Help my PC is dying of heat exhaustion

A

Andy

Desmond said:
It is so hot here this summer that my computer crashes when it gets hot. I
was looking into water cooled pc cases. Is it safe to buy a kit. Also the
prices vary from £199 to £1,400.

http://www.watercoolinguk.co.uk/p/W...2-240-Watercooled-Case-Black--Blue_35292.html

I am not sure about the options here. Do I NEED to pick one or can I just
transfer mother board and other things into the case. Any help would be
great.

Desmond.
Click to expand...

Hi Desmond,

Based on the cost of water cooling and the rarity of this heat wave, my
advice would be to remove your case side panel.

Andy
 
P

Paul

Desmond said:
It is so hot here this summer that my computer crashes when it gets hot.
I was looking into water cooled pc cases. Is it safe to buy a kit. Also the
prices vary from £199 to £1,400.

http://www.watercoolinguk.co.uk/p/W...2-240-Watercooled-Case-Black--Blue_35292.html

I am not sure about the options here. Do I NEED to pick one or can I just
transfer mother board and other things into the case. Any help would be great.

Desmond.
Click to expand...

That particular product, looks to be a combination of two products.

A CM-690 Version II, is a computer case. Coolermaster makes a number
of minor variations.

http://images17.newegg.com/is/image/newegg/11-119-259-Z05?$S640W$

The "240" in the title, could be referring to one of these kits.
It has a water block for the CPU, and a radiator that fits near the
top of the computer case. The radiator has exhaust fans, to move
the heat out into the room.

"COOLER MASTER Eisberg 240L Prestige"
http://www.newegg.com/Product/Product.aspx?Item=N82E16835103184

Your web site doesn't go into enough detail, for me to verify those
are the box contents.

*******

I wouldn't buy something like that, without first understanding
why the current cooling solution is not working.

Air cooling, is a two stage process. CPU to case air, is handled
by the CPU cooler and CPU fan assembly. That transfers the CPU heat,
into the case air, and makes a "warm cloud" around the CPU.

To move the warm cloud, you need an air intake vent (near the front),
and an exhaust fan (near the back). That causes a constant flow of
air through the case, and moves the "warm cloud" out of the computer.

You can trade off the two cooling subsystems. You could use a loud
back fan, and a quiet CPU fan. Or a loud CPU fan and a quiet back fan.
In other words, both parts have their own delta_T to contribute to
the final CPU temperature. If you're using the finest cooler for
the CPU, that money can buy, and the CPU still isn't cool enough,
it means the exhaust fan on the back is not enough.

A "well cooled case", has a 7C to 10C temp rise above room temp.
Say the room is 25C. Then if your motherboard has a surface mount
thermistor, it should register 35C or less. That tells you the
exhaust fan on the back is doing its job. On my previous computer
case, I had to remove some plastic trim on the front of the case,
to get sufficient air intake cross-section, to support the volume
the exhaust fan was moving. (I was getting the vacuum cleaner sound,
when the side panel was placed on the case. More vent in the front,
eased the vacuum cleaner effect.)

The CPU cooler is rated by "theta_R". (Only Zalman will tell us the
value, and most manufacturers won't admit the value, as it makes
shopping for the best one, too easy.)

Let's work a typical example, using our 25C room, and 35C internal
case air temperature.

The cooler that comes with the CPU, has a theta_R of 0.33C/W.
Say the processor is 100W, just to make the math simple.
0.33C/W * 100W = 33C. 35C case air, plus a 33C delta_T to
the CPU inside, gives a CPU lid running at 68C. This is a bit
on the warm side.

If I switch over to an enthusiast air cooler, perhaps it has
a theta_R of 0.15C/W. Now the delta_T is 15C, and the CPU is
35C+15C = 50C. This is below 65C (for certain AMD processors),
and I'm happy.

Notice the two contributions. The rear fan made things 10C warmer.
The CPU cooler made things 15C warmer. There isn't much room left
to improve either of them now. I would need a very loud rear
fan to get below 10C delta. And the CPU cooler, it's hard to
find something below 0.15C/W. Maybe I can find a 0.11 or 0.12 one.
Some of the coolers are so large, they prevent all the DIMMs from
being installed, so be careful what you wish for.

