The heat pump



  • Whilst working at a small OEM building shoddy PC's I had the chance to meet 'Stan' as I will call him to protect the stupid.

    One day whilst coming up with a new model he was perplexed by overheating issues. "I can't see why the CPU temperature is so high its got the best cooler we've got on it!".

    I took a look and noticed the system temperature was rocketting, it was a cheap case ill prepared for the sheer amount of kit he was trying to squeeze into it and everything was cooking.

    "Stan the system temperature is getting up to like 70ºC"

    "But the CPU has a fan on it, it should be cooling it down."

    "You can't make the CPU any cooler than the surrounding air without some kind of heat pump."

    "But its got a FAN on it, haven't you ever heard of wind chill?"

    I tried explaining the laws of thermodynamics to Stan but he wasn't up for any of my nonsense, he was older and had a posher accent so obviously he was right.


     



  • Next time there's a heat wave turn off the AC and give Stan a fan. If he complains about the heat tell him "But you've got a FAN on you, haven't you ever heard of wind chill?"



  • @EncoreSpod said:

    I tried explaining the laws of thermodynamics to Stan

    There's your own WTF :) It would probably have worked better if you'd said something like: "The air inside the computer is hot. The fan blows that hot air onto the CPU. Do you think blowing hot air onto something cools that thing down?"



  • @DOA said:

    Next time there's a heat wave turn off the AC and give Stan a fan. If he complains about the heat tell him "But you've got a FAN on you, haven't you ever heard of wind chill?"

    Unfortunately, that will backfire. Moving air (even hot air) cools a person because it evaporates perspiration on the skin. The better plan is to remind him that CPUs don't sweat.



  • @Jaime said:

    Unfortunately, that will backfire. Moving air (even hot air) cools a person because it evaporates perspiration on the skin. The better plan is to remind him that CPUs don't sweat.
    I think that might also backfire. Have you never worked with someone who'd take that as an instruction to moisten the CPU?



  •  Actually, if you did it right, that might actually work.

    Back when I had a car that tended to overheat late in the day, I took to dumping a couple of gallons of water over the engine before I started it up on the premise that there would be just a little evaporative cooling effect as soon as I started moving.  Not enough to really fix the problem, but enough to make the difference between making it home and almost making it home.



  • @Jaime said:

    @DOA said:
    Next time there's a heat wave turn off the AC and give Stan a fan. If he complains about the heat tell him "But you've got a FAN on you, haven't you ever heard of wind chill?"
    Unfortunately, that will backfire. Moving air (even hot air) cools a person because it evaporates perspiration on the skin. The better plan is to remind him that CPUs don't sweat.
    Nah, the room just needs to be hot enough. Using a fan in a hot sauna (say, 80°C / 175°F or above) is distinctly contraindicated, unless you want to experience some extra searing heat as the layer of cooler air against your skin is blown away.

     



  • @da Doctah said:

    Actually, if you did it right, that might actually work.

    Back when I had a car that tended to overheat late in the day, I took to dumping a couple of gallons of water over the engine.

     

    You should probably submerge the entire computer in water. Or ferrofluid.

     


  • 🚽 Regular

    If all other attempts at explaining why it doesn't work fail, you can always say the wind chill factor does not cool it enough. Of course, then I guess he'd just buy a bigger fan that can move air faster.



  • @RHuckster said:

    you can always say the wind chill factor does not cool it enough
     

    That would be the truth.

    @RHuckster said:

    Of course, then I guess he'd just buy a bigger fan that can move air faster.

    That would help, but may not enough. Or maybe it would.

     



  • @dhromed said:

    @RHuckster said:

    Of course, then I guess he'd just buy a bigger fan that can move air faster.

    That would help, but may not enough. Or maybe it would.

    Be a fun experiment to observe, mind.



  • [img]http://www.techlivez.com/wp-content/uploads/2007/11/pcfancase.jpg[/img]
    Be mindful of the "more fans solve everything" slippery slope



  • I'm curious to know if the bottom ones actually levitate the cabinet or mow the carpet.



