Linux - HardwareThis forum is for Hardware issues.
Having trouble installing a piece of hardware? Want to know if that peripheral is compatible with Linux?
Notices
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
If you have any problems with the registration process or your account login, please contact us. If you need to reset your password, click here.
Having a problem logging in? Please visit this page to clear all LQ-related cookies.
Get a virtual cloud desktop with the Linux distro that you want in less than five minutes with Shells! With over 10 pre-installed distros to choose from, the worry-free installation life is here! Whether you are a digital nomad or just looking for flexibility, Shells can put your Linux machine on the device that you want to use.
Exclusive for LQ members, get up to 45% off per month. Click here for more info.
My old CPU fan was failing, so I got a new one. The specifications are:
RPM: 2200 (lmsensors say ~2300)
CFM: 42
It's "silent" fan -- not because I care about the noise, but because it's the only one that the local store had that would fit in the socket (478).
However, it overheats when the CPU is at 100% for a long time. The CPU is a 3.0 GHz Pentium4 (one of those throttling things, so I've gotten a few errors in /var/log/mcelog about how throttling turns on and off when it's doing intense things).
The old one never overheated, and it was usually at 3700 RPM (I don't know what the CFM was).
Is it correct that I need a fan that spins faster? Higher RPM = better cooling? Does the RPM even matter? Is the CFM more important? What do I need?
In fact it's not completely inconceivable that I didn't apply the thermal paste correctly. The first time it was very hot, so I reapplied it and the temperature went down 5C (!). Also the thermal paste had a very hard consistency, it was not sticky at all and you could roll it up into a little gray ball. Do I just need a better thermal paste?
Thanks in advance.
[edit] By the way, the idle temperature is between 50-55 C (depending on the temperature in the room). But if I remember correctly it has always been that high. The difference is that now it goes up to >76 C when it's compiling things.
CFM is cubic feet per minute. essentially, how much air is going through. fans of similar size, but with higher rpm's will generally have more CFM, the other (probably less important) variables in that situation are the shape, number and angle of the blades.
my knowledge with thermal paste may be a little dated here. but artic silver 5 used to be the best, all you did was put about a drop the size of a grain of rice on a clean processor and then attach the heatsink. (twas liquid, so it spread out over the whole area) The thermal paste you got sounds like some cheap mess to me. The purpose of thermal paste is to help make a good smooth contact between 2 pieces of metal by filling in the microscopic imperfections, not provide a layer of insulation in between.
bottom line, you can try reapplying the thermal paste and see what happens to the temperature. ensure you do it correctly for whatever brand thermal paste you are using. if that doesn't work well enough for you, pick up a new fan, they are cheap on newegg and other sites. Might as well pick up some good thermal paste while you are at it. used to cost all of 5 dollars for a tube of AS5 back in the day, which was enough for quite a few processors.
You definitely need better paste. You don't need anything super expensive, but silver or zinc based pastes are preferable. Make sure the pastes are fluid not hard. If the paste came with the cooler you should usually discard it as it is often bad quality hard paste.
I agree on the silver or zinc based pastes. Plus, using a less powerful fan if the box has an extra spot for another fan couldn't hurt? My CPU is never at 100% especially for long but I guess it depends on what you do, I don't know much about throttling tho could be an issue or help and finding what possess are so intense?
I agree on the silver or zinc based pastes. Plus, using a less powerful fan if the box has an extra spot for another fan couldn't hurt? My CPU is never at 100% especially for long but I guess it depends on what you do, I don't know much about throttling tho could be an issue or help and finding what possess are so intense?
It can't be helped. Building some programs (ghc) takes several hours and it will be on 100% the whole time. Burning CDs also makes it go up to 100%.
Supposedly the Pentium 4 processors shouldn't be damaged by overheating because of the automatic throttling, but it can't be good.
I checked the BIOS clock settings (to underclock it) but they are locked and can't be changed.
The amount of heat a CPU cooler can transport is not only determined by CFM, but also by form and material of the cooler itself. If such a cooler is aimed at 90W CPUs it simply won't work reliably with a 130W CPU, regardless which fan you mount on it.
To determine which type of CPU you have you need the sSPEC number which is printed on the CPU. If you know the sSPEC number you can find your CPU on Intel's website and look what thermal design power your CPU has.
After that determine the maximum amount of power consumption that cooler can work with (usually printed on the packaging). If it is below the thermal design power of the CPU you need a better cooler.
Not having a shot at you Soderlund, but there is more to heatsinks and fans than just the fan CFM and RPM. The heatsink design and materials can make a big difference as well.
I dont know what heatsink you had before, or what you have replaced it with, but I'll assume that the heatsinks are fairly equal (big assumption there) and the main difference is the fan RPM/CFM, and the new heatsink has a smaller and/or slower fan. If thats the case, then your 'peak' temps will be higher, but idle temps can be very similar.
