Cooling the space behind mobo

[psykoikonov]

Hey Guys/Gals heres one for ya.
I bought a new case and had thought alot about the air space between my mobo and the cases ATX board (you know the 3mm gap where your mobos standoffs are). Sooo I bought a 80mm hole saw and cut a hole in my case side and the case's ATX board in a straight line right behind where the CPU mounts. I then put a fan blowing out in the case side and hooked it up. Well, I dropped 8C on my 1800+ running at 146FSB at idle. I then ran toast for an hour and didn't even make 40C (39 according to Asus tool and a quick check in bios immediately after showed 37C brrr). I played Unreal II for an hour or so, hopped out and checked again Asus util ... 35C, mobo at 25C. Air cooling is great, as long as you move ALL the air. That air has been imprisoned back there for months and I finally freed it and it thanked me!! BTW room temp is 20C, internal case temp is 22C, and air temp just to the right of HSF is 26C now (all acurately measured with digital thermometer). The case has a total of 7 80mm fans, 1 120mm fan, and 2 5.25mm hdd coolers (thats just the case btw).

Psyko


<A HREF="http://forums.btvillarin.com/index.php?act=ST&f=41&t=421&s=205efc501496823a64ba7fbd37152ff2" target="_new">Psyko's Rig</A>

 

[cjmurph]

awesome idea, follows the lines of logic quite well. Now where did I put that hole saw...

 

[LCARS]

I dunno... you didn't put it anywhere with something that can make a pile did you?

Everybody knows the hole's only natural enemy is the pile.

 

[cjmurph]

Hang on I'll check...

oh there it is, thanks.

 

[jihiggs]

thats the reason i liked the design of the slot proccessors, the other side of the cpu was more open to air circulation.

my computer is so fast, it completes an endless loop in less than 4 seconds!

 

[svol]

Yes... but the reason that it sucks is because you can't place big HSFs on it without touching the mobo or memory.

Anyway I'm going to try this as I'm planning to work around in my PC anyway. But I think I add a small HSF to the back for more cooling pleasure.

My dual-PSU PC is so powerfull that the neighbourhood dimms when I turn it on

 

[phsstpok]

Sooo I bought a 80mm hole saw and cut a hole in my case side and the case's ATX board in a straight line right behind where the CPU mounts. I then put a fan blowing out in the case side and hooked it up. Well, I dropped 8C on my 1800+ running at 146FSB at idle

I'm a little skeptical.

The backside of your motherboard has no direct access to your CPU. Between your backside fan sits the PCB which is your motherboard. Above that is 1/4 inch (or so) thick CPU socket. These are insulators rather than conductors of heat. Now considering how much we have to do on the topside, bigger massive heatsinks, powerful fans, etc, to make a difference as big as 8 degrees. Doesn't it seem unlikely that a fan alone, on the backside, could do this much cooling.

I think it is more likely that all you are managing to do is to cool the sensor that sits in the CPU socket without cooling the CPU one bit.

If you are really getting an 8 degree improvement then you should be getting measurably better overclocking. Are you?

Sorry, just one opinion.

<b>99% is great, unless you are talking about system stability</b>

 

[Lonemagi]

I had seen a test online once that tested the thoery of cooling the back of the motherboar. It showed that you could gain a few degrees in ambient temps, but cpu was not affected for the same reasons that you posted, phsstpok.

He might have been getting better airflow within the case because ofthis, and that might have a good effect on the cpu, but thats all I can think of.

<A HREF="http://tekkoshocon.com/" target="_new">http://tekkoshocon.com/</A> Southeast Pennsylvania gets an Anime Convention!

 

[phsstpok]

He might have been getting better airflow within the case because ofthis, and that might have a good effect on the cpu, but thats all I can think of.

Exactly! He could put an extra exhaust fan anywhere and achieve the same effect.

<b>99% is great, unless you are talking about system stability</b><P ID="edit"><FONT SIZE=-1><EM>Edited by phsstpok on 02/21/03 08:27 PM.</EM></FONT></P>

 

[psykoikonov]

Post deleted by psykoikonov

 

[phsstpok]

Not anywhere phsstpok, as then you still don't move the air behind mobo and it sits there and gets hotter and hotter. You guys continue to flame the chemical engineer who took 12 thermal dynamics courses in 3rd and 4th year LMAO. I'd already guesstimated/calculated the results to be around 7C before I did it based on the fact that the air is not moving at all. Don't believe me, then instead of implying heresay or some lame sites results why don't you do it.
PS: Does your post mean that if we all add another exhaust fan we drop 8C. Who needs water cooling then. Think I'll add 10 more and go -C LOL

Where did I flame you? Where did I cite anything? Where did I call you a liar? Where did I make a personal attack of any kind?

Before you accuse someone of flaming be sure they are.

All I did was state that I didn't see how your claimed cooling method would work and I expressed that I thought the motherboard sensor was being cooled but not the CPU. I did not claim you aren't getting the readings you say. I implied nothing except that I thought you were interpreting the results incorrectly.

If I erred then show me where I am wrong. You don't need to insult me.

My doing the same experiment won't prove your claim because I will still believe the sensor is being cooled and not the CPU. Besides my case has vented panels so the air is not so dead behind my motherboard.

Even without vented panels why would there be dead air? There are gaps both above and to the side of the motherboard. It's not a sealed off space. Plenty of room for air to circulate behind the motherboard, at least in my case. My case also has a removable motherboard tray, with openings. Air circulates through there too.

All this chest beating is getting us nowhere but I have a way to prove one way or the other. Find someone that has a motherboard that supports the CPU's thermal diode. Have them do your test.

If that shows a cooler CPU then I will believe. If it doesn't will you believe?

PS: Does your post mean that if we all add another exhaust fan we drop 8C. Who needs water cooling then. Think I'll add 10 more and go -C LOL

No I don't really believe that unless the case didn't have any cooling fans to begin with. I didn't realize that you have 8 case fans.

I would think that someone with 12 thermal dynamic courses under their belt can do better than 8 case fans. LOL!

[Now that was personal attack]


<b>99% is great, unless you are talking about system stability</b>