Sign in with
Sign up | Sign in
Your question

Wierd annoying cooling problem..

Last response: in Systems
Share
February 2, 2014 4:44:00 AM

So this is my build. ASRock B85M - Haswell, Kingston 8GB DDR3 1600MHz (2x4) HyperX, Intel Core i5-4670K Haswell - Box, Antec New Solution VSK-4000E - miditower, chieftec 550w, gainward gtx 580. Now the problem is the pc will not play any games it just crashes and pc wont boot until something has cooled. At about when my gpu reaches 70 degrees.

Now I have managed to fix this problem by installing a huge old dell fan on a zalman fan mate controller underneath the dvd rom with two open bayslots. And it only works when its used as an exhaust fan. Now the pc runs great never crashes, need to put the fan up to 3000rpm for the most intensive games not to crash, tried playing for hours and hours it never crashes now. Otherwise its hovering around 2500 idle. But this setup is a noisemachine, and I dont understand why it wont work without it when temps are low all across the board. And with this setup my gpu can handle 85C++ without any problems.

Now I tried to disconnect the noisy dell fan just to see if maby the problem has solved itself, but no now it wont even load windows without it running. Just get a blue restarting errors found screen and its stuck and wont even boot before something has cooled. Btw using windows 8.1 pro.

I was thinking maby the ram gets hot and since the fan is so close to them maby it exhausts the heat from them?, also my gpus upper side is right next to them and its pretty hot.

Its just so wierd that the pc runs flawless with this big exhaust fan but its dead in the water without it. Btw all parts are brand new expect the gpu and psu (which i cleaned off dust aswell). Any help appreciated :) 
February 2, 2014 4:57:40 AM

Why dont you try an Aida or some similar program to measure all temperatures, then you will know which part makes that issues. Maybe its CPU (bad thermal paste) or a mobo chipset or ram.
m
0
l
February 2, 2014 5:01:12 AM

I use aida 64 together with msi afterburner osd to monitor all the temps possible. And their all like under 30C, highest the cpu goes is 45C and thats with hours of playing marvel heroes or crysis 3. Gpu is 60-80C but it was like that on my old dell rig and it never was a problem. I even used prime95 for 8 hours max cpu temp is like 65C and no errors. So there is something not registering on any monitor software that gets hot. But yeah my guess is either ram or chipset, as im using a cheap mini atx board for gaming cpu and gaming gpu.. but thers no way to monitor temps for chipset or ram. Mainboard temps are atleast like 30C never goes above it
m
0
l
Related resources
February 2, 2014 5:05:56 AM

Well then so it seems Memtest for few hours it is... ,but first try the bios settings and check if RAM is set up properly. There can be some Turbo mode activated which leads to those issues.

If memtest will go fine, then you can try putting some small fan over the mobo chipset, so in that case you will know thats the faulty place. If it will be then you can always swap it with some bigger radiator bought on ebay or simply get some cooler over there.
m
0
l
February 2, 2014 5:08:20 AM

yeah might try memtest, in bios i think most are set to auto and all xmp profile settings are auto too.
m
0
l
February 2, 2014 6:10:48 AM

Yeah thing is that everything is running great now with the fan im using, and it is basically right next to both the chipset and the ram slots. its like 2cm right next to them right on the top backside of the gpu blowing the hot air away and out the front.

Problem is that running the fan at 3000rpm while gaming, even idle at 2500 which is the minimum I get with this zalman fan controller. It got this annoying high pitch squeeking noise that bleeds my ears, I just feel that a newly bought rig like mine shouldnt have these problems and be needing this noisemachine :)  I have tried placing my gtx 580 further down but its only a 16x(4x mode) pcie slot. Now it dont crash either, without the fan. But ofcourse the performance in that slot is worse.

So i do think its most likely the gpu heating up either the chipset or ram modules. The card runs so hot and its so huge had to remove chassis parts to even get it to fit. Also my mainboard is so small as its mini atx, so the card is covering the entire asrock chipset module aswell. Ill try to run memtest for a few hours now and see what happens atleast.
m
0
l
February 2, 2014 2:18:53 PM

Memtest gives no errors, but I found the problem is the memory sticks actually. Caus I tried a bit without the fan and this time one of the memory modules didnt wanna boot. So I took it out and ran with only one 4gb module in pc, and now everything works flawless. So put my other ram stick and put it in my freezer a few minutes and now it works again. Have to use the fan again when using both again.. but problem is obv the memory getting to hot then, for some reason god knows.
m
0
l
February 2, 2014 5:22:23 PM

read something about changing ram command rate if its running at 1T which mine seems to be and set it to 2T should be more stable. Dunno if that does anything about the heat or makes them handle heat better? My asrock bios wont let me chance it tho it seems, and can only choose xmp profile or auto so cant manually change each setting.. they just say auto, but seems i can like backspace the "auto" and write stuff there dunno
m
0
l
!