50C+ at full load for [1] minute? HELP PLEASE

[Pedro Sanchez]

Here is some info on my comp before I list some temps
Windows XP Home Edition [SP2] 5.1.2600
CPU Type AMD Athlon 64 3500+
MB Asus A8S-X (3 PCI, 2 PCI-E x1, 1 PCI-E x16, 4 DDR DIMM)
MB Chipset SiS 756, AMD Hammer
RAM 1024 MB (PC3200 DDR SDRAM)
Video Card NVIDIA GeForce 6800 GS(256 MB)
Monitor Proview 998/998A/998N(1239224727)
Disk Drive Maxtor 6L080P0 80 GB
Optical Drive NEC DVD_RW ND-3550A

Now for my question/problems

Before Arctic Silver, IDLE for 20 minutes C.P.U. 40C
After Arctic Silver, IDLE 20 Minutes, CPU 35C

Full Load Before Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 55C

Full Load After Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 49C

So, I went from 40C idle to 35C idle. And From 55C to 49C under full
load.

Now it's crazy temp time.

Boot Temp [Only startups loaded] 35C

OVERCLOCKED 30% IDLE FOR 5 MIN 43C
0% NO OVERCLOCKING IDLE FOR 5 MINUTES 42C

0% OC'D 51C STARCRAFT FOR [1 MIN]
30% OC'D 52C STARCRAFT FOR [1 MIN]

I also bought a digital thermometer to see the ambient temp in this
room. It is from a low of 79[F] to a high of 86[F]

So, did I do the arctic silver wrong?

Everything was fine until 2 weeks ago.

Please someone throw me a bone.
Please help ?

thanks,

 

[Ed Medlin]

Here is some info on my comp before I list some temps
Windows XP Home Edition [SP2] 5.1.2600
CPU Type AMD Athlon 64 3500+
MB Asus A8S-X (3 PCI, 2 PCI-E x1, 1 PCI-E x16, 4 DDR DIMM)
MB Chipset SiS 756, AMD Hammer
RAM 1024 MB (PC3200 DDR SDRAM)
Video Card NVIDIA GeForce 6800 GS(256 MB)
Monitor Proview 998/998A/998N(1239224727)
Disk Drive Maxtor 6L080P0 80 GB
Optical Drive NEC DVD_RW ND-3550A

Now for my question/problems

Before Arctic Silver, IDLE for 20 minutes C.P.U. 40C
After Arctic Silver, IDLE 20 Minutes, CPU 35C

Full Load Before Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 55C

Full Load After Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 49C

So, I went from 40C idle to 35C idle. And From 55C to 49C under full
load.

Now it's crazy temp time.

Boot Temp [Only startups loaded] 35C

OVERCLOCKED 30% IDLE FOR 5 MIN 43C
0% NO OVERCLOCKING IDLE FOR 5 MINUTES 42C

0% OC'D 51C STARCRAFT FOR [1 MIN]
30% OC'D 52C STARCRAFT FOR [1 MIN]

I also bought a digital thermometer to see the ambient temp in this
room. It is from a low of 79[F] to a high of 86[F]

So, did I do the arctic silver wrong?

Everything was fine until 2 weeks ago.

Please someone throw me a bone.
Please help ?

thanks,

Your temps are exactly what would be expected at the ambient temps of your
room. The temps are well within the AMD specs and should not be a problem.
Everything was probably fine earlier because the ambient temps have probably
gone higher. One other thing is that when you apply AS, do so with a very
thin layer. If you use too much it could act as a insulator and bring temps
up. What are your temps if you keep running under full load? High 50sC - low
60sC (under load) would not be unusual if your ambient temps are in the high
80sF. The CPU temps will be directly proportional to the ambient temps in
the room.

