Sign in with
Sign up | Sign in
Your question

New Case or More Fans?

Last response: in Overclocking
Share
June 26, 2011 12:28:44 AM

I was having some pretty bad heating issues for a while after building my first PC, and while I have finally gotten them under control, I have also developed somewhat of a passion for cooling (I guess I am a little weird. When I started things off I was hitting over ~90C at load due to an odd automatic overclocking set by my MoBo and a poorly set and poorly manufactured stock heatsink with a terrible thermal pad. After fixing the overclocking issue and resetting the heatsink with arctic silver 5 I got it down to ~79C, and with my new heatsink installed, I hit ~68C. My goal is to get it down to in the low 60's at load with low noise. I am sitting between a few options right now.

(1) I can make the big purchase and move up to the FT02 case which is supposed to be extremely quite and cool with its high diameter, low RPM fans, high airflow 90 degree MoBo design, and large amount of acoustic foam.

(2) Stick with my Antec 300 case and upgrade the fans. It currently has only the stock fans it came with, and I am running them on medium RPM which can get a little loud. I would not only replace these two fans (120mm rear fan and 140mm top fan) with quite fans, but also add a quite 120mm side "cpu" fan, as well as the two quite front 120mm intake fans. This is a cheaper option, but I would not get the build quality of the other case, the acoustic foam, or the 90 degree MoBo design.

I am also new to lapping and am considering this to reduce the need for the case and fans to pick up the slack.

*edit*
spelling and retroactive signature

*Update*
I just reinstalled my heatsink/fan as a few of you have suggested. This resulted in a slight improvement of temperatures at the low and high end.

0% Load Core Temp Range
Before: 29C – 41C
After: 26C – 36C

100% Load Core Temp Range
Before: 66C – 69C
After: 63C – 67C

I would still love input on my two options I mentioned. Would either reduce these temps and noise by a noticeable degree? I also noticed a few people suggested a completely different heatsink/fan. How much of a difference will that make in my temps (i.e. will the drop in temperature justify a $50 to $100 purchase; will the drop only be a few degrees, or over 10C)?

More about : case fans

June 26, 2011 10:01:17 AM

what is the current mobo, cpu, heatsing, case, psu.
m
0
l
a c 197 K Overclocking
June 26, 2011 12:21:12 PM

If you are getting these kind of temperatures, unless you have high ambient temps, you have something wrong with your setup.

The Antec 300 is a great budget case. If you are overclocking though, you do need more fans. I have several Antec 900 cases - think 300 with most of the extra fans included: 2 front, one back, and 1 top. I have Core2 systems overclocked to the edge of stability in them. My definition of stability includes "24 hour Prime95 stable with SpeedStep enabled".

I also do not exceed Intel's max recommended values for core voltage or load temps.

My fans are all set to LOW. Higher speeds did not improve cooling. Adding a side fan decreased cooling.

Lapping is generally a solution looking for a problem. Check the base of the heatsink for flatness. A steel machinist's rule works for this. Hold it on edge on the heatsink base. You should see very little or no daylight between the rule and the heatsink.

Most modern heatsinks have pretty flat bases.
m
0
l
Related resources
June 26, 2011 7:52:44 PM

@bisbang
Sorry, I had put my build in my signature, but I had never checked apply signature. It should show below now.

@jsc
When I was getting the terrible ~90C and ~79C temperatures, it was in a ~17C room during the winter. I did not have the money or time to install a new heatsink until this summer when the temperatures in the room are ~30C without the AC on (like it was the day I tested).
m
0
l
June 26, 2011 9:41:50 PM

is this temperature just on ur cpu? whats ur mobo temp?
m
0
l
June 26, 2011 9:57:25 PM

bigbang said:
is this temperature just on ur cpu? whats ur mobo temp?


Can you suggest a program for monitoring MoBo temperatures while running Prime (so I can test at load)? I currently use RealTemp to monitor my temperatures because I like its integration with Prime, but it only gives the CPU core temps. What reason would their be to look at my MoBo temps separate from my CPU?