Based on these examples, measure your CPU temperature, internal
case air temperature (use a glass thermometer), and room air
temperature. Compute your delta_T values, and see if something
isn't obviously out of whack.

When you fit the CPU cooler:

1) Check that it is fully seated. I had one cooler that sat
at an angle, because the heatsink was the wrong shape. It took
a solid half hour of filing in my bench vice, to fix it. As
another example of an oddity, on Intel Core2, the top of the
processor is convex, while the CPU cooler is flat. No, I don't
recommend lapping it. The convex shape is the result of the
soldering operation they use, to affix the top.

2) If using thermal paste, make sure the gap between CPU cooler and
CPU, has just a tiny bit of paste oozing out of the gap. That helps
prove you used enough. If you didn't use any thermal paste, and
there was no other thermal interface material (TIM) present, then
that is your problem right there. The CPU is going to throttle,
if there isn't some paste present.

3) Make sure the heatsink fastener is secure. On AMD, a certain brand
of cooler is known to snap off the plastic tab on the side of the
processor socket, leaving the cooler dangling loose, and causing
an overheat. If the fastener pulls on the plastic socket, inspect
it regularly to make sure it is still holding :)

If you're overclocking, I can see the attraction of a water cooling
system. If you're running stock clocks, then you should be able to
scrape by with air cooling. Some overclocking efforts, dissipate
200W at the processor, which is a bit much for a cheap air cooler.

To work out the required CFM of the rear fan, the cabinet cooling equation...

CFM = 3.16 * Watts / Delta_T_degrees_F

Here, watts would be the total watts of video card plus CPU plus
hard drives and so on. That could be a couple hundred watts. If the
video card vents out the back, a fair portion of the power is not
to be counted. If you do that math, and get 35 CFM, that's an
"ordinary" fan and not obnoxious. If the math works out to 100CFM,
you'd either need a very large fan (a case with a 220mm fan),
or be prepared for a deafening roar. I have a fan that will do
110CFM or so, and I have to keep it turned down. I'd have to leave
the room, if I fed the fan the full 12V. The fan in that case, is
37.5mm thick and thicker than a standard fan. And I think it draws
around 1 ampere of current (so it's not a good idea to run it
off a motherboard header). That's an example of an extreme. And
you don't really want to do that. A number of Dells use fans like
that, but normally the fan control is turned down on those as well.
Only under a failure, does the Dell fan start to wail.

But at least there is a cabinet cooling equation. If the computer case
air seems really warm, and you have a good idea of the total DC power
being used in the box, then you have the ingredients to figure
out the fan CFM (cubic feet per minute) needed. I'm not skilled
enough in the maths of CFMs, to tell you what happens when
there are seven fans, some pushing and some pulling :) That
is why I like to start with a simplistic cooling system, so
I can easily relate to the cabinet equation.

HTH,
Paul
 
F

Flasherly

It is so hot here this summer
Click to expand...
that my computer crashes when it gets hot. I was looking into water
cooled pc cases. Is it safe to buy a kit. Also the prices vary from
£199 to £1,400.
I am not sure about the options here.
Click to expand...
Do I NEED to pick one or can I just transfer mother board and other
things into the case. Any help would be great.
Desmond.
Click to expand...

I tend favor Intel over AMD for heat, although ideally I don't like
over 115F. Which is funny, because I'm getting around 130F on a 4000
or thereabouts AMD X2 core, and 105F from a somewhat less endowed
Pentium D 2.6GHz dual core. Temps achieved loaded and tortured, both
with six-legged water pipes, properly so called for condensation
combined with "wicking" action. Both cases are reasonably large,
accommodating, as are the heatsinks, and overall well enough aerated
for budget purposes. Money alone never was a requisite or intent to a
moderately designed computer, by far acceptable in all ways, I should
say.
 
D

Desmond

It is so hot here this summer that my computer crashes when it gets hot. I was looking into water cooled pc cases. Is it safe to buy a kit. Also theprices vary from £199 to £1,400.



http://www.watercoolinguk.co.uk/p/W...2-240-Watercooled-Case-Black--Blue_35292.html



I am not sure about the options here. Do I NEED to pick one or can I justtransfer mother board and other things into the case. Any help would be great.