Ed

 

[Pedro Sanchez]

Here is some info on my comp before I list some temps
Windows XP Home Edition [SP2] 5.1.2600
CPU Type AMD Athlon 64 3500+
MB Asus A8S-X (3 PCI, 2 PCI-E x1, 1 PCI-E x16, 4 DDR DIMM)
MB Chipset SiS 756, AMD Hammer
RAM 1024 MB (PC3200 DDR SDRAM)
Video Card NVIDIA GeForce 6800 GS(256 MB)
Monitor Proview 998/998A/998N(1239224727)
Disk Drive Maxtor 6L080P0 80 GB
Optical Drive NEC DVD_RW ND-3550A

Now for my question/problems

Before Arctic Silver, IDLE for 20 minutes C.P.U. 40C
After Arctic Silver, IDLE 20 Minutes, CPU 35C

Full Load Before Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 55C

Full Load After Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 49C

So, I went from 40C idle to 35C idle. And From 55C to 49C under full
load.

Now it's crazy temp time.

Boot Temp [Only startups loaded] 35C

OVERCLOCKED 30% IDLE FOR 5 MIN 43C
0% NO OVERCLOCKING IDLE FOR 5 MINUTES 42C

0% OC'D 51C STARCRAFT FOR [1 MIN]
30% OC'D 52C STARCRAFT FOR [1 MIN]

I also bought a digital thermometer to see the ambient temp in this
room. It is from a low of 79[F] to a high of 86[F]

So, did I do the arctic silver wrong?

Everything was fine until 2 weeks ago.

Please someone throw me a bone.
Please help ?

thanks,

Your temps are exactly what would be expected at the ambient temps of your
room. The temps are well within the AMD specs and should not be a problem.
Everything was probably fine earlier because the ambient temps have probably
gone higher. One other thing is that when you apply AS, do so with a very
thin layer. If you use too much it could act as a insulator and bring temps
up. What are your temps if you keep running under full load? High 50sC - low
60sC (under load) would not be unusual if your ambient temps are in the high
80sF. The CPU temps will be directly proportional to the ambient temps in
the room.

ED, I've seen the comp temp go as high as 56C playing starcraft which
is an old game.

I do recall the first few days after I used AC the temp for idle was
30-32C, It's never below 42C now.

And yes, the ambient has increased by about 15F since using AC.

What is a "Danger Zone" Temp for my CPU?

thanks,

 

[kony]

Here is some info on my comp before I list some temps
Windows XP Home Edition [SP2] 5.1.2600
CPU Type AMD Athlon 64 3500+
MB Asus A8S-X (3 PCI, 2 PCI-E x1, 1 PCI-E x16, 4 DDR DIMM)
MB Chipset SiS 756, AMD Hammer
RAM 1024 MB (PC3200 DDR SDRAM)
Video Card NVIDIA GeForce 6800 GS(256 MB)
Monitor Proview 998/998A/998N(1239224727)
Disk Drive Maxtor 6L080P0 80 GB
Optical Drive NEC DVD_RW ND-3550A

Now for my question/problems

Before Arctic Silver, IDLE for 20 minutes C.P.U. 40C
After Arctic Silver, IDLE 20 Minutes, CPU 35C

Full Load Before Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 55C

Full Load After Arctic Silver was used.
Starcraft with Mozilla in Background, C.P.U. 49C

So, I went from 40C idle to 35C idle. And From 55C to 49C under full
load.

Now it's crazy temp time.

Boot Temp [Only startups loaded] 35C

OVERCLOCKED 30% IDLE FOR 5 MIN 43C
0% NO OVERCLOCKING IDLE FOR 5 MINUTES 42C

0% OC'D 51C STARCRAFT FOR [1 MIN]
30% OC'D 52C STARCRAFT FOR [1 MIN]

I also bought a digital thermometer to see the ambient temp in this
room. It is from a low of 79[F] to a high of 86[F]

So, did I do the arctic silver wrong?

Everything was fine until 2 weeks ago.

Please someone throw me a bone.
Please help ?

thanks,

Your temps are exactly what would be expected at the ambient temps of your
room. The temps are well within the AMD specs and should not be a problem.
Everything was probably fine earlier because the ambient temps have probably
gone higher. One other thing is that when you apply AS, do so with a very
thin layer. If you use too much it could act as a insulator and bring temps
up. What are your temps if you keep running under full load? High 50sC - low
60sC (under load) would not be unusual if your ambient temps are in the high
80sF. The CPU temps will be directly proportional to the ambient temps in
the room.