I'm not really sure why everyone is worried about these temperatures. When I posted about this before, other users did not take issue with the temps where they were. At 100% load they are still below the conservative suggestions from Intel.
m
0
l
June 26, 2011 11:45:19 PM

if ur mobo temps are just as high, it indicates an air circulation problem.
but i bet that you havent applied the thermal compound correctly
m
0
l
June 26, 2011 11:45:44 PM

u can download asusprobe from asus which shows ur mobo temps too
m
0
l
June 26, 2011 11:46:35 PM

lower ur voltages/fsb, whats the lowest temp u can get ur cpu to run at?
m
0
l
June 27, 2011 1:04:58 AM

bigbang said:
lower ur voltages/fsb, whats the lowest temp u can get ur cpu to run at?


Are you asking about my idling temperature (0% load)? Running Realtemp, the cores range from 29C to 37C idling with my room temperature around 26C.
m
0
l
June 27, 2011 8:42:59 AM

damn. well lowering the clock and increasing the ratio can drop temperature in some cases. u havent increased the NB-cpu ratio have u?
m
0
l
a c 121 K Overclocking
June 27, 2011 5:26:45 PM

Be careful how you start down the quiet computing vs cooling road. There is no end to it.

You can improve cooling with faster fans, at the expense of noise.

You can improve both noise and quiet at the expense of cash. But there are diminishing returns.

Here are some ideas:

CPU cooler. The Zalman 9500 cnps is an attractive cooler that is better than the stock heat sink.
But it has a 92mm fan which will be relatively noisy considering the air it pushes.
A review database at frostytech.com shows it at 29.6c over ambient under load, and 53.3 dba.
http://www.frostytech.com/articleview.cfm?articleid=252...

The top units on the chart are 14 or so over ambient, and more like 50dba.
Check out the list, and consider something like the Noctua NHD14 as a replacement.
---see, more expensive---

There are many lists of thermal compounds, but I don't think the differences are that large to be really important.
Arctic silver-5 or MX2 are long time favorites, and are not expensive.
When applying thermal compound do not make the mistake of using too much. It will act as an insulator. It is hard to use too little.

I think the case is fine. Your idle temps are good.
The stock fans are OK, but noctua and scythe fans are a bit more efficient and quiet.
You could replace them.
There is also a device called a Scythe kama bay unit that can replace three front 5 1/2 bays with a 120mm intake fan.
---see, more expensive---

An important component of cpu temperatures is your graphics card. Particularly if the graphics card cooler dumps heat back into the case, heating up the ambient air that the cpu gets for cooling. Exactly which GTX460 do you have?
You could replace it with a GTX560ti which will run cooler.
---see, more expensive---

For more research on quiet computing, go to www.silentpcreview.com
They have good info, but are not particularly oriented to gaming systems.
m
0
l
June 27, 2011 5:37:39 PM

I believe the temperatures u have now 68C at full load running prime are not at all bad. Under realistic normal usage / high usage conditions ur temps should be much lower. Could u report back with those temperatures? I would think they will be quite normal.

I couldn't help but bring this up. That GPU is a little low for ur setup. What do u use this rig for?
m
0
l
June 27, 2011 6:01:39 PM

oi! gtx460 is a decent graphics card
m
0
l
a b K Overclocking
June 27, 2011 6:37:15 PM

FT02 all the way you will never look back and team it up with this CPU cooler, i have it and it has amazing cooling and is very quiet.

Be Quiet! Dark Rock PRO BK016 CPU Cooler Socket LGA775/1155 /1156/ 1366 /AM2/ AM2+/ AM3/ 754/ 940


Heres my set up.


m
0
l
June 27, 2011 7:17:27 PM

truely mahoosive fan
m
0
l
June 27, 2011 7:43:27 PM

@geofelt
I feel satisfied with the noise levels I am getting from the Zalman 9500 cnps. This may be because it is simply being drowned out by the case fans at this point, but the CPU cooler/fan seems quite enough for me. I am not looking for whisper quite, but just want to cut the noise back some. I will consider a new CPU cooler/fan in the future if it comes to required lower temps however as the over 10C difference is more than I would have expected.