  • @nexekho said:

     

    I'm glad I wasn't taking a sip from a drink.

    @nexekho said:

    Be mindful of the "more fans solve everything" slippery slope

    I think The Hussie had to learn that the hard way.

     



  • @Cassidy said:

    I'm curious to know if the bottom ones actually levitate the cabinet or mow the carpet.

     

    It is basically a dirigible.

     



  • @dhromed said:

    I'm glad I wasn't taking a sip from a drink.

    I think limited resistance to spills was part of the reason this design never took off.




  • The secret is that his name really is Stan, isn't it?

    I bet most of you have seen this, but it bears relinking for the 200th time

    The concept is not that different from using liquid cooling, but the execution is so much awesomer.



  • @nexekho said:


    Be mindful of the "more fans solve everything" slippery slope

    Must be a Fujitsu box.

    Back in the olden days, a friend and I were allowed to take apart servers before deploying them (I don't know why our boss let us do this). We took apart a 3u server which had 24 fans. Each cooling point had a fan, and each fan had a friend. What with redundant power supplies and cpu/memory banks... yeah. That server sounded like it was taking off when we finally powered it on.
    We figured there must be a Fujitsu subsidiary which did nothing but make fans.



  • @tweek said:

    We took apart a 3u server which had 24 fans. Each cooling point had a fan, and each fan had a friend. What with redundant power supplies and cpu/memory banks... yeah. That server sounded like it was taking off when we finally powered it on.
    An HP blade chassis has 10 cooling units that are tunnel fans with five individual fans per unit. That's 50 fans with an exit temperature of around 150 degrees. If you stand behind a rack of these things, your skin goes dry in like 60 seconds.



  • I used to have an 80386 AT&T Unix machine with three 10 inch fans on the bottom of the machine, two power-supply fans and an intake fan on the top, and three ten inch fans on the rear of the machine. When you powered up the machine, one pair of 10 inch fans would come on low, followed by a second set on low when the redundant supply came up, then the final set right around when the LED display on the front started displaying the POST results. 

    The two by two process repeated itself during POST, as all the fans went from low to high. I knew the system had started up when the jet engine throttled back to low and I could hear myself think.

    I eventually gutted it and put two motherboards, three power supplies, five disks, and six CDROM drives in it. It was waaaay quieter then. 

    The only system I've had louder was a CAD workstation made by Inter-somebody back in the days of custom multiprocessor Intel boards. It actually had a 24V 12x12x6 cage fan in it, the same size as the ones used in the old CISC AS/400s. Unlike the one in the AS/400 the damned thing ran on high all the time and had no attempt to deaden the noise. No rubber mounts and the intake 'duct' vibrated in time with the fan.



  • (I was looking for "cool story, bro" but couldn't find it.)



  • @Jaime said:

    @tweek said:
    We took apart a 3u server which had 24 fans. Each cooling point had a fan, and each fan had a friend. What with redundant power supplies and cpu/memory banks... yeah. That server sounded like it was taking off when we finally powered it on.
    An HP blade chassis has 10 cooling units that are tunnel fans with five individual fans per unit. That's 50 fans with an exit temperature of around 150 degrees. If you stand behind a rack of these things, your skin goes dry in like 60 seconds.

    Are those the ex-compaq blades or hp-designed ones? I had a Compaq blade server that was somewhat logically made (IIRC) back 'round the turn of the century, I'm assuming hp reengineered that (put blue and white plastic on it and threw more fans at it).