ED, I've seen the comp temp go as high as 56C playing starcraft which
is an old game.

I do recall the first few days after I used AC the temp for idle was
30-32C, It's never below 42C now.

And yes, the ambient has increased by about 15F since using AC.

What is a "Danger Zone" Temp for my CPU?

thanks,

- A difference that large between AS and the prior compound
is most likely due to the prior compound being applied too
thick. Unless the *core* or 'sink is irregular, AS tends to
drop temps about 2C, "maybe" 4C at full load on high heat
(approaching 100W or more) CPUs.

- Starcraft is a somewhat arbitrary load, determination of
fitness needs be made at full load, for an extened period of
time at the highest ambient temp the room encounters
year-round (or at least factor for this higher ambient
additively, if room is 6C lower than expected, simply add
that 6C to the result). Prime95's Torture Test is among
those apps known to come closer to full load.

- Your CPU will have an (in)stability threshold lower than
the "danger zone" that damages the CPU. It is above 60C,
usually above 70C but varies based on specific core (clock
speed relative to the core design's inherant ceiling, and
voltage actually used (supplied by the motherboard and the
stability of the board in supplying it)). As for permanent
damage, your motherboard should have a setting to shutdown
if it exceeds a set threshold, and the threshold is best set
higher than this, minimum of 70C but often 80C is more
appropriate.

 

[SteveH]

Here is some info on my comp before I list some temps
Windows XP Home Edition [SP2] 5.1.2600
CPU Type AMD Athlon 64 3500+
MB Asus A8S-X (3 PCI, 2 PCI-E x1, 1 PCI-E x16, 4 DDR DIMM)
MB Chipset SiS 756, AMD Hammer
RAM 1024 MB (PC3200 DDR SDRAM)
Video Card NVIDIA GeForce 6800 GS(256 MB)
Monitor Proview 998/998A/998N(1239224727)
Disk Drive Maxtor 6L080P0 80 GB
Optical Drive NEC DVD_RW ND-3550A

What is a "Danger Zone" Temp for my CPU?

thanks,

I have the same CPU, isn't the top end temp 65c?
I would look more towards airflow through your case, than worrying about the
effects of AS too much.
I'm currently running prime 95, which is thrashing my CPU at 100% usage. I
can't get the full load temp to exceed 46c. However. I do have a thermaltake
Xaser III case with 5 80mm fans in it.
Another example is my G/f's PC. We just took all the parts out of her old
Globalwin 802 case and put them in an Antec P180 - this has taken 20c off of
the HDD temps and 12C of of the temp of her PIV 3.0Ghz Prescott CPU!

SteveH

 

[Pedro Sanchez]

I have the same CPU, isn't the top end temp 65c?
I would look more towards airflow through your case, than worrying about the
effects of AS too much.
I'm currently running prime 95, which is thrashing my CPU at 100% usage. I
can't get the full load temp to exceed 46c. However. I do have a thermaltake
Xaser III case with 5 80mm fans in it.
Another example is my G/f's PC. We just took all the parts out of her old
Globalwin 802 case and put them in an Antec P180 - this has taken 20c off of
the HDD temps and 12C of of the temp of her PIV 3.0Ghz Prescott CPU!

SteveH

This is the case I am using
http://www.newegg.com/Product/Product.asp?Item=N82E16811154013

Other than the rear fan, I see no place to put another fan.
Any ideas?

Thanks

 

[Pedro Sanchez]

As for permanent
damage, your motherboard should have a setting to shutdown
if it exceeds a set threshold, and the threshold is best set
higher than this, minimum of 70C but often 80C is more
appropriate.

