pretty nice. I shudder to think how much time that took - it looks so polished. Did you find a new case or are you hoping the next one will be the same dimensions?

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Wed Jan 09, 2008 11:25 am

Heh....I try to make the pictures tell the whole story. Before I covered the duct with black felt, you can see the airflow path into the drive cage. It's simply a hole cut in the side of the duct.....I added a little air-scoop in the later pictures. The copper drive cage has an oblong hole in the one end that matches the hole in the main duct. That duct is under positive pressure....any hole cut into it will allow airflow to escape. Of course there's a limit, determined by the CFM of the fan. Like I named this thread....this is a maximum duct. No other airflow is necessary. The Sparkle on the top could run fanless.

You'd be disappointed if you saw my work area. Nothing much. That last stand/filter housing could have been constructed using a cordless drill, and a jig-saw. I worked on/off for about six hours total. The first stand in the photos was finished in two hours.....the stain drying time took the longest. If I need to make another for a different-sized case....no big deal. Making the first ones always takes the most time.

On this Sparkle project....what you cannot see is how easy everything comes apart. It's even easy to change the MB. Changing out the main fan...about a five minute job.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sat Jan 12, 2008 2:36 pm

The first stage of this project is finished.....and it was a success. This computer will run on the single fan you can see. The fan in the Sparkle only turns on if I run CPUBurn for quite a while, or if I run with the case open. Otherwise it could handle fanless no problem. I can measure temps on it of a maximum of 37C.....

The ambient lately has been about 23C. The CPU usually idles around 32C, with the Scythe S-Flex at about 725rpms.....the filter installed. Both WD drives in there right now run about 34/35C, and even less when the CPU is being stressed. Stressed out the CPU runs about 47C @1275RPMs. At that point I can hear the airflow from about 3 feet away. But that's to be expected with a relatively hot CPU like this P4-3.4.

This project was built mostly from older parts, the Sparkle being new of course. I am looking to find a new case, but haven't found one I like. This short tower size works out pretty well. The only thing that stands out so far...this Silverstone, which doesn't get very good reviews....somewhat flimsy. But it's exactly the right size, and could be easily modified for this setup.

While I'm chasing after a new case, I did a small bling job to the Dell. Added a Scythe Kama-panel with a laptop drive, and an Antec temp read-out panel......not to mention the new wood top, with the "dental" moldings. I like it. It can be duplicated for a different sized case without much trouble or cost. If I find a new case, I'll post about it in the Gallery section. A realistic single-fan computer, made possible by a maximum duct.......

Image

Image

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun Jan 13, 2008 10:25 am

On to phase two.....tweaking the setup for better temps, at lower rpm from the main fan. And for you non-believers in the value of ducts, I ran this setup without the duct.....everything else the same. What happened.....at max usage, the CPU temp shot up past 50C and was rising, and the S-Flex was past 1400rpms and rising. And the GPU temp was up over 7C. No good.

Next.....I hung a Scythe 80x15mm fan under the Ninja, blowing upward. It is controlled by the Scythe panel manual fan control. I adjusted the speed till I can not hear anything.....that's about 7V. Then I repeated the no-duct test. Better this time, The CPU maxed at 49C with the S-Flex at 1400rpms. Still not as good as the duct alone.

Finally for now, I replaced the duct, leaving the little 80mm fan attached and running at 7V. Well this way does give better temps than any of the setups thus far. The CPU maxes at 46C with the S-Flex at 1210rpms. A slightly better set of numbers than running without the little 80mm fan. As to noise.....hard to say. Probably slightly quieter because I cannot hear the 80mm and the rpms of the S-flex dropped somewhat at max usage, and at an idle....about 650rpms right now. About 50rpms lower than without the small fan.

So I think this second fan on the Ninja may be worth the trouble, not by much however. The duct however, is critical to the good temps, and the slow fan speeds. Next up......maybe a different fan to replace the S-Flex. Here's what it looks like without the duct, with the little fan attached with zip ties and rubber spacers. (I rigged up a temporary scoop for the drives).