Desmond.
Click to expand...

It is nearly 30C in my room and I believe that the room temperature is to high. Yes I have got a good fan on my AMD 4200+ dual core processor. but if air is blown out the back of the case only to be replaced with warm air coming in I can't see a way around it.I have taken the fan off and blown the crap out that gets into it. Currently I have taken the side of the case and I have a desktop fan blowing into it. It seems to be OK that way.
I have a tall tower system 14U (24.5 inch) high. I have a fan at the top but the lead is not long enough to reach the connector on the main board. I do agree that better circulation is needed.

These water cooled cases are not cheep but I have heard that the temperature can be below 15C and you can even remove CPU fan but that sounds a bit dodgy to me.
 
P

Paul

Desmond said:
It is nearly 30C in my room and I believe that the room temperature is to high. Yes I have got a good fan on my AMD 4200+ dual core processor. but if air is blown out the back of the case only to be replaced with warm air coming in I can't see a way around it.I have taken the fan off and blown the crap out that gets into it. Currently I have taken the side of the case and I have a desktop fan blowing into it. It seems to be OK that way.
I have a tall tower system 14U (24.5 inch) high. I have a fan at the top but the lead is not long enough to reach the connector on the main board. I do agree that better circulation is needed.

These water cooled cases are not cheep but I have heard that the temperature can be below 15C and you can even remove CPU fan but that sounds a bit dodgy to me.
Click to expand...

"temperature can be below 15C"

The water cooler cannot go below ambient. It isn't a swamp
cooler, and does not rely on evaporation for cooling. It
just relies on ordinary heat transfer, into room ambient.

Those readings come from inaccurate temperature sensors.
The Intel built-in sensor for example, is accurate at high
temperatures (like one degree below the throttle point).
But the Intel sensor is pretty bad, at the low end. You
could probably manage a 15C reading on Intel, if the actual
temperature was 25C.

A water cooler still has to have a delta_T, as the thermal
resistance is not zero, and the power dissipated on the
CPU is finite. So the CPU should end up at a higher
temperature than ambient. To get some idea of the
performance, you want to run Prime95 or CPUBurn or the
like, dissipate max power, then check to see what
kind of delta_T results. And give time for any liquid
reservoir in the water cooling loop, to reach dynamic
equilibrium. That could take ten minutes or more.

If you wanted to try a water cooler, you don't need to use
a particular PC case. As long as you could route the two
hoses outside the computer case, you could mount the radiator
outside the case, rather than inside it. I've seen at
least one computer case, that had large grommet holes, for
passing a couple water cooling hoses out of the case. (The grommet
material, is to prevent the hoses from getting cut on the case
metal.)

You could also use a removed PCI slot cover for that purpose. Lots
of possibilities for roll your own solutions. But I'd prefer
to see some thermal performance numbers first, executed by
someone who knows how to test. Just to see if it's worth
$180 or $100 or whatever, for a kit.

Paul
 
J

John Doe

If you use compressed air, you don't need to take the fan off of
the CPU. Just blow around it in various ways (keeping the can
upright). I don't bother to take it outside, I just turn on
filtering devices like my vacuum cleaner and my filter fan with
them right next to the case.

Good luck and have fun.
 
F

Flasherly

It is nearly 30C in my room and I believe
Click to expand...
that the room temperature is to high.
-

85F degrees is not too hot, nor is that explicit reason for a computer
to crash if the manufacturer rates the chip as safe somewhere around
twice that temperature. Needless to mention, earlier AMD chips, to
me, seemed more fault prone over 115F, although can't say I've noted
that aspect on this newer dual core AMD X2 4000 (paid $10US for the
chip off Ebay, actually). Processing video along with an added layer
of programming for simultaneous audio evinces tempertures consistantly
between 120F to 130F. The case is set to its side and laid
horizontally with a front panel fan. The other fan, I personally made
to also attach to a PS lead, is affixed to the upward-side's drilled
holes for venting. Since fans are the first failure points to PCs, I'd
as soon rather see them attached to a robust and replaceable PS, than
a MB header and what lesser voltage tolerance/supplies a MB is capable
of regulating;- OTH, a heatsink fan may be built better, and I do have
them, regardless, plugged into MB headers -- it's only until perhaps
some other time, when to splice them off into PS adaptor plugs. This
P4 dualcore is running 4 degree F higher than ambient room
temperature, or 24 degrees lower than my hottest hard drive, which is
over 10 years old. A brutish six-legged CPU cooler, stretching pretty
much the case width, cools it, at one generation below the newest,
heartier coolers of the same dimensions capable of handling 6 cores;-
some of the latter indeed do approach water-cooling costs, but my days
of needlessly chasing technology at higher costs are contentedly over,
as I've sufficient means on the middle ground to occupy.
 
