Something wrong wit case temps.

[phatboi]

ok, there's definitely something wrong with my position of fan in my case. i have a SkyHawk Aluminum Case and i have a Vantec Stealth 80mm intake fan up front that i cut a hole for, an Enermax adjustable fan (80mm) on a bracket in the middle of my case, an Enermax 60mm exhaust fan. i also have an the Dual Fan Enermax Power Supply that has the fan right over my processor to suck out heat from my case. i have an adj. 80mm Enermax on my AX-7 too. with the side of the case on, i get reasonable temps. around 45C for the Processor and 41-42 for the Case. but just with one side off, i get it down to 40C for the Processor and about 39 for the Case. it seems pretty high to have a case temp in the 40's, especially for an aluminum case. but am i getting negative pressure? i can't put in an 80mm exhaust fan, that's why i have the 60mm Vantec. should i put in 2 80mm side fans on the side of my case that will blow in cool air? or have 2 80mm fans that will suck out hot air? i want to run my system with the side of the case on. which i usually do, but with a 5 degree difference, i need to know how i can keep that w/ the side of the case on. i don't think i'm doing to bad if i can get my Xp 2000+ to stay at 40C w/ those case fans. any suggestions? thanks.

[Cricket]

Hi phatboi,

Just as an experiment, try reversing the fans you already have installed. Have the 60mm on the rear blow into the case and the 80 in the front blow out. Put your case side back on and then get a temp reading after your system has been on for awhile.

This should give you negative pressure as long as you have a lot of vent holes around the case. You just want to see what your temps would be like with negative case pressure.

If your temps do adjust down, then you're going to have to do some modding (do you have a Dremel Tool?)...or leave the fans the way they are.

Cricket

[phatboi]

thanks, cricket, i'll try that. yes, i do have a dremel tool. i just bought one the other day so i can cut a hole for my intake fan. i like it, pretty fast (35,000 rpm).

[mike breck]

Obtaining negative pressure will also depend on how much air your intake and exhaust fans are moving.

For example, my gaming system is a T/Bird 1200 266.

I have one 45.2CFM bottom/front, one 45.2CFM middle back, and one 45.2CFM top/back.

So I have approx twice as much air being extracted from the case. This gives me a mobo temp 21c. In fact, my mobo temp has alway been 1-2c above ambient room temp. Pity I'm not an overclocker, because most overclockers would be very pleased with that.

I question your case fan in the middle of the case on the bracket. When I tried such an arrangement, it disrupted the "S" flow (visualise the optimum air flow in the shape of a letter "S", travelling from bottom/front to top/back) flow thro the case and on my HSF.


I think you should remove this fan and concentrate your efforts on the intake/exhaust side fans. If you do not have room for an extra case fan at the back, then consider getting a PCI slot fan. These are cheap and very effective with an airflow of 42CFM.

Also try removing the filter from the fan bottom/front and see how that affects your temp. I found having filter material increased the case temp considerably; so I removed it.

I have not found the dust intake has increased dramatically without it. Even if it does, you just give the case a good blow with compressed air every couple of months. It will still be cleaner than most PCs that are cleaned on a one-two basis.

Finally see this thread regarding case fan placement. Check out the experiments conducted by Deadeye and the link to his diagram.

http://forum.pcmech.com/showthread.p...threadid=25733

HTH

[Trebuchet]

Two things here:

1) Negative pressure draws in air through every nook and cranny, including through optical drives. Dust = Not a good idea. A slight overpressure is considered better for cooling by most overclockers. It is also essential to filter the intake air.

I think you might wish to look at getting more airflow through your case, possibly by cutting blowholes. Case cooling needs to be planned, it's not a matter of just randomly slapping in fans.

2) I am very suspicious of your temp readings. It is inconceivable that your processor is only 1C higher than your mobo and/or case with the setup you describe, that is simply not possible without refrigeration. What are you using to get your temp readings? Some software for this are notorious for their inaccuracy, this includes ASUSprobe and SiSoft Sandra. I personally like Motherboard Monitor. (http://mbm.livewiredev.com/)

I would first check temps in the BIOS to see if they at all agree with your posted readings, then calibrate your temperature sensors. The procedure can be found here: http://www.arcticsilver.com/diode_calibration.htm

[morriswindgate]

The CPU temp is not bad, but it sounds like you have bought into all the hype and fear of AMD meltdown.
First you don't need any more fans, as you probably have one to many already.
And second, It is quite possible that you are getting negative pressure in the case. If this is true, you actually get less of a volume of air to cool everything and at the same time you can get fan surging which will cause them to actually generate heat.
The first thing that I would do is unplug the 60mm exhaust fan and see if the temp come down or stay the same. Then if they do you might want to setup the 80mm front fan to run on 7 volts to cut down on the noise.