Image

PretzelB
Posts: 513
Joined: Tue Feb 11, 2003 6:53 am
Location: Frisco, TX

Post by PretzelB » Sun Jan 13, 2008 10:56 am

I re-read this again and I think I finally get it. So you replaced the old duct that tool the air and sent towards the top and replaced it with the duct mounted to the HD assembly to just steal some airflow from the bottom fan. I would guess not much air is going towards the drives but it sounds like it works really well. I'm also surprised the bottom filter assembly doesn't knock the temps up a bit as opposed to no filter.

What are you using this machine for anyway?

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun Jan 13, 2008 11:19 am

Well not exactly. The HD scoop in this last photo was there only for the no-duct test. The main duct now has a built-in scoop. It does steal some airflow, but not much. You can feel it when you put your finger over the drives when the setup is running. It's enough airflow to keep the drives under 35C, normally.

As I said in the earlier testing.....if I pull the filter out when the thing is idling, the CPU drops 1C, and the main fan-speed drops about 50rpms. So there is some restriction on the airflow..... The fan controller is sensitive enough to lower the speed almost immediately when the restriction is removed.

I've been doing minor tweaking as I go along, so some of the numbers might be off slightly. The ambient has been 23/24C lately in the test room, and that does make a difference. This thing is only a test project for now. I don't need another computer.....got too many now. But when I saw the Sparkle SPCR review, it was obvious to me that a new airflow setup was possible. That Sparkle has been a complete success. I did hear the fan the other day when I ran with the case open for a long time. The Sparkle is the only thing that gets hotter when the side is open. :)

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun Jan 13, 2008 11:46 am

You know....it's beginning to look like the NoiseMagic controller might be next bottle-neck to going much lower on the fan speed. It's running on it's minimum voltage at 650rpms. I tried reducing it's operating voltage slightly. I can reduce the voltage to it by 1.2V......then it has a lower minimum when running. If I go lower however, it's control circuit is upset and it maxes the voltage to the fan. The only way around that is to install a different fan that will run slower at 5V or so.......and still have enough airflow at maximum loading.

The Noctua 9-blade 120mm is a good possibility....but it costs a bunch, compared to this S-Flex, 2.5x as much to be exact. :(

PretzelB
Posts: 513
Joined: Tue Feb 11, 2003 6:53 am
Location: Frisco, TX

Post by PretzelB » Sun Jan 13, 2008 3:32 pm

Bluefront wrote:Well not exactly. The HD scoop in this last photo was there only for the no-duct test. The main duct now has a built-in scoop. It does steal some airflow, but not much. You can feel it when you put your finger over the drives when the setup is running. It's enough airflow to keep the drives under 35C, normally.
I think I get it now. The picture of the duct on 1/3 is probably what makes it easier for me to imagine. I assume the only escape route for output air is out the back of the psu right? I see the top you made but I think that's just for show and there is no hole in the top.

Maybe it's just me but one thing that might make it easier to visualize is arrows for the air flow. It's hard to keep track of where all the holes are and where you are sending the air. Probably just me.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Sun Jan 13, 2008 4:00 pm

I think you've got it. There are no holes on the top. The only exhaust is through the perforated Sparkle, but mostly around the Sparkle through the new grill I made. I do think an exhaust vent hole directly above the Sparkle on the top, would provide slightly better PSU temps, and maybe better overall temps......but probably not enough to make a big difference.

A 3D drawing program would show the whole setup better I guess.....but that's too much for me.