Its easy to replace the fan on a heatsink. I probably would have done that before buying a whole new heatsink and fan. You can even use 'ghetto' mods like replacing a 70/80mm fan with 120mm fan..yeah, its going to be lareger than the heatsink, and wont mate up to the mounting holes..but you can just use one mounting hole then use a ziptie or 3 to attach the fan better.
Ugly, but cheap and very effective. I've pulled that trick many times to make a system quieter. Replacing a 70mm/6000RPM/50CFM/47dBA fan with a 120mm/1300RPM/55CFM/20dBA fan makes a big difference to noise, and the temps should be very similar.
Yeah, artic silver is good stuff..but its not as amazing as some people make out. Toothpaste can beat it!
Quote:
Anyway, I greatly reduced the amount of grease on the bottom of the cooler and tested again, and now Arctic Silver 3 scored 0.48°C/W. This is about as small a performance difference as my test rig can discern. But hey, the thinner layer did work better.
Snip!
Firing up my Minty Fresh Thermal Transfer Experiment produced, rather hilariously, an excellent score - 0.47°C/W.
That's right - Toothpaste Superior To Arctic Silver 3!
Its a good article to read if you dont know a huge amount about thermal paste.
Just to make things even more complex, some 'dry' thermal paste can be good as well (like the expensive diamond based compounds). But as far as the standard silicon based stuff, and even the silver based thermal pastes goes, if its dry its no good.
Thanks for your replies. They are very informative.
Before turning on the computer to read your replies, I picked it apart and applied the new paste as soon as I came home today. (So I didn't pay any attention to the text on the CPU.) I saw that the old paste had been squashed out to a tiny little circle covering about 1/9 of the processor's area; I didn't see that earlier because you have to re-apply the stuff after you remove the heatsink. But I'm quite sure it was the right amount, so it must have been like a thick cylinder between the CPU and the heatsink.
I put on the new paste (white, sticky, fluid) and made sure the whole thing was evenly covered. Now the average idle temperature is consistently between 50-52C. I can't get it above 65C now, but I haven't tried any long-running task yet.
If it doesn't work, I'll do as you say, TobiGSD. I never checked if they matched.
I might try a case fan like you say, aus9, because the temperature is 38C when I start it up, and it slowly climbs to 50C. That seems to indicate that the temperature builds up in the case, or am I wrong? It is a tower.
Opening the case makes a big difference, by the way.
Quote:
Not having a shot at you Soderlund, but there is more to heatsinks and fans than just the fan CFM and RPM. The heatsink design and materials can make a big difference as well.
Don't worry, I suspected that. I would never have guessed that toothpaste could be equivalent though.
Quote:
I dont know what heatsink you had before, or what you have replaced it with, but I'll assume that the heatsinks are fairly equal (big assumption there) and the main difference is the fan RPM/CFM, and the new heatsink has a smaller and/or slower fan. If thats the case, then your 'peak' temps will be higher, but idle temps can be very similar.
Its easy to replace the fan on a heatsink. I probably would have done that before buying a whole new heatsink and fan. You can even use 'ghetto' mods like replacing a 70/80mm fan with 120mm fan..yeah, its going to be lareger than the heatsink, and wont mate up to the mounting holes..but you can just use one mounting hole then use a ziptie or 3 to attach the fan better.
Ugly, but cheap and very effective. I've pulled that trick many times to make a system quieter. Replacing a 70mm/6000RPM/50CFM/47dBA fan with a 120mm/1300RPM/55CFM/20dBA fan makes a big difference to noise, and the temps should be very similar.
I'd say the old heatsink was about 50% larger, otherwise they look pretty much the same.
I had no idea you could do that with CPU fans. I know someone who is good at that type of modifications though -- he attached some metal thing that he had made himself to a harddrive so he could fit a fan on it.
Maybe I should use the old heatsink with the new fan!
I would never have guessed that toothpaste could be equivalent though.
It is, for a short while. Long term, it would just dry out and become worse than useless.
Quote:
Originally Posted by Soderlund
I'd say the old heatsink was about 50% larger, otherwise they look pretty much the same.
I had no idea you could do that with CPU fans. I know someone who is good at that type of modifications though -- he attached some metal thing that he had made himself to a harddrive so he could fit a fan on it.
Maybe I should use the old heatsink with the new fan!
Its hard to know for sure, but it might be a good idea. Unless there is some difference that you're not seeing or discounting (e.g., the old heatsink is bigger, but has less fins that are more widely spaced) I'd doubt that temps will be any worse. If the heatsinks are very similar overall, just the old one is bigger it should help your temps, at least a little.
It explains it much better as I could have done in a post here. It all boils down to creating a low thermal resistance between CPU core and ambient air. The less resistance, the easier the heat flows away from the processor core. Having said that, if the ambient temperature is high, the CPU temperature will be higher as well. And the inside of the case is to be considered as ambient. But a low capacity fan to slowly flush the air through the case should suffice to solve that problem.
LinuxQuestions.org is looking for people interested in writing
Editorials, Articles, Reviews, and more. If you'd like to contribute
content, let us know.