[Alfie]

Those temps seem high for your setup.
Have you tried reversing the fan on the heatsink?
You may want to try this and see how the temps read with the fan reversed.
Also make sure the PS's fans are exhausting(you never know!)
My current temps are cpu 24c and case temp 22c and the setup has 3 exhaust(including PS) and 2 intakes.
Your heat sources are the cpu(#1) hard drives,how are they stacked and do you have any cooling for them?
Video card,is it overclocked? Is the fan working on it?

[Trebuchet]

>My current temps are cpu 24c and case temp 22c and the setup has 3 exhaust(including PS) and 2 intakes.

Alfie, if you are not using a peltier then there is no way your CPU temp is only 2C above your case temp. I think you must have your sensors set incorrectly, or perhaps labeled wrong. I would buy that your mobo is only 2C above case or maybe even ambient temps, but the CPU is going to be at least 8C higher even at idle.

As an example, my overclocked PIII system currently is showing (at idle) :
CPU: 34C (It will climb to 45C under full load)
Mobo:27C
Ambient: 25C (Sensor at air intake)

I have 130+ cfm of air flow through my case, including 2 120mm blowholes directed at my CPU, mobo, and cards. I doubt your cooling setup moves as much air.

[Alfie]

I have a direct read from the cpu and in addition have software programs reading the same,the temps are correct.(bios also show the same temps)
The additional cooling to the cpu is by using v-cool,it is able to drop the temp by 8-10c.
With that in mind,the actual read off the cpu is 38.7c with v-cool disabled and cpu under heavy load,32-36.4c. using it normally.
It took me a while,but with experimenting with fan placement and knowing where the hot spots were,I was able to get the temps way down.
By the way,I've been at it for awhile

[mike breck]

My view is dust is going to get into a case whether you have filters or not. The important thing is to the case and fans a good clean out on a regular basis. Also keeping the working environment clean, hovered and dusted, helps as well.

I've monitored dust intake on two systems, one with negative pressure and one without, and there was no perceptable difference in dust intake.

See this thread for a discussion about negative air pressure.

http://forum.pcmech.com/showthread.p...threadid=25365

However, I agree experimentaion is the key. What works well on one system is not necessarily the best arrangement for another system.

Anyone who's interested enough will try various strategies and see what works best for them.

[HAL9000]

Originally posted by Trebuchet Two things here: 1) Negative pressure draws in air through every nook and cranny, including through optical drives. Dust = Not a good idea. A slight overpressure is considered better for cooling by most overclockers. It is also essential to filter the intake air. I think you might wish to look at getting more airflow through your case, possibly by cutting blowholes. Case cooling needs to be planned, it's not a matter of just randomly slapping in fans.

Positive air pressure means that air in the case is not escaping fast enough. If it cannot escape, it has more time to take on more heat. Taking on more heat leads to expansion of the air in the case which increases pressure. As you can see, it's self defeating. A slight negative air pressure promotes air flow, positive hinders it.

A positive air pressure will do the opposite that you mention the negative air pressure will do. Dust will come into the case and find its way out one way or another, whether it be through floppy drives or optical drives.

Also, take a look at this experiment. It clearly shows one intake, 4 exhaust, and one fan promoting a upward flow, all which would easily bring on a negative pressure, being the coolest scenario.

[Trebuchet]

HAL9000, while I quite agree that is a well done test as far as it goes, it is incomplete. None of the configurations shown produce positive pressure or even balanced pressure. They are ALL negative pressure, so it only demonstrates variations on a single theme. Also, the rear intakes merely recycle already heated air from the other rear exhausts. I have tried every single one of those exact configurations in my system, and others, including the CPU duct (both in and out) and slot cooler before I settled on my current arrangement.

In any case, I suggested a slight overpressure, not a 30+ cfm difference which would be created by such an imbalance of intake vs. exhaust fans as in the referred tests you linked. A severe pressure imbalance either way would not be good. I use variable speed fans or my baybus to adjust fan speeds and thus "tune" my pressure.

