Entire Computer Running Hot

My entire computer has been heating up, as of late. I fixed my last graphics card issue, and it had been running fine for the past few weeks. But yesterday and today, everything has been hot. The motherboard and GPU are both running 60 C, despite the fact that I'm not doing anything heavy. The CPU is running between 50 and 55, which is almost a 10 degree change from where it was two days ago. But the most prominent source of heat has been my power supply. It's spewing extremely hot air, much hotter than the rest of my computer, which leads me to think that it's the source of my woes.
Long story short, my computer is hot and I think it's my power supply's fault. Everything has been dusted thoroughly, and all fans are working, so I'm not sure what the issue is. I don't think it's the sensors, because the entire computer feels pretty close to what Speccy and MSI Afterburner are saying.

 

140 Fahrenheit is not that hot. Computers do not start shutting down, until they get more around 210 degrees. As long as you have good air flow, the house or room temp is not the same as it is outside, which means keeping the house air conditioned, you will be fine.

There are programs that sit in your taskbar that will warn you when the machine gets and stays over 190, then start the shutdown.

 

Well, I don't see anything wrong with your temps. 60°C for a GPU is barely warm. And 50° - 55°C for a CPU is still in the "warm" range. I don't think about worrying until my CPU temps sit and stay above 60°C.

There are many things that affect entire computer temps - the ambient (room) temperature is one of the main things. Has the computer's "operating environment" (the room's temperature) changed with the coming of Summer? Or is the room air conditioned and the environment stays fairly constant?

Have you scanned for malware? You say YOU are not doing anything heavy, what is your computer doing? When you are idle, what does Task Manager say for CPU and RAM utilization? Note when you go idle, Windows will start running background processes - so some activity is ALWAYS going on. But not a lot, and TM (or TM's Resource Monitor) should show you what that is.

MSI Afterburner? If you are overclocking, STOP! At least until this is resolved.

The PSU pushing out more heat than normal could simply mean it is being pushed harder by the computer inside demanding more from it. It is normal for the air from the PSU to be much "warmer" than the rest of the computer - that small confined place is where all the power conversion takes place, and that AC to DC conversion takes a lot of work. Even the best (most efficient) PSUs lose 15 - 20 watts for every 100 watts generated - and it is lost in the form of heat! If your motherboard, CPU, GPU, RAM, drives, etc. are pulling 300 watts from the PSU, the PSU is pulling around 360 watts from the wall - and that 60 is seen (or felt) as heat - as much heat as from a 60w lightbulb.

But, still it should never be "hot" (where the air feels like it might burn you). Of course, "warmer than usual" may be perceived as "hot!", if it is unexpected.

Still, anything that plugs into the wall can kill - there are deadly, destructive voltages and current exist inside computer PSUs. And ALL of the many hardware components depend on quality power. So as an electronics technician, I always want to ensure I have good, clean stable power - right at the very start of the troubleshooting process (after cleaning the interior of heat-trapping dust and conducting a visual inspection of the insides for loose cables, etc.). That just makes good technical sense - even if the PSU is not suspect.

The only true way to thoroughly and conclusively test a PSU is while the PSU is under a variety of realistic loads using an oscilloscope or power analyzer. I advice against, and do not recommend using a multimeter to test PSUs. For one, most do not test for ripple. And to test the voltages properly (under a variety of proper loads) requires jamming two sharp, highly conductive meter probes into the heart of the motherboard where one tiny slip of an unsure, unsteady hand can slice through dozens of circuit traces. Not to mention potential ESD damage to ESD sensitive devices. So my suggestion is to do what I always do, swap in a known good PSU of ample power and see what happens.

If you don't have a spare PSU you can use (or borrow), get a PSU Tester. The advantage of this model is that it has an LCD readout of the voltage. With an actual voltage readout, you have a better chance of detecting a "failing" PSU, or one barely within the required tolerances as specified in the ATX Form Factor PSU Design Guide (see “Table 2. DC Output Voltage Regulation” on Page 13). Lesser models use LEDs to indicate the voltage is just within some "range". However, none of these testers test for ripple and they only provide a little "dummy load", not a variety of "realistic" loads. So while not a certain test, these testers are better than nothing. They are also great when using a spare PSU for testing fans and drive motors as they signal the PSU to turn on when plugged in.

If your system is clean of malware, you don't have any processes going wild pushing your resources and demanding more power, and your PSU checks out okay, then look at your case. It is the case's responsibility to provide sufficient cool air flow through the case. I generally recommend at least one large (120mm or larger) fan in back (in addition to the PSU fan) pulling exhaust air out, and one large fan in front, pushing cool air in. Not all cases support front mounted fans so look for additional mounting positions in the back, or even on top for a "blowhole" fan (which are very efficient). Some cases come with 80mm fans, but support larger (92 or 120mm) fans. Look for additional fan screw holes for larger fans.

One final question - does the computer work fine otherwise?

