Hot CPU
Hippopotomonstrosesquipidelian title
Join Date: Oct 2006
Location: is everything
Posts: 1,826
Likes: 0
Received 0 Likes
on
0 Posts
I'm in an environment which appears to be vastly hotter than OFSO's. The only thing I do is make sure a machine is not sitting on a hard surface: if you can space it a couple of cm above the hard deck, convection will take care of the rest.
I've also converted a couple of machines to liquid cooling. Only for the experience, I didn't think it had any absolute benefit until recently. Right now, I'd go for liquid cooling simply because it leaves the case generally uncluttered, and you get a good exhaust fan for free. I've never had a liquid system leak, even in the early days when I used them simply to say I had one. In my particular case, the graphics cards are a greater concern than the CPU. The CPU is the low-cost bit.
Finally, CPU's are cheap. Why not have one in stock?
I've also converted a couple of machines to liquid cooling. Only for the experience, I didn't think it had any absolute benefit until recently. Right now, I'd go for liquid cooling simply because it leaves the case generally uncluttered, and you get a good exhaust fan for free. I've never had a liquid system leak, even in the early days when I used them simply to say I had one. In my particular case, the graphics cards are a greater concern than the CPU. The CPU is the low-cost bit.
Finally, CPU's are cheap. Why not have one in stock?
Join Date: May 2001
Location: Newcastle/UK
Posts: 1,476
Likes: 0
Received 0 Likes
on
0 Posts
The heat sink on my CPU (AMD Phenom 3.6 Ghz Quad Core) has a arrangement of copper pipes running through it,one assumes they has some kind of fluid in them in a closed system came with the CPU when I bought it,pretty heafty fan as well.
Did install a couple of extra four inch fans as well,but it proved to be irritatingly noisy so I took em out,
Of course a lot of cooling is achieved by the case not having a lid,dont believe in lids, built a one that didn't have a case once,worked fine for years.
Behold
Did install a couple of extra four inch fans as well,but it proved to be irritatingly noisy so I took em out,
Of course a lot of cooling is achieved by the case not having a lid,dont believe in lids, built a one that didn't have a case once,worked fine for years.
Behold
Last edited by tony draper; 27th Aug 2012 at 19:48.
Join Date: Jan 2012
Location: .
Posts: 2,173
Likes: 0
Received 0 Likes
on
0 Posts
depends on the machine and case design, but leaving it open can actually impede the air flow and - counterintuitively - hinder coolng
PCs depend very much on a directional air flow, in through one vent, over the CPU/GP, then out via another. On older machines the PSU was the exit, newer ones via a rear vent. The actual internal routing depends on whether the system is ATX, mATX or BTX. But it all boils down to one thing - on a properly designed system taking the side panel off can actually make the machine run hotter
I notice in that photo that there appears to be no exhaust fan, and the CPU fan isn't ducted to the outside, so in your case, taking the side panel off may help.
As for those copper heatpipes - I think you'll find they're solid copper
PCs depend very much on a directional air flow, in through one vent, over the CPU/GP, then out via another. On older machines the PSU was the exit, newer ones via a rear vent. The actual internal routing depends on whether the system is ATX, mATX or BTX. But it all boils down to one thing - on a properly designed system taking the side panel off can actually make the machine run hotter
I notice in that photo that there appears to be no exhaust fan, and the CPU fan isn't ducted to the outside, so in your case, taking the side panel off may help.
As for those copper heatpipes - I think you'll find they're solid copper
Last edited by Milo Minderbinder; 27th Aug 2012 at 21:24.
Join Date: May 2001
Location: Newcastle/UK
Posts: 1,476
Likes: 0
Received 0 Likes
on
0 Posts
Had a four inch fan on that side vent plus one on tother side when I first built it but as I said it sounded like a Phantom spooling up so I dumped em with no ill effects temperaturewise.
Hmmm prompted to take a look due to this thread I reckon tiz time I unhooked the heat sink/fan and give em a good scrape out again.
Hmmm prompted to take a look due to this thread I reckon tiz time I unhooked the heat sink/fan and give em a good scrape out again.
Last edited by tony draper; 27th Aug 2012 at 21:15.
Join Date: Jan 2012
Location: .
Posts: 2,173
Likes: 0
Received 0 Likes
on
0 Posts
never a good idea to remove the heat sink unless you really need to, especially with AMD CPUs.