You are one of a few people who has said that I may have applied the thermal compound incorrectly, so I think I will finally go in an reinstall the heatsink since this is such an easy task anyway.

@gametstr
I will check the temperatures while running some regular programs. The 1GB 460GTX is plenty powerful for for video editing, which is what I built the PC for. Video editing is processor heavy and the GTX 460 is actually overkill.

@Uther39
Man that case is beautiful. Do you have any complaints about it? I checked out the cooler as well, but (1) I just bought mine recently, and (2) that thing is supposed to be heavier than sin, which worries me a bit.
m
0
l
June 27, 2011 10:08:40 PM

I just reinstalled my heatsink/fan as a few of you have suggested. This resulted in a slight improvement of temperatures at the low and high end.

0% Load Core Temp Range
Before: 29C – 41C
After: 26C – 36C

100% Load Core Temp Range
Before: 66C – 69C
After: 63C – 67C

I would still love input on my two options I mentioned. Would either reduce these temps and noise by a noticeable degree? I also noticed a few people suggested a completely different heatsink/fan. How much of a difference will that make in my temps (i.e. will the drop in temperature justify a $50 to $100 purchase; will the drop only be a few degrees, or over 10C)?

*I will be copying this post into my original post for new viewers of the topic because I know a lot of people do not want to read each and every comment to participate in the conversation.
m
0
l
June 27, 2011 10:22:31 PM

damn!!! i thought i was wrong. but i still think u put too much thermal paste. call me paranoid -.-
m
0
l
a b K Overclocking
June 27, 2011 10:30:44 PM

Absolutely no complaints about the FT02, its the best case ive ever owned and trust me ive had alot, as ive been building my own PCs since 2002, these include Antec 300, 902 and the 1200, Fractal Design cases, Thermaltake Armour & Mozart just to name a few. The FT02 is understated Art, Form and Function combining in a perfect marriage, truly the best and quietest Air Cooled Case money can buy and before you ask, no im not a share holder at Silverstone, haha.

As for the Be Quiet CPU cooler, it fits with the FT02 like a hand in a glove, both working together to not only bring amazing cooling ability at almost silent levels but also enhancing the aesthetics of your internal layout, resulting in you not knowing whether or not to look at your PC or the monitor. The weight is of no concern either as it comes with a very large back plate to spread its heftiness across the motherboard. I have heard people complain about the coolers installation process, saying it was fiddly, but i found it to be just as easy as most after market coolers, to install.
m
0
l
June 28, 2011 3:30:45 AM

Get all the fans going you can. Three fans if thats all that fit. I do not know if your GTX460 is blocking air flow in this case. Or your cable management.
m
0
l
June 28, 2011 10:37:20 PM

@uther39
You're making me lean further and further towards getting that case. It seems like everyone that I talk to who actually bought the case has no remorse and recommends it to everyone considering it.

Probably going to stick with my current heatsink/fan for now. The one you suggested seems like it works great, and it definitely looks good, but since I am not a fan of windows myself, I would never see it anyway.

@walterm
I do not believe the GTX is causing any issues. It runs parallel to the exhaust fan of my CPU cooler, with the cooler being above it. Since hot air rises, and the main exhaust from my case is the top rear, this would cause an issue. While I do have a few cables that may be blocking some airflow, their is no alternative way to route them in my current case. Everything that can be hidden away is.
m
0
l
a c 145 K Overclocking
June 29, 2011 3:59:04 AM

Lapping is a good idea on a poorly manufactured / finished HS....but it voids the warranty on both the CPU and many hi end heat sinks which are designed with a specific profile which is destroyed by lapping. For example:

Quote:
Prolimatech does not condone any type of lapping done to the CPU or to heatsink base. Every Prolimatech's heatsink base is designed on a pin-point scale of how the base is to be flat and/or curved where it's needed to be. We have programed our machines to machine the surface in a very calculated way. Any after-manufacture lapping or modding done to the base will alter the design, hence negating its performance factor as well as its warranty.