M

Mike Tomlinson

En el artículo <[email protected]>,
Desmond said:
Currently I have taken the side of the case and I have a desktop
fan blowing into it. It seems to be OK that way.
Click to expand...

I's suggest leaving it like that until the heatwave is over. it won't
be long and isn't worth the cost of a water cooling kit.
 
P

Paul

Desmond said:
It is nearly 30C in my room and I believe that the room temperature is to high. Yes I have got a good fan on my AMD 4200+ dual core processor. but if air is blown out the back of the case only to be replaced with warm air coming in I can't see a way around it.I have taken the fan off and blown the crap out that gets into it. Currently I have taken the side of the case and I have a desktop fan blowing into it. It seems to be OK that way.
I have a tall tower system 14U (24.5 inch) high. I have a fan at the top but the lead is not long enough to reach the connector on the main board. I do agree that better circulation is needed.

These water cooled cases are not cheep but I have heard that the temperature can be below 15C and you can even remove CPU fan but that sounds a bit dodgy to me.
Click to expand...

One other editorial comment.

Check your hard drives. They register temperature via SMART.

You can use the free version of this, and the Health tab,
to display SMART for each drive.

There are a few complete data sheets for hard drives, that
provide a temperature/humidity graph, which shows allowed
operating regime. The more humid it gets, the lower the
allowed operating temperature. While no explanation is
provided, I feel this is related to the drive not being
hermetically sealed, and moist air entering the HDA. At
high humidity (60%), the allowed temperature might be 35C.
If the air was quite dry (35% RH), you might be allowed
50C or a bit more. The only time I can hit 35% here, is
using central air, and by the time I get to 35%, the air
temperature is back down to 21C anyway. So if you met
the graph completely for the drives, it more or less requires
air conditioning.

Now, in the real world, hard drive endure more than that.
But all I can tell you is, I've had one failure of a hard
drive, when the house was uncontrolled for a month in summer,
and the RH hit 60% (carpets start to mildew at 60%). I can't
say whether that is a coincidence or not. Haven't had a failure
since. (And now have a brand new 2 ton AC installed for my small
house.) Due to the pricing structure of electricity here, I
tend to run the AC in the evening, meaning the room environment
here swings wildly during the day. I can't quite reach your
30C temps though. At 6PM I'm still a bit below that.

One of the reasons I have the intake vents in front of my
hard drives open, is to get lots of fresh air over them.
And that's why the delta_T on the drives is so small in
my system. Only a few degrees above ambient, typically.
Right now, both drives register 27C. My RH is 50%, so I'm
in the safe zone on the graph. Current room temperature
is 25C, so my delta on the hard drives is 2C. The temperatures
though, are measured with different sensors, so we don't
really know what the delta is exactly. And at least one
specific model of drive (from years ago), the sensor value
is "stuck" and is bogus. When SMART came out, someone thought
it would be fun to pretend they had a temperature sensor, when
none was provided. Now, temperature sensing is quite common.

The worst computer cases, are the ones that hide the hard
drives in a "dead spot", and no cooling air is available.
That's when I do mods, to fix it.

Paul
 
M

Michael Black

Well, there's your problem right there. Get a cheap air conditioner. Less
hassle than rebuilding your PC.
Click to expand...
One time someone I knew, her website disappeared. The next time I saw her
I asked, and she said "the server melted". It was a computer somewhere in
a closet, and either the door was closed when it shouldn't have, or a fan
died, or something, and it just expired from the heat.