  • @tweek said:

    @Jaime said:
    @tweek said:
    We took apart a 3u server which had 24 fans. Each cooling point had a fan, and each fan had a friend. What with redundant power supplies and cpu/memory banks... yeah. That server sounded like it was taking off when we finally powered it on.
    An HP blade chassis has 10 cooling units that are tunnel fans with five individual fans per unit. That's 50 fans with an exit temperature of around 150 degrees. If you stand behind a rack of these things, your skin goes dry in like 60 seconds.
    Are those the ex-compaq blades or hp-designed ones? I had a Compaq blade server that was somewhat logically made (IIRC) back 'round the turn of the century, I'm assuming hp reengineered that (put blue and white plastic on it and threw more fans at it).
    HP.  To be fair, 50 fans sounds more reasonable when you consider that this is a 10U unit that can house up to 32 servers.  I like almost everything about it, but it is quite an impressive generator of heat.  With 1200 watts of power supplies (for a total of 4800 watts per rack), that heat has to go somewhere.



  • @nexekho said:

    Be mindful of the "more fans solve everything" slippery slope

    Oddly, it appears that SOME of the fans in front are exhaust, whilst others are intake ...



  • How can you.. oh.. the labels.

    Actually, that's something I've been meaning to find out: for a processor fan, should it be intake-up (blowing air down onto the chip) or exhaust-up (venting hot air away from the chip)?

    I've always thought it the latter from the time I had old Pentium-S with integrated fans that created a lukewarm breeze.



  • @Cassidy said:

     exhaust-up (venting hot air away from the chip)?

    This way



  • @Cassidy said:

    Actually, that's something I've been meaning to find out: for a processor fan, should it be intake-up (blowing air down onto the chip) or exhaust-up (venting hot air away from the chip)?

    The correct answer is: read the goddamned directions before installing.



  • @Cassidy said:

    Actually, that's something I've been meaning to find out: for a processor fan, should it be intake-up (blowing air down onto the chip) or exhaust-up (venting hot air away from the chip)?

    Both are used. As Blakey pointed out, read the instructions first.



  • That would presume some were provided...



  • Put enough (Exhaust) fans on a system and suck all of the air out. No convection cooling and watch the temperature skyrocket.

     Actually had this as a problem on a mil embedded system about 20 years ago.



  • I saw an experiment (possibly a video on Tom's HW) where it was shown that component temperatures on a case with cover removed was higher than the enclosed case - cooling (presumably through airflow via the case fans) was impacted without the cover.



  • I've seen a retard who destroyed a processor because instead of the regular thermal paste that you use between the CPU and the cooler, he used toothpaste.

    He thought that since it gives a chilling sensation to his mouth, it would also "cool the CPU the same way".



  • @Renan said:

    I've seen a retard who destroyed a processor because instead of the regular thermal paste that you use between the CPU and the cooler, he used toothpaste.

    He thought that since it gives a chilling sensation to his mouth, it would also "cool the CPU the same way".

    So let me get this straight... this retard was smart enough to open up his computer, where to find the CPU on the mother board, how to remove the heat sink (or cooler or whatever the heck it is called), knew that a paste is used to conduct the heat away from the CPU, but yet was stupid enough to think that toothpaste could do the same job as thermal paste?  Did you by any chance suggest he try to brush his teeth with thermal paste since it would "give a chilling sensation to his mouth just like it does for a CPU".



  • @Anketam said:

    @Renan said:

    I've seen a retard who destroyed a processor because instead of the regular thermal paste that you use between the CPU and the cooler, he used toothpaste.

    He thought that since it gives a chilling sensation to his mouth, it would also "cool the CPU the same way".

    So let me get this straight... this retard was smart enough to open up his computer, where to find the CPU on the mother board, how to remove the heat sink (or cooler or whatever the heck it is called), knew that a paste is used to conduct the heat away from the CPU, but yet was stupid enough to think that toothpaste could do the same job as thermal paste?  Did you by any chance suggest he try to brush his teeth with thermal paste since it would "give a chilling sensation to his mouth just like it does for a CPU".

    This was somewhat common in my country (of course most people here don't have a clue what thermal paste is and never ever use it).  It seems that the toothpaste worked because I never heard of somebody who burned a cpu from that (however they pulled the toothpaste from the market, something to do with it being bad for your health as people also ate it when hungry)



  • @serguey123 said:

    @Anketam said:

    @Renan said:

    I've seen a retard who destroyed a processor because instead of the regular thermal paste that you use between the CPU and the cooler, he used toothpaste.

