Just thought I'd share my cooling results with you guys.

[georget]

Hi all,

I'm relatively new to posting in these forums, but I've been actively reading them for some time. I recently overclocked my processor (Athlon T-bird 1.2 266 AYHJA-9 oc'd to 1.37 10x multiplier on a 137 fsb) and with the temps I was getting, I knew I needed some cooling help. Even at default speeds, I was having a wee bit of a heat problem with my processor.

For my HSF, I had gotten a (hangs his head) Coolermaster DP5-6I11A. On my case, I had 1 80mm fan in the front at the base as an intake, and in the rear were 2 80mm fans (1 in the power supply) that were performing exhaust duties. At default speeds (9.5x133), with my case closed up, I was getting idle temps of 46.4c (115.5f) and load temps of 49.1c (120.4f). After going through some of the cooling tips on this site, and reading many many posts, I decided the time had come to take some action.

The first thing I did was cut out a hole in two of my 5.25" drive bays to accomodate another 80mm fan. I then hung a small "L" bracket under the power supply (directly over the HSF) blowing down at a slight angle (angled toward the back of the case.) Finally, I cut the metal grille surrounding the lower intake fan, and increased the size and number of the holes in the front plastic cover. Regardless of which combinations of intake and exhaust that I tried, the temps barely budged (I lost 1c from idle and load no matter how I set the airflow of the fans.) I removed the side of the case and got a more noticable drop (40.9c idle, 47.4 load).

The highest I could overclock stable was (and still seems to be) 1.37, and my temps at this speed (10x173) were 41.4c idle and 48.9c load. This was, of course, with my case cover off. I didn't want to leave my case open, and knew that it was time for me to spend a little coin. Not having much to spend, I opted for the "Copperman" from 1CoolPC (Vantec CCK-6035D) and a new 80mm case fan (Sunon 50.6 cfm) along with a tube of ASII. My mainboard is a FIC AZ-11EA, so the Alpha was unfortunately, not an option for me. Well, my parts came, and I began the installation. I cleaned all the old thermal grease (generic) off the cpu, and then applied ASII as per the instructions on their website (substituting an old healthcare card in place of the credit card suggested ). I then replaced the fan I added to the drive bays with the new sunon fan (as an intake, seems to work better that way.) Well, I'm happy to say that my case is closed up and I'm now getting temps of 39.4c at idle and... 41.9c at Load!! This is at 1.37, after a 3-hour long flight in FS2002! I can finally keep my case closed AND not have to constantly worry about the heat!

One other thing to note, I added a second power supply (an old 235 watt AT) to the top of the case (outside it, actually ) and I drilled a hole in the top of the case to run the wires into it. The old switch is run to the bottom 5.25 drive bay that the extra intake fan is in. I got the tips for adding the extra power supply, and how to use drive bays for extra fans, right on this site!

So to those of you (like myself) that are new to this site, make sure you read all that you can find here, and take in all the info that you can. Most of these folks really know what they're talking about. And to those of you who regularly contribute to this site, and help those with questions, a very hearty thank-you.

Sincerely,
GeorgeT


System Specs:

AMD Athlon T-Bird 1.2 (AYHJA-9 Week 23, 2001) @ 1.37 (10x137)
FIC AZ-11EA Rev 2.1 (BIOS VDC-47)
384 MB PC-133 Micron RAM
VisionTek Xstasy GeForce 3 Ti200 @ 210/430
Enlight 7237 Case (300w ps)
Vantec CCK-6035D "Copperman" HSF

One other thing to note, the additional power supply only controls the "extra" fans, and all wires in the case have been bundled together (using black electrical tape) and routed neatly (as neatly as I could get it ) and all this has been done since adding the first fan to the 5.25 bay. Ok, I think I covered everything. Thanks for listening, and thanks for all the information I've learned here. Oh, and sorry for the long-windedness, I can be that way sometimes... ok, pretty often!