 

Why not reformat it. Enter Bios system and adjust the speed of fan.

 

Why do that, when there is nothing wrong with the install, nor with the temps.

 

Reformatting should ALWAYS be a last-resort option - after all other options have been tried and failed.

You lose everything with a format - all your data, any downloaded programs, all your security, and you put your computer months, if not YEARS behind in critical updates.

 

Agreed

Reformating will not help this issue it has to deal with Air Flow from the fans if the fans are not pushing enough Air Flow then of course your computer is going to get hot.

Recommending that the Fans are working Need to be cleaned with a Can of Air.

IF Any of your Fans are not working then i would check them all while the computer is running.

 

Ok, lets separate fact from fiction.

60c, if an accurate reading, is incredibly hot for an idle temp on a motherboard or a gpu.

It is virtually impossible for a cpu to run cooler then your ambient (motherboard). The cpu will be hotter, especially loaded.

My guess is something is amiss with your VRM (voltage regulator module) section of that motherboard. Or else these readings you are getting are inaccurate.

Are these IDLE temps being reported???

 

Well i would have to disagree 60 c is Very hot and very close too the GPU's Max Temp rating

Mostly all GPU's can handle 160 F But in a sense 140 is a little close for comfort.

You did not tell use what kinda Video card you have.

 

I think it may be the temperature in the room, seeing as it cools down it night and we don't have air conditioning. I stopped overclocking my graphics fan, for the time being, which provided no significant change. I have an Nvidia GeForce 8800 GT. All of my fans have been cleaned and are running properly, and the side of my case is ON. My computer is in a place where it can draw in cool air and expel warm air, not trapped under a desk or next to a printer or anything. All of the temperatures I've been reporting have been idle temperatures.
How do I go about looking at the VRM section of my motherboard?

 

Yes, hot temps in a room can do it. I have lived in a trailer that would get to 110 during the day, and if you were lucky when you ran fans in there, you could get it down to 95, maybe 92, and at night maybe 86.

I keep my air at around 72, and my stuff stays maybe at the max around 114f. I have seen it get as high as 128f to 132f, so really no big deal for a few hours. Running it at high temps all of the time, no matter what the electronic equipment is, is just shortening its life.

You really should get some 8,000 to 12,000 btu room a/c units to run, if you plan on using computers. May not get the room down to 68-70, but at least if you can keep it around 74-76, things would be better for the computers, tvs, etc, including your fridge.

 

Please post your specs. You can use Everest, or Speccy if you'd like.

 

Why. We have already determined the nature of the problem. Too hot of an environment. Those are not going to do anything other than satisfy your curiosity. Will not magically make the environment of the room cooler, or tell you why the room is too hot.

When it is 95 outside, it is 95 inside at the most. Start turning on electronics like tv's, computer, you can turn the room into a oven, and nothing can be done to cool the room at that point, but get a room a/c, if the apartment or house does not have central air.

 

I agree with Goldenskull. While 60°C is pretty warm (and 60°C is my personal threshold for starting to get nervous) 60°C is not "very hot", or even just "hot" for a CPU. And definitely not for a GPU. However, I do agree with LTLSU that 60°C at full system "idle" is a concern, as it leaves little wiggle room when the computer is being tasked.

No - sorry. While a proper layer of TIM (thermal interface material) and a properly matched HSF assembly are essential, NO conventional heatsink and fan can cool a CPU cooler than the temperature of the air being blown on to it. The Laws of Physics don't allow chunks of metal to be affected by "wind-chill factor" - that's living flesh only.

So if the room is 100°F, there is no way the CPU can be cooled lower than that - unless using an "active" alternative cooling solution - such as refrigeration or Peltier cooling. As I noted before, "It is the case's responsibility to provide sufficient cool air flow through the case." But even the case needs the ambient temps to be in the "comfort zone" to do any good.

Also, TIM does not "go bad" or get old on its own. It can, and will last for 10 years or longer, AS LONG AS the cured bond between the mating surfaces is not broken from bouncing the PC off the floor, or twisting on the heatsink. So if it's possible the cured bond of the old TIM has been disturbed, then thoroughly cleaning the mating surfaces and applying a proper (can't get too thin) layer of TIM is a good idea - as long as essential ESD precautions are taken (unplug from wall, touch bare metal of case interior BEFORE and while digging around inside).

 

Sorry, I did not mean to imply you said that. But when you said, "efficient heat sink, coupled with a good fan, laughs at outside temps", that suggested to me (perhaps incorrectly) that outside temps don't matter if you have a good cooler. I took "outside" to mean outside the case, not outside the house. If you meant outside the house, then again, my apologies. If you meant the computer's ambient, or room temps, then I stand by statement. And I picked 100° as number all would agree is a hot room.

But to your comment about AC, no. Sorry again, but there are many many computer users (probably more than not in the world) who do NOT have AC. Many live in climates that don't have enough hot days to justify it, and many just cannot afford it.