According to AMD's site, my cpu tops out at 65c
And my motherboard has that C.o.p. shut down feature too.
So, if it WILL shutdown before damage is done just don't worry???

thanks

 

[kony]

According to AMD's site, my cpu tops out at 65c
And my motherboard has that C.o.p. shut down feature too.
So, if it WILL shutdown before damage is done just don't worry???

thanks

There is never a need to worry.

Simply make sure your system cooling, including heatsink, is
never allowing the CPU to get anywhere near 65C, or even 60C
for more margin. If it can't do that, change things till
it does.

65C is not a threshold for damage though, only stability,
and has to be borne out with testing in the final
implementation (system) to ensure that stability.

I don't rely on motherboard shutdown features and don't
advise anyone else to do so. Build for long-term
reliability and the odds of ever needing this shutdown
feature are lower than getting killed in a car accident or
some other calamity that makes what your CPU is doing,
trivial.

This includes selecting a good quality (typically a real
major fan manufacturer brand, not PC parts brand) low RPM
fan. The difference in fan lifespan can be an order of
magnitude between the cheap junk and the ones that cost a
mere $3 more.

 

[kony]

This is the case I am using
http://www.newegg.com/Product/Product.asp?Item=N82E16811154013

Other than the rear fan, I see no place to put another fan.
Any ideas?

This is a nice picture, I refer to it with the following
comments;
http://images10.newegg.com/NeweggImage/productimage/11-154-013-12.jpg

The case could be better but isn't so bad. For best results
the rear fan grille (stamped into the case wall) would be
cut out, and the holes along the bottom right edge of the
rear would be covered, as they create a short-loop which
interferes with properly flow.

It does have space in front for a couple of (80mm?) fans,
but they will be mostly noisey and of low effectiveness
unless you also cut out the stamped-in grills on the case
wall. The problem with this is structural integrity, that
after cutting out two holes there your case will be flimsy
in that area and any fan imbalance will cause more noise
than it otherwise would.

Overall, the case has fair bezel intake so any significant
drop in CPU temp will come more from replacing the CPU 'sink
than increasing case flow, unless you had very low RPM fan
installed currently.

 

[Conor]

ED, I've seen the comp temp go as high as 56C playing starcraft which
is an old game.

So what?

I do recall the first few days after I used AC the temp for idle was
30-32C, It's never below 42C now.

And yes, the ambient has increased by about 15F since using AC.

What is a "Danger Zone" Temp for my CPU?

80C.

 

[Don Taylor]

According to AMD's site, my cpu tops out at 65c
And my motherboard has that C.o.p. shut down feature too.
So, if it WILL shutdown before damage is done just don't worry???

On the subject of odd sudden temp changes:

Ancient old Antec case, old ECS motherboard, 2000 Athlon,
no overclocking, Vantec Aeroflow cpu fan, Antec 350 supply.
Been running fine with CPU temp just below 50C for years!

Then suddenly recently, no changes of any kind that I can
locate, the temps jumped 10C+. Room temps didn't change.

I thought dead fan, yanked opened the case, nope all the
fans are still running fine. So I thought dust bunnies,
nope checked and and found a thin layer of dust on things,
blew some of that off, but this was sudden and a big jump.

With the side off the case the temp drops 10 or 15 C,
back to normal or below.

I try blowing air across the Aeroflow thinking maybe that
lint in there somewhere has insulated it a bit. Sometimes
this seems like it gives a jump but more I think it is just
having the side off the case.

I just can't figure this one out. I'm hesitant after all
this time to pry that heat sink off the CPU, half expecting
that if I mess with it I'll break it or make something worse

If I don't button the case back up I'm going to get whacked
because of all the fan whine. But for the life of me I can't
figure out what changed or how to get the temps back down 10C.

After I figure this out and fixed it, I'm thinking of wrapping
the case with Celotex (an insulating sound board) and finding
a way to have that soak some up of the noise but still have
the fans get the heat of of the case.