-GeorgeT

[edited for a typo in "plastic"... forgot the "L" ]

 

[KFB]

Welcome to the forums GeorgeT - you came to the right place. I stumbled upon this site myself over the summer when I bought a new system. Now I'm hooked. You definitely did your homework as that is a nice overclock (don't be ashamed about buying a CoolerMaster - I've made even worse purchase decisions)

 

[plague]

nice work, and welcome to the forums. I too started messing around here this past summer, and now I'm water cooled thanks to all these crazy people .

 

[bobt17]

nice work, and welcome to the forums. I too started messing around here this past summer, and now I'm water cooled thanks to all these crazy people .

DITTO lol WELCOME TO THE FORUMS!!!!

 

[georget]

Thanks for the welcome! Wow, into watercooling Plague? I haven't got the guts myself, I figure with the kind of luck I usually have, I'd turn my processor into a rather expensive aquarium decoration! As for the overclock, I was very fortunate to stumble upon some info about some of the athlons (I guess from the info I've gathered, the 266fsb ones) weren't locked at the factory. I checked my voltage and saw I was running at 1.87 without making any mods, so I took a chance and upped the multiplier (didn't want to take off the coolermaster beast until the new hsf arrived... wasn't too confident of the clips, and there are some capacitors VERY close to the socket.) As luck would have it, the CPU took the multiplier changes without a problem. I did manage to get it to 1.4 on a 10.5 multiplier, and it was rock solid in Sandra's burn-in and 3dmark2001, however prime95 would error out in less than 10 minutes, so I downed it to the 10x multiplier, and all was right with the world. Now that I'm running a little cooler, I'm going to give the ASII a couple of days to run and "set" (I guess from the website that it takes 48 hours or so to achieve maximum thermal conductivity) so if my temps manage to drop another 1 or 2c, I may try it again at 1.4 just to see what happens. Thanks again for the welcome!

-George

 

[Tiger]

Welcome - and may you enjoy this hobby as much we all do.
Another thing that will probably get you a few degrees is ducting.
The common one is to cut a hole in the side (even make an acrylic side for your case and then you can see whats going on in there as well its easier to cut than metal) and add a 80mm fan so that it is blowing into your heat sink fan. Then you can make a light cardboard duct so that the external air is ducted straight into the 60mm fan. It makes the world of difference.

 

[georget]

Thanks for the tip Tiger. I've been thinking about toying around with some ducting, but I was thinking of trying something a little different from the norm. The two ways I'm considering trying (and to be honest, I don't know how feasible either method will be or what kind of temp change to expect) are:

1) Instead of placing a single duct on the side blowing down on the hsf, I was thinking something along a "Y" duct. One side of the "Y" (the bottom side, as I would place it vertically instead of horizontally) would be the intake, and the other side of the "Y" (mounted on the top half) would work as an exhaust. I think this would improve airflow over the cpu/hsf even more than just an exhaust or intake duct would, but I haven't tried it yet to see.

2) The other method I'm considering trying is very similar in concept to what I stated above, but I'm thinking of running a duct the full length of the pc (from the 5.25 bays, over the cpu/hsf, and then all the way to the rear of the case. The fan installed in the front half of the duct (in the 5.25 bay) would work as the intake, and the fan at the rear of the case would be an exhaust.

I'm just trying to decide what material to make the ducts out of for testing, and, more importantly, how to mount them.

Let me know if you think either method would be viable, I'm definitely open to suggestions.

[edited for typos - boy I make 'em alot hehe]

 

[Tiger]

Here's my opinion and I will try to analyse what I am saying;
Type 1 with the y junction; You have a fan blowing in and a fan exhausting. The HSF is blowing down; the inlet is helping it but the exhaust will be opposing it. If the fans are identical the effect will be nil. The effect may in fact be counter productive in that it causes so much turbulence that the HSF flow my be radically affected.
Type 2. May be a bit better but it depends on the relative power of the exhaust fan. If the exhaust fan is stronger than HSF fan you will be forming a partial vaccum at the entrance of HSF and therefore starving it of air.
Some important aspects about air movement;
don't have fans opposing one another, have them supplementing each other. If the exhaust fan is close to the HSF it is best to shield the HSF from it. The exhaust fan will cause a partial vacuum behind the shield which is re-inforced by the HSF.
The reason why the side duct is so successful is because the fan is a) introducing ambient air and b) it is giving a large volume of it to the HSF.
I used one and it increased the HSF performance (RPM) by 10%.