    He thought that since it gives a chilling sensation to his mouth, it would also "cool the CPU the same way".

    So let me get this straight... this retard was smart enough to open up his computer, where to find the CPU on the mother board, how to remove the heat sink (or cooler or whatever the heck it is called), knew that a paste is used to conduct the heat away from the CPU, but yet was stupid enough to think that toothpaste could do the same job as thermal paste?  Did you by any chance suggest he try to brush his teeth with thermal paste since it would "give a chilling sensation to his mouth just like it does for a CPU".

    This was somewhat common in my country (of course most people here don't have a clue what thermal paste is and never ever use it).  It seems that the toothpaste worked because I never heard of somebody who burned a cpu from that (however they pulled the toothpaste from the market, something to do with it being bad for your health as people also ate it when hungry)

    Maybe you guys use different stuff to make toothpaste, maybe it has to do with the amount applied... I'm just not willing to test the feasibility of substituting toothpaste for the thermal paste used between the CPU and the heat sink (thanks, Anketam) in a controlled environment. That could be worth an ignobel if one ever published a paper on it.



  • I did some Internet research so I cannot guarantee the accuracy of this, but people have claimed that you can use a very thin layer of tooth paste just like thermal paste, with only two caveats.  First tooth paste conducts electricity which means you have to clean up all excess or else you can short something in a very bad sort of way.  Secondly it only lasts 2-3 months before it dries out and turns from a heat conductor to a heat insulator.



  • @blakeyrat said:

    @Cassidy said:
    Actually, that's something I've been meaning to find out: for a processor fan, should it be intake-up (blowing air down onto the chip) or exhaust-up (venting hot air away from the chip)?
    The correct answer is: read the goddamned directions before installing.

    Hmmm, the correct answer is check it out before buying it.

    @morbiuswilter said:

    Both are used. As Blakey pointed out, read the instructions first.

    I have never seen a cpu cooler pushing air into the cpu.  Maybe they are uncommon and the design seems retarded from a thermal point of view

     



  • @serguey123 said:

    I have never seen a cpu cooler pushing air into the cpu.  Maybe they are uncommon and the design seems retarded from a thermal point of view

     Quite rare (not to menion ineffective) when the fan is moving ambient air. Quite common when there is a "refigerated air source". You will find the latter in many embedded/industrial applications, but I have only seen one "PC" application (more common to use water cooling)



  • @TheCPUWizard said:

    @serguey123 said:

    I have never seen a cpu cooler pushing air into the cpu.  Maybe they are uncommon and the design seems retarded from a thermal point of view

     Quite rare (not to menion ineffective) when the fan is moving ambient air. Quite common when there is a "refigerated air source". You will find the latter in many embedded/industrial applications, but I have only seen one "PC" application (more common to use water cooling)

    Jesus fuck, the point is always read the instructions for the specific component you're about to install, just don't fucking "guess and check" based on what you've seen before. That's what I'm trying to get across, read the goddamned instructions. It's not hard.


  • ♿ (Parody)

    @blakeyrat said:

    That's what I'm trying to get across, read the goddamned instructions. It's not hard.

    Yeah, but with the broken-ass Chinglish you get, making sense of what you read is funny at best, impossible at worst.



  • @boomzilla said:

    @blakeyrat said:
    That's what I'm trying to get across, read the goddamned instructions. It's not hard.
    Yeah, but with the broken-ass Chinglish you get, making sense of what you read is funny at best, impossible at worst.

    That is why I love Noctua, pretty pictures!



  • @blakeyrat said:

    Jesus fuck, the point is always read the instructions for the specific component you're about to install, just don't fucking "guess and check" based on what you've seen before. That's what I'm trying to get across, read the goddamned instructions. It's not hard.

    I think you missed my response of "that would presume there were any instructions".