Any ideas what I can check? (the fins on the Aeroflow are
really tightly packed and the fan blocks most access to this, I
can't think of a good way to really clean that with minimal risk)

thanks

 

[SteveH]

It does have space in front for a couple of (80mm?) fans,
but they will be mostly noisey and of low effectiveness
unless you also cut out the stamped-in grills on the case
wall. The problem with this is structural integrity, that
after cutting out two holes there your case will be flimsy
in that area and any fan imbalance will cause more noise
than it otherwise would.

Overall, the case has fair bezel intake so any significant
drop in CPU temp will come more from replacing the CPU 'sink
than increasing case flow, unless you had very low RPM fan
installed currently.

I think he should at least try getting a couple of 80mm fans into the front,
with out any further mods, just to see if it does make any difference. He
should be able to pick up some quiet fans, although it will mean spending
more than normal on them, as as this is only a cheapo case, is it worth it?
I think it may be, the temps I've quoted in my post above are only with the
stock AMD HSF, so it can only be down to airflow through the case that I get
better temps than the o/p with the same CPU.

SteveH

 

[Blinky the Shark]

This is a nice picture, I refer to it with the following comments;
http://images10.newegg.com/NeweggImage/productimage/11-154-013-12.jpg

The case could be better but isn't so bad. For best results the rear
fan grille (stamped into the case wall) would be cut out, and the
holes along the bottom right edge of the rear would be covered, as
they create a short-loop which interferes with properly flow.

It does have space in front for a couple of (80mm?) fans, but they
will be mostly noisey and of low effectiveness unless you also cut out
the stamped-in grills on the case wall. The problem with this is
structural integrity, that after cutting out two holes there your case
will be flimsy in that area and any fan imbalance will cause more
noise than it otherwise would.

Overall, the case has fair bezel intake so any significant drop in CPU
temp will come more from replacing the CPU 'sink than increasing case
flow, unless you had very low RPM fan installed currently.

New to group, so I don't know its concensus on blow hole fans. That
said, see photo of this system at bottom of page below. Hole was neatly
cut with a metal nibbler. Fan is filtered and blows on CPU/chipset area
of mobo. All other fans blow outward.

http://blinkynet.net/comp/bhard.html

 

[kony]

On the subject of odd sudden temp changes:

Ancient old Antec case, old ECS motherboard, 2000 Athlon,
no overclocking, Vantec Aeroflow cpu fan, Antec 350 supply.
Been running fine with CPU temp just below 50C for years!

Which old ECS board? Some were susceptible to capacitor
failure, might be worth a closer look.

Then suddenly recently, no changes of any kind that I can
locate, the temps jumped 10C+. Room temps didn't change.

Seems odd, might you have a virus or something running in
the background that now has the idle time of the system,
greatly reduced?

I thought dead fan, yanked opened the case, nope all the
fans are still running fine. So I thought dust bunnies,
nope checked and and found a thin layer of dust on things,
blew some of that off, but this was sudden and a big jump.

With the side off the case the temp drops 10 or 15 C,
back to normal or below.

15C is a pretty significant drop IF you mean drop in CPU
temp. Are you sure all the intake and exhaust areas are
free of dust? It tends to build up between the front bezel
and the metal front wall of the case too.

I try blowing air across the Aeroflow thinking maybe that
lint in there somewhere has insulated it a bit. Sometimes
this seems like it gives a jump but more I think it is just
having the side off the case.

I just can't figure this one out. I'm hesitant after all
this time to pry that heat sink off the CPU, half expecting
that if I mess with it I'll break it or make something worse

Maybe, but if that's the problem how would you ever resolve
it otherwise? Take your time, use the right tool, and don't
hesitate to take other parts out first if they're in the
way.

If I don't button the case back up I'm going to get whacked
because of all the fan whine. But for the life of me I can't
figure out what changed or how to get the temps back down 10C.

Which fans, just the CPU heatsink fan? If the system
weren't as old, I might suggest that it was a good
investment to buy a better heatsink with a large low RPM
fan, but today... hard to say what value you place on the
system or how long you'd continue using it.