 

[georget]

Hmmm, I hadn't given any consideration at all to a vacuum effect. I definitely see your point. My thoughts behind figure 1.1 was going on the very simple principle that heat rises. I figured with the lower duct as an intake, it would push in cooler, room temperature air, and that the hotter air would rise in the chamber and be sucked out by the top fan. Obviously, I don't know very much about cooling, so my ideas are kind of primative as the most primitive of priniciples are the foundation for them. I wonder this, and perhaps you can assist me with my ever-changing mind What if I made both fans intakes without the separator between them? Would that make more of a difference than a single fan ducted? I still have no idea how I'd go about making something like this, but I'm certainly willing to give it a shot! Thanks for shedding some light on this for me, I appreciate the help.

-George

 

[Tiger]

Let's look at the logic here;
Lets say you have moderate 80mm fan moving 50cfm and a HSF fan moving 40cfm. What's happening is the hsf is not trying so hard to accelerate the air through the hs and is therefore being aided by the 80mm. If you now add another 80mm to the equation and flood the hsf with an additional 50cfm then what will happen is not an additive effect. The most likely scenario is that the air stream from each 80mm will collide and cause a significant amount of turbulence. This turbulence could result in reduced air flow into the hsf.
Perhaps you are being confused with convection currents. Forget about convection. Once you add a fan you start using forced convection. Try to focus on air flows and methods of improving air flows. The one thing that I found while experimenting was that if there was a rough balance between the air being introduced and the air being removed then there was harmony.
Hope that helps.

 

[georget]

It helps a great deal. Thanks for all the info. I think the single fan ducted over the cpu should be enough from what I've learned here. One last little question (sorry to be a bit of a pain) is, if I did add a duct over the cpu with the duct fan sucking ambient air into the case and down onto the cpu/hsf, what's the best way to remove this air after it's been heated? Should I add an extra fan (perhaps as a "chimney" as an exhaust?)

Thanks again for all your help. It's been most appreciated!

[edited because I forgot one more question: With the duct, should it reach all the way down to the socket, or should the end of it be above/level with the top of the hsf? Thanks, hit "Post Reply" and forgot to add this question. ]

 

[Tiger]

Duct from the case fan to the entrance of the HSF . The heated air will come out of the fins and direction will depend on the orientation of the HS. If you have an exhaust fan in the normal place i.e. back top and an exhaust fan in the bottom of the PSU then they will take that heat out of the case immediately.

 

[Pepsi]

Just a thought but you never mentioned (or I flew right over it) what you are using for thermal compound? Believe it or not it does make a difference. Did you try lapping your HS to get better contact? Fans are great you'll get your temps down once you the air moving in and out the case then about that time the noise will get to you... heheh..
Welcome to the forums... BTW I started hanging out here about two months ago. Since then I built two decent water cooling systems something I'd never of considered trying to do 6 months ago. Good luck
Pepsi

 

[georget]

Thanks for the welcome Pepsi.

For thermal compound, I'm using ASII. With the vantec copper cooler (listed in sig) and the ASII, I got a pretty decent drop in temps over what I was running previously (coolermaster with generic thermal compound.) My temps now are running cooler with the case all buttoned up with the new hsf/ASII than they were running with the case wide open using the old hsf/ASII. But I see lots of people around here running reasonably lower temps than I on air, and the case temp is only running about 3c over room temperature. So it doesn't seem that airflow in and out of the case is the problem, but rather airflow around the cpu/hsf to get the heat out and away more quickly. I think I'm going to try Tiger's duct suggestion and see how much lower I can get my temps.

As for these forums, been reading them for a while, but just started posting. Seems like a really nice community, and I'm enjoying being part of it. Thanks again for the welcome.