With the AMD designs theres too much chance of the thermal paste acting like glue, and ripping the CPU out of the socket - possibly killing the board.
Intel designs aren't so much of a problem, though some motherboards object to having the heat sink twistlocks removed and reseated too many times
With the AMD designs theres too much chance of the thermal paste acting like glue, and ripping the CPU out of the socket - possibly killing the board.
Intel designs aren't so much of a problem, though some motherboards object to having the heat sink twistlocks removed and reseated too many times
Join Date: May 2001
Location: Newcastle/UK
Posts: 1,476
Likes: 0
Received 0 Likes
on
0 Posts
Been building me own puters since the days of the 486 Mr Milo, I know how to remove and replace CPUs/heatsinks, the one in the photo has been yanked and cleaned twice since it was built.
Only ever had one CPU fry itself on a grossly overclocked pentium, fan was stalled by the IDE cable,mind you it did die very swiftly,spotted what had happened almost instantly and grabbed for it but too late.
Only ever had one CPU fry itself on a grossly overclocked pentium, fan was stalled by the IDE cable,mind you it did die very swiftly,spotted what had happened almost instantly and grabbed for it but too late.
Join Date: Aug 2012
Location: on the beach
Posts: 359
Likes: 0
Received 0 Likes
on
0 Posts
Just a stray thought, unless you are running a dp center or doing heavily numerical finite element analysis, do you really need all that computing power?
I'm writing this on a teeny 7" mid tablet and it doesn't even get warm. Arm/risc/linux/android-gingerbread.
Micellaneous Cooling Issues
We used to have thick copper buses on all boards to conduct heat away from chips because there's no convection in zero-g and the fans might fail.
I dimly recall reading of piped sulfur as a cooling medium. Don't try this at home.
Coolth may best be found in your dungeon and accessed by airflow through donkey dick or piped liquid through car radiators with fans. Biggest problem is trapping condensation.
Aerospace parts may have wider temperature specs but they tend to lag behind domestic parts.
I'm writing this on a teeny 7" mid tablet and it doesn't even get warm. Arm/risc/linux/android-gingerbread.
Micellaneous Cooling Issues
We used to have thick copper buses on all boards to conduct heat away from chips because there's no convection in zero-g and the fans might fail.
I dimly recall reading of piped sulfur as a cooling medium. Don't try this at home.
Coolth may best be found in your dungeon and accessed by airflow through donkey dick or piped liquid through car radiators with fans. Biggest problem is trapping condensation.
Aerospace parts may have wider temperature specs but they tend to lag behind domestic parts.
Mistrust in Management
Milo
Not sure I understand your post:
Why would the Mboard be killed with no power?
Like Tony I've built (asembled I mean) computers from the 486 up. They are just a poor mans meccano set.
Regards
Exeng
never a good idea to remove the heat sink unless you really need to, especially with AMD CPUs.
With the AMD designs theres too much chance of the thermal paste acting like glue, and ripping the CPU out of the socket - possibly killing the board.
Intel designs aren't so much of a problem, though some motherboards object to having the heat sink twistlocks removed and reseated too many times
With the AMD designs theres too much chance of the thermal paste acting like glue, and ripping the CPU out of the socket - possibly killing the board.
Intel designs aren't so much of a problem, though some motherboards object to having the heat sink twistlocks removed and reseated too many times
Like Tony I've built (asembled I mean) computers from the 486 up. They are just a poor mans meccano set.
Regards
Exeng
Join Date: Jan 2012
Location: .
Posts: 2,173
Likes: 0
Received 0 Likes
on
0 Posts
socket physically fractured
seen it happen several times, with AMD chips, where the CPU is ripped from the socket even though the CPU retention clip arm is still down. Result is the socket can fracture, or in some cases be partially pulled from the board.
The heat transfer stuff (especially the silver colloid types) can simply stick too hard. One way round it is to release the tension clips on the heat sink, and then cut through the heat transfer medium by passing fishing line under the heat sink.
Can't happen with Intel CPUs because of the way they have a kind of retention "shield" over the socket, stopping the CPU pulling away.
Thats an AMD CPU he's got there
seen it happen several times, with AMD chips, where the CPU is ripped from the socket even though the CPU retention clip arm is still down. Result is the socket can fracture, or in some cases be partially pulled from the board.