    My second thought was to ask people that may know this stuff, rather than guess and check. However, judging by your response, that may not have been the best course of action either...



  • @Cassidy said:

    @blakeyrat said:

    Jesus fuck, the point is always read the instructions for the specific component you're about to install, just don't fucking "guess and check" based on what you've seen before. That's what I'm trying to get across, read the goddamned instructions. It's not hard.

    I think you missed my response of "that would presume there were any instructions".

    If there's no instructions, in the box or online, you need to stop buying cheap shit.



  • @blakeyrat said:

    If there's no instructions, in the box or online, you need to stop buying cheap shit.

    You also presume I bought it.

    To head off anymore of your presumptions at the pass: the machine was "inherited", the fan was already placed in a certain position hovering over a Zalman flower, but this bought it close to the side of the cabinet, and some of us were debating which was the correct way (compared to a Dell of a similar make in which the fans are arranged across the passive heatsink to duct airflow from the front out of the back).

    Thanks to those that did give a useful answer - which I've understood as "it's generally blowing hot air up away from the heatsink but it could be either"



  • @Cassidy said:

    Thanks to those that did give a useful answer - which I've understood as "it's generally blowing hot air up away from the heatsink but it could be either"
    I've yet to see a CPU fan that doesn't blow air on the heatsink - heck, some cases have a tunnel above the CPU fan to directly bring in the cooler air from outside. As for Zalman flower, I presume it's one of those old heatsinks with a fan that isn't mounted directly on the heatsink - I'd expect the cooling there will be much better if the fan blows on the heatsink than the reverse.



  •  It's one of these (finally found an image of it - it's not on Zalman's website anymore)

    Okay...so the fan should be blowing DOWN onto it? Hmm...

    (BTW - what's broken with your tags, Ender? It's difficult to read your post)



  • Yeah, I was thinking of this heatsink. If you have a case that has an opening above the CPU, it's best that you screw the fan there, and have it pull air inside
    As for my tags, they're just fine :)



  • @Cassidy said:

    (BTW - what's broken with your tags, Ender? It's difficult to read your post)

    He's the asshole* who reversed all of the tag cloud.



  • @ender said:

    @Cassidy said:
    Thanks to those that did give a useful answer - which I've understood as "it's generally blowing hot air up away from the heatsink but it could be either"
    I've yet to see a CPU fan that doesn't blow air on the heatsink - heck, some cases have a tunnel above the CPU fan to directly bring in the cooler air from outside. As for Zalman flower, I presume it's one of those old heatsinks with a fan that isn't mounted directly on the heatsink - I'd expect the cooling there will be much better if the fan blows on the heatsink than the reverse.

    Yeah, that's what I thought. It's been a decade since I've built my own overclocked computers, but I'm pretty sure most fans blow down. However, I rarely used axial fans. I had a huge, high-CFM, 115-volt AC radial blower mounted on the outside of the case that blew in right over the CPU and GPU. No other fans needed.. :)


  • Garbage Person

    @morbiuswilters said:

    @ender said:
    @Cassidy said:
    Thanks to those that did give a useful answer - which I've understood as "it's generally blowing hot air up away from the heatsink but it could be either"
    I've yet to see a CPU fan that doesn't blow air on the heatsink - heck, some cases have a tunnel above the CPU fan to directly bring in the cooler air from outside. As for Zalman flower, I presume it's one of those old heatsinks with a fan that isn't mounted directly on the heatsink - I'd expect the cooling there will be much better if the fan blows on the heatsink than the reverse.

    Yeah, that's what I thought. It's been a decade since I've built my own overclocked computers, but I'm pretty sure most fans blow down. However, I rarely used axial fans. I had a huge, high-CFM, 115-volt AC radial blower mounted on the outside of the case that blew in right over the CPU and GPU. No other fans needed.. :)

    Axial fans are for suckers. Real men use squirrelcages!

    [i]Real[/i] real men use automotive blowers. 40A worth of fan dangling off your 12v rail.


Log in to reply