Michael
 
F

Flasherly

Now, in the real world, hard drive endure more than that.
But all I can tell you is, I've had one failure of a hard
drive, when the house was uncontrolled for a month in summer,
and the RH hit 60% (carpets start to mildew at 60%). I can't
say whether that is a coincidence or not.
Click to expand...

Coincidence. I'm running much higher humidity, although in cycles
according to a fair degree of leeway permitted summer AC usage. With
drives 5 to 10 years old, that would still be well under manufactured
stipulations for higher heat tolerance. But, there's really not much
reason to go by to account industry failure, so as a reflection within
endusers ratings, experience and expectations;- That they all do fail
reflects as much a wide regard for disgust in underscoring an
inconvenient loss of storage and its repercussions. I've bought
severly discounted Western Digital HDs from BestBuy event sales that
failed as if set to clockwork, to the day, when their warantee
expired. To say that with some bitterness when regarding Western
Digital, is to say Western Digital will be the first brand I'd as soon
personally skip over;- furthermore to discount such admittance for one
especially comforting, in being what irony has to condition by
imposition upon reasoning as a result of the subjective experience.
Oh, and that's entirely above what filth underlies carpet pads that
don't get regularly vacuumed, otherwise, for perfectly serving intent
beyond superficially little else above to note;. . .berber, shag,
industrial fibers, I've haven't the foggiest clue, beyond it's plain
and brown, and that nothing has sprouted or nefariously inhabits its
recesses to climb out for some induibitably ill-conceived purpose.
 
P

Paul

Flasherly said:
Coincidence. I'm running much higher humidity, although in cycles
according to a fair degree of leeway permitted summer AC usage. With
drives 5 to 10 years old, that would still be well under manufactured
stipulations for higher heat tolerance. But, there's really not much
reason to go by to account industry failure, so as a reflection within
endusers ratings, experience and expectations;- That they all do fail
reflects as much a wide regard for disgust in underscoring an
inconvenient loss of storage and its repercussions. I've bought
severly discounted Western Digital HDs from BestBuy event sales that
failed as if set to clockwork, to the day, when their warantee
expired. To say that with some bitterness when regarding Western
Digital, is to say Western Digital will be the first brand I'd as soon
personally skip over;- furthermore to discount such admittance for one
especially comforting, in being what irony has to condition by
imposition upon reasoning as a result of the subjective experience.
Oh, and that's entirely above what filth underlies carpet pads that
don't get regularly vacuumed, otherwise, for perfectly serving intent
beyond superficially little else above to note;. . .berber, shag,
industrial fibers, I've haven't the foggiest clue, beyond it's plain
and brown, and that nothing has sprouted or nefariously inhabits its
recesses to climb out for some induibitably ill-conceived purpose.
Click to expand...

This is the curve for a 1TB hard drive from Hitachi.

I added the yellow coloration, to show the operating region.
The outer curve is for storage conditions for the drive.
The yellow section, is for when the drive is powered.

http://imageshack.us/a/img268/8350/o87.gif

I think I have another drive spec, which has a slightly worse
curve than that one.

And it's interesting, that the last two drives I bought, they
actually came with a bag of silica gel, inside the antistatic
bag with the hard drive.

Paul
 
M

Mike Tomlinson

Paul <[email protected]> said:
And it's interesting, that the last two drives I bought, they
actually came with a bag of silica gel, inside the antistatic
bag with the hard drive.
Click to expand...

What's interesting about that? It's just to absorb any moisture which
was trapped in the bag when it was sealed up with the drive inside.

You're over-analysing things as usual, Paul.
 
M

Michael Black

What's interesting about that? It's just to absorb any moisture which
was trapped in the bag when it was sealed up with the drive inside.

You're over-analysing things as usual, Paul.
Click to expand...
Those bags have become so common, I don't think we even notice them most
of the time. If I open up a radio box and find a pack of that silica gel,
then why wouldn't I find one in with a hard drive? The hard drive probably
costs more than the radio.

Michael
 
F

Flasherly

This is the curve for a 1TB hard drive from Hitachi.

I added the yellow coloration, to show the operating region.
The outer curve is for storage conditions for the drive.
The yellow section, is for when the drive is powered.

http://imageshack.us/a/img268/8350/o87.gif

I think I have another drive spec, which has a slightly worse
curve than that one.