Yes, but not if it is just blowing hot air on the fins. Again, you cannot cool anything with a fan to a lower temperature than the air the fan is blowing. If the room temperature is up 10°, you can expect your CPU temps to rise similarly. And that is the point I was trying to make.

As for TIM, I am going on 40+ years of experience and tons of personal research and I have yet to come across a TIM maker that says their TIM needs periodic replacing. That would NOT be a good selling point, in my book. Note in these Arctic Silver Application Instructions, page 2 it says (my bold added),

Your CPU temps are well above your room temps.

And if you think your HD temps are correct (assuming they have been spinning for awhile), I have some excellent swamp land in Florida you might be interested in. Sensors used in computer components cost about 10 cents...for 10! And are notorious for not being precise. They tend to be consistent, which is typically more important than absolute accuracy. By consistent, I mean if off by 2° at 50°, they tend to always be off by 2° at 50°.

Also, note that drives get their cooling from the cool exterior (room) air being drawn across them through vents in the front of (most) cases. CPUs are blasted with the warmed case air.

 

Absolutely.

Test? Well, no, but I know for a fact, none of MY computers have been dropped, kicked, or mishandled. And I bet for almost everyone who has built their own systems, mounted their own CPUs, and have been in possession of their computers ever since know if they have been dropped, kicked or mishandled - or subject to heavy footfalls and other vibration sources. Don't you know if your computer was dropped? I mean, you are right, there's no test, but just because there's no test, that does not mean it should automatically be replaced if you have no reason to suspect it was dropped, kicked or mishandled in the first place.

So, if you are working on someone else's computer, or if your computer was just moved by someone else, or subject to some other physical abuse, (or you have a mischievous kid who is suddenly being very quiet), then you might assume the TIM needs replacing. But if you built your computer (or since new, have remounted your HSF), then it seems to me you should know if the bond might be broken. Note that fan and drive motor vibrations are not enough. The same TIM is used in other very heavy duty electronic applications, subject to much more abuse than found sitting next to a desk at home. And they hold up just fine, not needing regular replacement.

I'm afraid I am not following. ANY TIME you loosen the clamps once the TIM has been through a heat/cool (curing) cycle, you MUST start over by cleaning the mating surfaces and re-apply a "fresh" new layer of TIM. You never reuse old (cured) TIM.

Small bond releases? That's where you lost me. I have never heard of that and my friend Bing Google isn't helping. Got a link?

It is common, if a little too much TIM was applied, for some TIM to squish out the sides during the first few hours of use. It must be remembered that almost all TIM requires some time to cure, and temperatures can improve as much as 5° as it cures. But curing (break-in period) times vary by TIM formulas (as seen in my Arctic Silver link above) from a few hours, up to 300 hours! That's almost 2 weeks! During that time, some excess TIM may work its way out. If more squishes out after that, there was WAY too much applied in the first place.

And those HSF assembly clamps exert a lot of force, and take a lot of force to set and remove for a good reason - to hold the heatsink firmly to the CPU. They are designed to compensate for the fan and motor vibrations, and picking up the computer to take outside for cleaning, moving to add RAM, etc.

And for the record, if you are using a large, aftermarket cooler, you should always unmount it before shipping (even in the backseat of the car). The weight hanging (potentially bouncing - thanks to unexpected pot-holes), through the socket can put excessive stresses on the motherboard. If you don't remove, make sure the computer lays on it's side with the motherboard facing up and the CPU and heavy heatsink "sitting" on top.

No, sorry. Yes, that is a good monitor, but as I noted above, the motherboard (and HD) sensors are cheap - very cheap, low-tech devices. ANY software monitor you run on your system will be using the exact same, el cheapo-supreme, low-tech devices. And for that matter, most use the exact same Windows api to get that information as used when you type msinfo32 and hit return from the Start orbs Search box. They just arrange using their own GUI design. If you want to accurately measure temps, you need an accurate sensor. And again, accuracy is not their strong point.

No it doesn't. They are not heat storage devices, they do not retain heat. In fact, heat sinks are designed to be one-way heat transmissions devices, with heat moving towards the fins, away from the heat sensitive device.

So you suggesting, "if" the heat isn't released has no bearing here. If that chunk of metal does not release (radiate) heat, then it is not a heat sink - certainly not one properly designed.

Even if your computer (and hard drives) was off all night, there is no way it can be cooler than the air it sat in. Right? This assumes the corner where your computer sits is the same temp as where your little weather station sits. And this assumes the computer (and drives) had time to fully (inside and out) "acclimate" to the "current" room temperature - and that could take a couple hours for the dense matter of the drives to heat, if the sun suddenly heated the air in the room.

I mean if you truly believe a fan can blow 82° air onto a chunk of metal containing running motors (a hard drive) and lower the temperature of that running drive to 77°, then I don't know what more I can say.