Solve me this: Case temp above CPU temp in retail?

Me and my friend are debating this. Here is the story:

My friend has a P3 733MHZ Coppermine, with a Retail Intel cooler, 256MB SDRAM and WinXP. His Asus Probe reports this relatively too low of a temp, 17º, and 25º Full load. I simply find this too low and unbeleivable!
Next, I tell him that he should use a thermometer and measure Case temps, so maybe then we'd see more detail. I claimed that usually, on conventional Retail cooling, Case temps are almost always below CPU temps. But that makes it impossible almost, to have below 20º Case temp, in a Canadian weather of 25º outside!
He claims it is possible and he has that. Currently he can't measure his case temp cuz he has no thermometer, but my point still stands, either Asus Probe for his 733MHZ is really screwed up or something in the logic is not working.

Usually, can case temps on retail cooling be HIGHER than a CPU temp which is THIS low? And what is to beleive?

Thanks for any argument or suggestion!

He's now going to check in BIOS the temps.

EDIT: Ok his BIOS reports 24º after reboot. This makes a bit more sense but still it is a low temp don't you guys think, for retail?

--
Meow<P ID="edit"><FONT SIZE=-1><EM>Edited by Eden on 06/06/02 10:19 PM.</EM></FONT></P>
31 answers Last reply
More about solve this case temp temp retail
  1. Completely different system but anyway:

    My old PII 350 and AT case used to run at:

    <b>Winter:</b>
    CPU: 24ºC (idle) 34ºC (load)
    Case:25ºC (idle) 32ºC (load)

    <b>Summer:</b>
    CPU: 32 (idle) 42ºC (load)
    Case:33 (idle) 40ºC (load)

    As you see, at load the CPU is hotter than the case.

    The northbridge of this motherboard used to be very hot so think it is producing some “distortion” with the temp readings. I think it’s just a heat distribution problem within the case because if I set up a better ventilation (open case and home fan or connecting the vacuum cleaner) the temperatures behave “normal”, I mean CPU always hotter than motherboard.


    BTW, With those extremely low temperatures what is your friend waiting to overclock?


    <b>(<font color=yellow>as good as it looks</font color=yellow>)</b>
  2. u do know that analprobe is a complete pile of sh1te right?

    just use Motherboard Monitor.

    <font color=blue>Pants Down! Turn Around! Bend Over! You're about to Experience Telstra broadband! :lol:
  3. Yeah he should. But what is weird is his BIOS also reports 25º. I still have a hard time beleiving it, I mean really, a P3 733MHZ at 133MHZ FSB this low with Retail fan? Where have all the "cool" NWs gone!
    I should tell him to try OCing, but I wanna get down to this story, as it just doesn't make logic. I don't recall once putting my hand inside a case and feel it rather cold, when it has all PCI ports used, possibly the AGP, and a retail cooler.

    --
    Meow
  4. A chip can never be cooler than the room its in.

    Its a thermal impossibility.


    :wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
  5. Thank you, that is what I wanted to know. Unless water cooled, it makes no sense to have a bigger case temp than a CPU, under retail cooling solutions, case closed, although the CPU has a good aeration and is not stuck around it.

    But if you say in the room it's in, how about the case it's in?

    --
    Meow
  6. Quote:
    A chip can never be cooler than the room its in. Its a thermal impossibility

    Fully agreed, true CPUs temp always is hotter. However, <b>readings</b> of the case temperature sensor can be hotter because it could be located in a bad ventilated area. In may case it was between the northbridge and CPU heatsinks.


    <b>(<font color=yellow>as good as it looks</font color=yellow>)</b>
  7. Quote:
    A chip can never be cooler than the room its in.

    Its a thermal impossibility.


    I think you mean an air cooled chip. I've seen several processors at -20c, and I don't think the people using them were in a room that cold.

    <font color=blue>Hi mom!</font color=blue>
  8. yes it can.
    eg my current situation with a senfu thermometer measure (+/- 0.1°C):
    23.8°C ambient air temp.
    22.2°C iron case temp.

    add-ons:
    25°C mb temp.
    45°C cpu temp.