After I figure this out and fixed it, I'm thinking of wrapping
the case with Celotex (an insulating sound board) and finding
a way to have that soak some up of the noise but still have
the fans get the heat of of the case.

That seems like a bit of overkill unless you just happened
to have the Celotex lying around and had plans to reuse the
case again for the next system build. Better to focus
more on improving air intake and exhaust paths and use
largest lowest RPM fans possible, which will prolong fan
lifespan too.

Any ideas what I can check? (the fins on the Aeroflow are
really tightly packed and the fan blocks most access to this, I
can't think of a good way to really clean that with minimal risk)

Oh, somehow I'd assumed you had cleaned it. Yes it should
be taken off and cleaned out periodically... I can't spec a
period since it varies per environment.

 

[Don Taylor]

In alt.comp.hardware.homebuilt you write:

Which old ECS board? Some were susceptible to capacitor
failure, might be worth a closer look.

K7S5A. Others have reported problems with this but it has
been astonishingly reliable for me. I'd forgotten about the
cap problems so I got out the floodlight and looked really
closely at all of them. None of the electrolytics appear to
be deformed at all, and I've seen pictures of marginal and
bad ones, and built electronics so I suspect I'd notice this
when I look. But that was an excellent suggestion.

Seems odd, might you have a virus or something running in
the background that now has the idle time of the system,
greatly reduced?

DSL modem firewall, Linksys BEFSR41 "firewall", Norton
firewall, Norton AV, Ad-Aware, Spybot, Trend Micro now and
then, (haven't had an infection since 2001 when I tried to
help some poor SOB in the office with his spreadsheet and
he gave me a macrovirus on his floppy, but it wasn't clear
that I even opened that before we all discovered he was
infected) I just scanned everything AGAIN and it appears
clean. But again, an excellent suggestion.

15C is a pretty significant drop IF you mean drop in CPU
temp. Are you sure all the intake and exhaust areas are
free of dust? It tends to build up between the front bezel
and the metal front wall of the case too.

Yes CPU temp. And that is almost how much it jumped up when
this started. I've inspected it pretty well with the flood
light, there are thin layers of accumulated dust, but not
enough to choke off airflow anywhere inside I don't think.

Maybe, but if that's the problem how would you ever resolve
it otherwise? Take your time, use the right tool, and don't
hesitate to take other parts out first if they're in the
way.

It seems like there is
A: pretty conclusively narrowing the problem to one likely
cause then confirming and making changes to fix that
or
B: like the original poster, changed a bunch of stuff, pulled
off the heat sink, put it back on, fooled with thermal compound,
don't know if I screwed something up in the process, don't know
what the original problem was, never will know now because of
all the changes and fiddling AND THEN maybe STILL don't know
the problem is or what to do to fix it.

If I can figure it out before adding upmteen different
variables to the problem then I'll certainly do what is needed
to fix it.

Which fans, just the CPU heatsink fan? If the system
weren't as old, I might suggest that it was a good
investment to buy a better heatsink with a large low RPM
fan, but today... hard to say what value you place on the
system or how long you'd continue using it.

The overall system noise with the side off the case to keep
the temperatures below 50 instead of above 60.

That seems like a bit of overkill unless you just happened
to have the Celotex lying around and had plans to reuse the
case again for the next system build. Better to focus
more on improving air intake and exhaust paths and use
largest lowest RPM fans possible, which will prolong fan
lifespan too.

The fan noise was annoying me before this happened, compared
to a Dell that I bought here and which is as close to silent
as any I could imagine. Now that I've been sitting here with
it whining I thought that as soon as I resolve the heat problem
I'd try to even further reduce the noise. (And sheets of
Celotex are $5.95 at the supply store down the street)

If $6 would cut the noise after the temps were solved then
this box will be good enough for another year or so. And
$6 a year for a little quiet is a bargain in my book. Or I
could spend a few hundred for new "quiet" parts, which may
after a few hours not seem subjectively any quieter, doesn't
seem like nearly the bargain to me. $6 of Celotex can go
in the dumpster if it didn't work and I'd still think it was
worth trying to see if it would work.