Slight positive pressure with filtered intake offers the best general method to keep dust out of delicate optical drives by being sucked in through every tiny crack in the case and drives IMHO.

>The additional cooling to the cpu is by using v-cool,it is able to drop the temp by 8-10c. With that in mind,the actual read off the cpu is 38.7c with v-cool disabled and cpu under heavy load,32-36.4c using it normally. <

Alfie, don't you perhaps think artificially dropping your CPU temp by lobotomizing it with HLT software might have been a significant enough detail to include in your original post? I can achieve similar temps by turning my system "Off". ;-)

[Alfie]

Trebuchet,If you knew what you were talking about,perhaps I'd listen.

[Trebuchet]

Ah, I see. You have lots of posts, I have only a few, therefore you must know more than I do about every subject. I wondered exactly how that worked in this particular forum, so I'm glad you took the time and effort to explain it to me.

My heart is filled. I eagerly await your next insightful and profound utterance, maestro.

[HAL9000]

Trebuchet, sacrasm will get you nowhere fast.

Also, take a look at the simple physics behind a positive air pressure. You're not letting it out, therefor you are building more heat. This heat cannot escape freely, therefor takes longer to excape, since it's in there longer, it builds more heat. Put it in terms of a heat sink. If I were able to "power" a heat sink to give it a positive amount of heat, and keep the power maintained at a certain level then turn the system, it would definitely run much hotter than a CPU with a peltier on it which would be the equivilant of running it with a "negative" amount of heat.

Example 2 which I had posted before pertaining having more exhaust area than intake. I had posted this in previous threads. As an example that I have given before, I used to fly Radio controlled model airplanes and helicopters. With fully enclosed engines, it was important to have the exhaust area to be at least twice the area of the intake. The cool air intake could be quite small, but with a large enough exhaust area, the heat was easily dealt with and the engines didn't overheat. I had one person comment about propeller blast, but this is non-existant with an engine fully enclosed in a model helicopter. The cooling fan would draw air through a 2 inch circle (about 3.14" total area), but exhaust through a 4x3 inch opening in the bottom. I originally had it as 2x2 inch, but it would overheat. Increasing to 4x3 and everything ran fine.

Cool a car engine. The thermostat allows the coolant to stay in the engine longer similar to leaving the air in the case longer. What happens, the engine heats up. Now remove the thermostat which removes the restrictive flow, what happens, the engine runs much cooler than with the restricted flow. Don't want to use a thermostat as a potential arguement because it opens and closes the flow, then from my hot rodding days, you replace the thermostat with a restrictor plate. The water pump is trying to cram coolant in, but the restrictor plate keeps it from flowing out at an optimal rate. The result, again, hotter than without the thermostat or restrictor plate.

Does all of this relate to a computer, definitely as it all relates to the simple physics of transferring heat from one medium to another. Lower the pressure in your case and watch it cool off.

[Alfie]

Trebuchet,
Never said that I did,but in response to phatboi's post,i simply pointed out his temps were kind of high and in turn I pointed out what my temps were,as i've posted about v-cool and different cooling options for AMD processors,thought i'd lend a helping hand,if you haven't noticed,that's what this forum is about,now if you wish to go on about how certain temps are impossible and we must all be wrong and you are the new overclocking king who knows everything there is to know about cooling Athlon systems,perhaps you could instruct phatboi on how he can set up his system for optimal cooling.

[Trebuchet]

HAL, I think we'll have to agree to disagree on this issue. I have heard equally plausible arguments for positive pressure (Lets in cool outside air faster, keeps out dust, etc.) I know which has worked best for me. As I said, I have tried ALL those methods in that diagram. What I have now works better. I make no claim it is a perfect solution for everyone. Certainly I would be happier if my system sounded less like a 707, which is why I will soon be downvolting all my fans for those rare occasions when it is not running at full load (I crunch molecules for UD pretty much 24/7).

I appreciate your courteous reply.

Alfie, we seem to have gotten off on the wrong foot here. My apologies, I'm probably being overly touchy. I recently reluctantly left a forum I'd been at for 3 years because of obnoxious people and moderators, and I'm afraid I'm still seeing them behind every bush. I'm sorry if I took out my frustrations on you, I should think before I shove my foot in my mouth any deeper. Humble pie isn't much better.