The heat transfer stuff (especially the silver colloid types) can simply stick too hard. One way round it is to release the tension clips on the heat sink, and then cut through the heat transfer medium by passing fishing line under the heat sink.
Can't happen with Intel CPUs because of the way they have a kind of retention "shield" over the socket, stopping the CPU pulling away.
Thats an AMD CPU he's got there
Last edited by Milo Minderbinder; 28th Aug 2012 at 00:05.
Spoon PPRuNerist & Mad Inistrator
Milo,
More recent AMD Opteron CPUs have used LGA format (socket F, C32, G34) with similar CPU retention clips to intel - I expect that the next generation of AMD desktop chips will do the same. The socket AM2 / AM3 architecture seems to be nearing end of life.
I've had no problems with intel or AMD ZIF sockets over the last 15 years. I've occasionally had a CPU stick to the heatsink and pull out with the ZIF lever still locked, but never any damage to the mobo or socket - or the CPU or pins.
Maybe I've just been lucky, but I'm sure that the ZIF lever is designed to apply a much smaller retaining force than would damage the socket or CPU in the event of a tug of war. I am talking about a straight, vertical pull-out force, not a twisting or rotating force here, as I'm sure that either of these could indeed damage the CPU and / or socket.
SD
More recent AMD Opteron CPUs have used LGA format (socket F, C32, G34) with similar CPU retention clips to intel - I expect that the next generation of AMD desktop chips will do the same. The socket AM2 / AM3 architecture seems to be nearing end of life.
I've had no problems with intel or AMD ZIF sockets over the last 15 years. I've occasionally had a CPU stick to the heatsink and pull out with the ZIF lever still locked, but never any damage to the mobo or socket - or the CPU or pins.
Maybe I've just been lucky, but I'm sure that the ZIF lever is designed to apply a much smaller retaining force than would damage the socket or CPU in the event of a tug of war. I am talking about a straight, vertical pull-out force, not a twisting or rotating force here, as I'm sure that either of these could indeed damage the CPU and / or socket.
SD
Join Date: May 2001
Location: Newcastle/UK
Posts: 1,476
Likes: 0
Received 0 Likes
on
0 Posts
Used to peruse a overclockers website a few years back,some of their home made cooling systems were a bit far out,this was before one could venture to the puter boutique and buy exotic cooling systems orf the shelf as it were
Guest
Thread Starter
Join Date: May 2008
Location: Somewhere between E17487 and F75775
Age: 80
Posts: 725
Likes: 0
Received 0 Likes
on
0 Posts
To wind this thread up (and for the benefit of anyone else with CPU cooling problems in hot weather).
My extra fan blowing air into the CPU cooler duct shows:
a marginal reduction (2º) with an ambient temperature over 30ºc;
a significant reduction (5º) w.a.a.t. under 28ºc.
a probable reduction in CPU temperatures immediately after switch off (CPU fan stopped but extra fan still running).
Conclusion: if you are bored, and you've got the parts lying around in the shed/garage/bedroom, it's worth doing.
My extra fan blowing air into the CPU cooler duct shows:
a marginal reduction (2º) with an ambient temperature over 30ºc;
a significant reduction (5º) w.a.a.t. under 28ºc.
a probable reduction in CPU temperatures immediately after switch off (CPU fan stopped but extra fan still running).
Conclusion: if you are bored, and you've got the parts lying around in the shed/garage/bedroom, it's worth doing.
Join Date: Apr 2004
Location: Civ/HAL/SHY/FYY/PWK/AAS/WAD/AVI/GPT/BZN/BSN/WAD/BAS/FLK/WIT/MND/WAD/WIT/WAD/Civ
Posts: 373
Likes: 0
Received 0 Likes
on
0 Posts
Liquid Cooling
The best to use for liquid cooled systems is cheap supermarket VODKA!
The alcohol prevents algae/mould growth within the pipes.
(and you can drink the rest)
The alcohol prevents algae/mould growth within the pipes.
(and you can drink the rest)
Join Date: Jan 2012
Location: .
Posts: 2,173
Likes: 0
Received 0 Likes
on
0 Posts
as I mentioned earlier, most of the cooling system manufacturers recommend (or did the last time I looked) a glycol-water mix.
I suspect ethanol may be too agressive on the seals in some systems. It certainly may not mix too well if you already have glycol in the pipework
I suspect ethanol may be too agressive on the seals in some systems. It certainly may not mix too well if you already have glycol in the pipework