PretzelB
Posts: 513
Joined: Tue Feb 11, 2003 6:53 am
Location: Frisco, TX

Post by PretzelB » Sun Jan 13, 2008 5:42 pm

Bluefront wrote:A 3D drawing program would show the whole setup better I guess.....but that's too much for me.
Beggars can't be picky but I won't say you need a 3D drawing program. All I was thinking was taking what you have and then adding some arrow icons to the images. Of course that means you have some kind of photo editing software to cut and paste the arrows to the picture (and the time to take the effort). Again, not trying to tell you what to do, just saying what would make it easier for me personally to see it easier. Feel free to ignore as I won't mind a bit. :wink:

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Mon Jan 14, 2008 1:06 pm

More about airflow with this setup......With the addition of the second "helper" fan under the Ninja, I can now reduce slightly the airflow from the main fan. Not by much however, because the helper fan only cools the CPU.....nothing else. And the HD temps remain a concern, since they are dependent on the airflow from the main fan. It's proven tricky getting everything right. I changed the Scythe 80x15 for a San Ace 80x15, which is quieter with somewhat less airflow.

Then I attached this helper fan to the main fan controller, so now both fans increase/decrease with the CPU temperature.....I think it's quieter this way. And because of the added airflow from the helper fan, I think I can now install the new 1300rpm Noctua 9-blade fan, which might have somewhat less maximum airflow than the 1600 S-Flex now in there.

But there are a few other ideas I have which cannot be used on this case....so. I just ordered a new larger case for this maximum duct project, along with a few other pieces, including the Noctua 9-blade. Before I start the next phase, I'll try out the Noctua in this setup, which is tweaked to the max right now.

Here's the new case I'm waiting for......

Image

It's an I-Star S8 from NewEgg. It looks to be a good candidate for some new stuff I'll be trying. It's slightly taller, wider, and longer, but I intend to try out a slightly different lower filter housing. This entire Maximum duct concept has many possibilities....I've only touched on a few.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Tue Jan 15, 2008 9:19 am

While I'm waiting for the new pieces to arrive, I thought I'd try some more things with this duct setup. So I replaced the little air scoop for the HD chamber, with a 60X10mm Yate Loon. That's the biggest fan that I can fit....a 70mm fan hits the edge of the video card. SpeedFan can control this fan....so I have it set on 30%. The drives run about 33C with this very low airflow.

Amazingly to me.....the max CPU temps went down with the addition of this fan. This proves that little scoop does have a measurable affect on the main airflow. I don't think this changed the sounds at an idle, but it did help when the CPU is stressed. It now maxes at 44C @1110rpms. It idles about 31C @ 650rpms today........23C ambient.

Conclusions here....mixed bag. If you can find a quiet small fan, and control it to run slow enough, you can have better/easier control over the HD temperatures. Of course this adds complication, cost, and more cables. Take your pick......the duct by itself can keep the drives cool enough.

Image

sanse
Posts: 399
Joined: Thu Dec 22, 2005 9:48 am
Location: Netherlands

Post by sanse » Tue Jan 15, 2008 10:29 am

Bluefront wrote:While I'm waiting for the new pieces to arrive, I thought I'd try some more things with this duct setup. So I replaced the little air scoop for the HD chamber, with a 60X10mm Yate Loon. That's the biggest fan that I can fit....a 70mm fan hits the edge of the video card. SpeedFan can control this fan....so I have it set on 30%. The drives run about 33C with this very low airflow.
another 3 fans and you are at six, like me in the solo. ;-)

the opinion of mike, that any slow spinning fan is better than convection is soo true.

Bluefront
*Lifetime Patron*
Posts: 5316
Joined: Sat Jan 18, 2003 2:19 pm
Location: St Louis (county) Missouri USA

Post by Bluefront » Tue Jan 15, 2008 12:49 pm

Nah...This is a test setup. The final computer will be different. The hard drives weren't running passive. The little air scoop diverted airflow from the main duct between the drives, and they run cool enough like that.

This last HD fan just leaves most of the airflow in the main duct, mainly untouched, which makes the CPU run slightly cooler at max. The little helper fan on the Ninja, does the same thing. Neither are 100% necessary.

I've been considering trying a 140mm Yate Loon as the only fan in this new case arriving soon, since it's slightly wider.. The other possibility.....two 120mm fans on the bottom, with a modified HD mount to take advantage.

Post Reply