Oh, somehow I'd assumed you had cleaned it. Yes it should
be taken off and cleaned out periodically... I can't spec a
period since it varies per environment.

When I opened up the side I blew the dust out of the system.
But the CPU heatsink is sheets of aluminum spaced less than
a mm apart, I've blown air through it but I'm still looking
for anything else I could try before making changes to just
make it harder to figure out what the original cause was.

thanks for any ideas

 

[kony]

The fan noise was annoying me before this happened, compared
to a Dell that I bought here and which is as close to silent
as any I could imagine. Now that I've been sitting here with
it whining I thought that as soon as I resolve the heat problem
I'd try to even further reduce the noise. (And sheets of
Celotex are $5.95 at the supply store down the street)

Not all Dells stay quiet though, I've seen several which had
their rear exhaust fans get noisey after a couple years.
They weren't particularly bad quality fans either, it just
seems the motherboard PWM control for them wasn't as stable
as it could've been.

If $6 would cut the noise after the temps were solved then
this box will be good enough for another year or so. And
$6 a year for a little quiet is a bargain in my book. Or I
could spend a few hundred for new "quiet" parts, which may
after a few hours not seem subjectively any quieter, doesn't
seem like nearly the bargain to me. $6 of Celotex can go
in the dumpster if it didn't work and I'd still think it was
worth trying to see if it would work.

Maybe, I can't see your system. I can't hear my heatsink
fan at all, the only noise escaping the system is from fans
mounted on the case walls with a direct exit- something
Celotex can't fix, but because they are large (and so
possible to use very low RPM, the noise is low.

When I opened up the side I blew the dust out of the system.
But the CPU heatsink is sheets of aluminum spaced less than
a mm apart, I've blown air through it but I'm still looking
for anything else I could try before making changes to just
make it harder to figure out what the original cause was.

thanks for any ideas

Often when I get 'sinks with such tightly spaced fins I take
a piece of sandpaper folded into a V shape and slope the
sharp corners of the 'sink fins some. Result is less dust
getting stuck. The other alternative that works even better
is to just put a filter on the case but that takes more
pre-planning for best results, part of why I always want a
case on hand well before any parts to put into it.

If you used silicone based thermal compound, it may be time
to change it. It is a good maintence to change it every
couple years even if the temps aren't excessive yet when on
a high-heat open flipchip (no head spreader) like an Athlon
XP or older.

 

[Don Taylor]

On Fri, 02 Jun 2006 16:45:24 -0500, Don Taylor

Maybe, I can't see your system. I can't hear my heatsink
fan at all, the only noise escaping the system is from fans
mounted on the case walls with a direct exit- something
Celotex can't fix, but because they are large (and so
possible to use very low RPM, the noise is low.

I wasn't thinking of just heatsink fan but power supply fan,
drive fan, case fan...

I was thinking of having the Celotex extend past the back edge
of the case 8-12 inches and maybe even closed across the back
perhaps half way down the height of the case. That would leave
an exit path for the fan air and cables but would give the fan
noise a chance to bounce off the acoustical tile a few times
before what was left found its way to the room.

Often when I get 'sinks with such tightly spaced fins I take
a piece of sandpaper folded into a V shape and slope the
sharp corners of the 'sink fins some. Result is less dust
getting stuck.

I'll give that a try and see what happens

Decades ago I had a carton of sound absorbing Sonex (tm) foam,
I don't know whether those folks are still in business or not
but they made that stuff for serious sound control, boiler
rooms and such.

thanks

 

[Don Taylor]

....

Someone suggested that yes the fans might all still
be turning, but that the logic controlling them has
slowed them down, in spite of the temperatures being
high.

I'm not sure how to diagnose whether that is really
the cause for the sudden cpu temp jump from 50- to
60+ or the drop to something back near where they
were when I pull the side off the case, and it goes
right back up when I put the side back on.

At least I can work on the problem now when the room
temp is 20+ instead of august when it is 40-.