As to the cooling problem being discussed, the initial post in this thread by phatboi was specifically requesting info on improving his case cooling. I attempted to answer that as best I could by first gathering more information, but if you've got a better solution by all means present it. I can't use CPU cooling applications, although I've used them in the past with some success, because I'm virtually always running at 100% CPU utilization. I've always viewed those apps as a nice auxilary method of cooling, but not as a primary one. So I tend to forget them.

And as you correctly surmised, I am much less familiar with AMD processors than with Intel.

[Alfie]

Perhaps I'm a little testy myself,so let it rest.
Each system varies and cooling solutions seem to apply to each system differently.
After trying different cooling approaches,the one I use now,has a fan in the bottom of the case pulling air in,a second fan is rigged to the outside of the drive cage and blows air across the memory and cpu.
I cut a hole in the bottom plate of the PS directly above the cpu's heatsink,there are 2 exhuast fans on the back of the case.
I believe this is a negative system cooling arrangement.
The original atx design by Intel had a positive design with the PS fan blowing in,they have since reversed this.
V-cool is and should not be the primary form of cooling,it simply is a good program that helps cool and has a shutdown option when preset temps are exceeded,most useful for those that may leave their computers on and unattended.
The heatsink phatboi is using should be getting the cpu cooler than it is,as it's one of the best you can get!
The main element in system cooling is the room temp.
The cooler the environment,the cooler the system.
phatboi,what is your room temperature?

[HAL9000]

Trebuchet, all I'm saying is understand basic thermal physics and I guarantee you can bring your case (and CPU) temps down a bit more. You can't run cooler by keeping the coolant in longer.

[glc]

Well - here's what I personally would do.........

Disconnect ALL fans except for the CPU fan and get a baseline temp reading. Then reconnect each fan ONE at a time and get readings - then reverse each one and get readings ONE at a time. Then try various combinations till you find the one that gives you the best temps!

I have a hunch that this will give you the best results without sounding like a 747:

Front 80mm fan blowing in
Rear 60mm fan blowing out
Fan on bracket in the middle of the case REMOVED
CPU fan blowing into the heatsink
Power supply fans running at low speed blowing out

Don't sweat a thing if the CPU stays below 60 degrees at full load on a hot day.

[phatboi]

well i checked my powersupply and for some reason, i think the fan is blowin air. i thought it was supposed to suck air out. i have the enermax powersupply w/ 2 fans. the one at the bottom is supposed to suck air right? i'm really just worried about my motherboard temp. it's pretty high, i used to get 35C wit my other board. my cpu temp seems fine to me, i overclocked it to 1.7ghz and it worked. temp stayed under 50C. i was happy. but i couldn't finish a loop in 3dMark2001 SE. i have a hard drive cooler for my seagate, it has 3 fans in front. and a harddrive cooler for my IBM which has 2fans mounted underneath. then i have the other case fans. i took out the one in the middle b/c it really didn't do anything i noticed. i am using MBM5. everything is overclocked, my videocard and processor. i can only fit a 60mm exhaust fan. so i'm thinkin' bout cutting a blow hole in the side of my case. right around the processor to exhaust air. i'm kinda hesitant b/c i have an aluminum case and don't wanna mess it up. but if it helps, i'm all for it. thanks for all the info. guys. really appreciate it.

[mike breck]

Maybe you should stop overclocking for a bit until you get your temps sorted out.

I would follow glc's advice and start your fan strategy from scratch again - adding one fan at time and monitoring the results.

I would suggest leaving the HD coolers out of the equation for the time being because it may be disrupting your thro case flow - just like the fan you removed. It seems to me you have a lot of fans that blowing air and may be counteracting each other.

If you're seriously considering cutting your case, then why not put a fan in the top (by top, I mean the roof) of the case. This is probably the most effective place for an exhaust fan because, as Hal has always said, hot air rises.

And you can't argue with that simple truth.

[Alfie]

In order to test the flow of air on a fan,simply hold a thin piece of paper up to it,if the fan holds it,the fan is exhausting,if it blows it away it's intaking,sounds better than sucking and blowing

[mike breck]

Also,

small animals and children stuck to the outside of the case = too many intake fans

guys with toupees or "brushed over hair" avoid your house = too many exhaust fans

HTH