Knowing your CPU and Heat Sink would be nice but when looking for reviews, make sure that the tests are on the processor die you are using. For example, frostytech uses a "Mk.II Synthetic Thermal Test Platform - Intel LGA775 Version" .... as CPU's for this platform present an entirely different heat signature than the 1366/1155/1156 platforms, the data results are of lesser value than those tests on the specific platform you are using. For 1366 / 1155, one of my 1st stops is here:

http://benchmarkreviews.com/index.php?option=com_conten...

That's a quarterly roundup, you can see how the later release stand up by looking at the latest tests in each category:

http://benchmarkreviews.com/index.php?option=com_conten...
http://benchmarkreviews.com/index.php?option=com_conten...

The temps you are seeing look appropriate for an overclocked 2600k at about 4.5 - 4.6 Ghz under Prime95 with a decent cooler.

The Scythe Mugen 2 is a standout cooler (SCMG 2100) .... AT $40 it outperforms the comparably priced Hyper 212 by about 7C, and comparing with coolers costing twice as much /..... beats the DH-14 by 1C and matches or in some tests beats the venerable Megalems.

As for the AS5, check the AS website and read about the curing issues on the home page. Expect performance to improve after over 200 hours of thermal cycling from 70C back down to room temp. Shin Etsu matches AS5's thermal transfer performance but w/o the curing issue.
m
0
l
June 29, 2011 7:30:24 AM

you know noone actually asked u the most important questions, what voltage/frequency is ur cpu running at? cause if u got that temp at 5Ghz or something thats just perfectly normal or even magically low
m
0
l
June 29, 2011 8:02:21 AM

bigbang said:
you know noone actually asked u the most important questions, what voltage/frequency is ur cpu running at? cause if u got that temp at 5Ghz or something thats just perfectly normal or even magically low


Unfortunately, I'm running without any OC at the moment.
m
0
l
June 29, 2011 4:29:37 PM

well according to zalman, the zalman cnps9500 has a thermal resistance of 0.16°C/W (in silent mode!), the 2500k's overclocked power consumption is roughly 358W, under load, so that translates to roughly 57 degrees celcius. thats how much it should be running at. i can point at 3 scenarios.
the CNPS9500 doesnt officially support the 1155 sockets, so it might be that you have been forced to install it in a skewed angle blocking the airflow. refer to this website for an instructive image:
http://www.zalman.com/eng/product/Product_Read.asp?idx=...
but guessing by the fact that 775 and 1155 have very similar dimensions i guess ull cross this one out.
other option is that the values they are giving are unrealistic for your particular case and theres a general lack of fans or air blockage inside. i again think this is highly unlikely.
and the last option is that the thermal paste hasnt been put on correctly. unfortunately most articles on the web are misguided or dont point out to the main principles. firstly you must realise that the thermal paste itself is not a good heat conductor, not even comparably close to the metal copper or aluminium of ur heatsink/cpu. the purpose of the paste is to cover the infinitesimally small holes on the surface the metals of the heatsink/cpu. so get a pair of plastic gloves. put the thermal paste on and start rubbing really hard on the surface, both on the heatsink and on the cpu. you can also do it with pressing a credit card shaped card but it doesnt work aswell as rubbing. at the end you want to be able to see through the paste on ur heatsink/cpu. so you want to be able to see the metallic surface everywhere, but ull notice bits of the thermal paste have gotten stuck in places you didnt realise there were holes. and you dont want giant patches of thermal stuff on the surface anywhere as that would imply a metallic irregularity in the surface. I can't for the life of me find a single picture of a correctly applied thermal paste on the net. thats how simple people think it is, and miss the point entirely.
there is a way of testing wether you have applied the thermal paste correctly. using a thermometer at the base or midway on the heatsink and finding the difference between the cpu and the heatsink. in general there will be fluctuations and you should be aiming to minimise it by using the thermal paste.
the CNPS11X from zalman (also) promises to give you a 10% increase in thermal resistance (according to the website numbers) so achieving around 50 degrees on silent i would imagine.
all the best
** have u checked that ur heatsink is not wobbly?**
m
0
l
!