And it's interesting, that the last two drives I bought, they
actually came with a bag of silica gel, inside the antistatic
bag with the hard drive.

Paul
Click to expand...

That's 106-146F at the halfway mark, at an even split for degenerative
humidity as heat increases to its spec'd limit. Hitachi - OK, that's
what happened with IBM when they got tired of holding that ball;- I've
never had so many drives manufacturered from such a wide diversity of
countries, as when I owned IBM drives (can't even remember correctly,
but suspect they might have been under 100Meg/byte class drives). No
doubt some failures in there, too, as they vaguely come back into
recollection;- Hitachi, though, is coming up a semi-blank. ...Possibly
I was into Western Digital when they had contracts with the US Navy,
and Seagate just past that and ever since, except for a "few"
monsterous Samsungs (1-2T/byte drives) from when the floor fell
through prices prior to a hurricane a year or two ago.

Those specs are nasty. I'd really expect better, but it's hard to
argue with the black&white. You could pray for my drives, if vaguely
close to those specs, as they're living a life in hell;- in fact,
don't even bother - I've already got my money out of most the
steady/ready Eddies that have sufficed for system builds, the oldest
ones. :)
 
S

Skybuck Flying

You are not alone.

My DreamPC for 2006 and fixed throughout the years also cannot handle the
heat.

Not even with an Antec 1200 case, somewhat cleaned.

At 27 celcius degrees, the CPU gets so hot that the motherboard's bios shuts
down the PC.

The temperature shutdown seems to be at 50 according to motherboard
settings.

So there I will be... playing Company of Heroes and all of a sudden...
booom... it shuts off.

Well at least this will prevent frying damage or so... so I kinda like it as
long as it doesnt happen well I am busy with something important ;)

So far it has not happened with anything important just gaming.

The solution for me for now is to keep the doors and windows open and allow
some cool breeze to come in... this will drop the temperature down to 26
degrees celcius.

So my computer crashing or not crashing depands on 1 or 2 degrees.

Apperently the AMD X2 3800+ CPU is rated at about 85 watts or so...
apperently that's way too hot.

I am not happy with how CPU's are marketted today...

They all have the same name and same model number... sometimes there will be
a letter behind the cpu's model number for example for latest intel haswell
processor a T.

Apperently the T versions run at reduced clock rates, which makes them
consume less energy... instead of 85 watts it will be 65 watts or 45 or 35
watts or something.

So I was thinking... maybe it's time to ditch the AMD x2+ 3800+ crap cpu...
and switch to something else...

But this could be an expensive joke:

Probably new power supply needed, new motherboard, new memory... maybe even
new graphics card or maybe not.

What's further annoying is the cheap shit that's between the wafer and the
cpu heatspread apperently this could be replaced and make it drop some
degrees.

Not sure what happens if my finger would touch a waver if that would be the
end of it. Does seem interesting but at 35 watts or 45 watts doing that
might not be necessary.

For now I will see how it goes however... there are very hot days ahead...
so keeping the doors open might not be possible.

Right now I am thinking... maybe there still is a socket 939 processor
somewhere.... with much lower heat... that could be nice... but then me
would be a little bit worried about the performance ;) Maybe even a single
core.

Bye,
Skybuck.
 
P

Paul

Mike said:
What's interesting about that? It's just to absorb any moisture which
was trapped in the bag when it was sealed up with the drive inside.

You're over-analysing things as usual, Paul.
Click to expand...

They were the first two hard drives I've ever seen with
silica gel. They didn't waste money on that in the past.
The drives in question, were a 2TB and a 3TB.

Paul
 
M

Mike Tomlinson

En el artículo <[email protected]>, Paul <[email protected]>
escribió:

They didn't waste money on that in the past.
Click to expand...

It isn't a waste of money. It's to prevent moisture damage to the drive
in transit - if the drive is put somewhere cold, any moisture inside the
sealed bag will condense out onto the platters, which is Bad News.

The reason you haven't seen the silica gel packs previously is because
you've bought drives which have been bulk shipped in large containers
with large silica gel bags. These drives were then repacked by the
seller in individual nags for retail sale, by which time there is no
longer a need for the silica gel.
 
Top