    <i>if <b>you know</b> <font color=white>you don't know</font color=white>, the way could be more easy ...<P ID="edit"><FONT SIZE=-1><EM>Edited by Labdog on 06/07/02 06:56 PM.</EM></FONT></P>
  9. Quote:
    A chip can never be cooler than the room its in.

    Its a thermal impossibility.

    yes it can.

    eg my current situation with a senfu thermometer measure (+/- 0.1°C):

    23.8°C ambient air temp.
    22.2°C iron case temp.

    add-ons:
    25°C mb temp.
    45°C cpu temp.


    <i>if <b>you know</b> <font color=white>you don't know</font color=white>, the way could be more easy ...
  10. So you just proved his point. What am I missing?

    <font color=blue>Hi mom!</font color=blue>
  11. </font color=red><font color=red>eg my current situation with a senfu thermometer measure (+/- 0.1°C):

    23.8°C ambient air temp.
    22.2°C iron case temp.
    </font color=red>


    <i>if <b>you know</b> <font color=white>you don't know</font color=white>, the way could be more easy ...
  12. Quote:
    add-ons:
    25°C mb temp.
    45°C cpu temp.


    You might need to re-read what Matisaro said.

    <font color=blue>Hi mom!</font color=blue>
  13. did i read that right? anal probe? lol .... sorry south park redefined anal probe and it's a funny word to me...
    *shrugs*


    <A HREF="http://www.anandtech.com/mysystemrig.html?id=9933" target="_new"> My Rig </A>
  14. Quote:
    <font color=red>ADD-ONS</font color=red>:
    25°C mb temp.
    45°C cpu temp.



    <i>if <b>you know</b> <font color=white>you don't know</font color=white>, the way could be more easy ...
  15. Quote:
    <i>xxsk8er101xx says:</i>
    did i read that right? anal probe?


    Yeup, it's what people have started calling Asus Probe. It's fitting, I think :lol:

    Quote:
    <i>Labdog says:</i>
    <insert spam here>


    What exactly are you trying to say?

    <font color=blue>Hi mom!</font color=blue>
  16. what?


    <i>at least, that discussion between 2 blind deaf ppl fit the posts counter increase. thus everything is not lost.</i>


    <i>if <b>you know</b> <font color=white>you don't know</font color=white>, the way could be more easy ...
  17. Yeah, since his "MB temp" as Asus Probe states, is 32º. Now how exactly is that possible when CPU temp is around 25, I dunno. In fact today his temp was up at 30º (Probe even reported 13º for a second, talk about cheap monitoring!!!), which I find plausible since it's one of the early makes of 0.18m, right?

    His Mobo is an Asus CUV4X IIRC, he scanned me his manual page stating the internal diode reader, which makes his statement a tad firmer, but nonetheless it proves nothing as to case temps. He also has made a pic of the case, so if anyone wants to see how I described it and maybe evaluate my case here, PM me with your e-mail!

    My point isn't to say he's wrong, it's just that I am looking from a logical POV, with the following facts:
    -Retail Cooled P3
    -Aircooled
    -His other house comps like a 900MHZ Cely reported at 40º although I find it very faulty, since celeries don't generate this much!
    -Good airflow in case, no wiring in front of CPU fan
    -Average weather of 20º outside, with hotter temps inside

    There are more but you get my point. Anyway thanks for the arguments, any way I can learn more of this, is good.

    --
    Meow
  18. Asus probe often reports temps BELOW AMBIENT for Intel CPU's. Now we know this is impossible. I found by experimenting with various versions that at least one version is usually fairly accurate for a certain board.

    What's the frequency, Kenneth?
  19. NOTE: the product name "Anal Probe" when used to refer to "Asus Probe" is copyrighted by yours truly. :smile:
    A one off home use license costs $5 US.
    A commercial license to use the name 'anal probe' an unlimited number of times costs $25 US. :lol:


    <font color=blue>Pants Down! Turn Around! Bend Over! You're about to Experience Telstra broadband! :lol:
  20. Quote:
    I think you mean an air cooled chip. I've seen several processors at -20c, and I don't think the people using them were in a room that cold.


    Yes, and if water cooled the chip can never be cooler than the temprature of the water, unless one is using a peltier, but then the chip can never be cooler than the cold side of the device.

    :wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
  21. Ok I just finished my debate with him.

    Here are the results: By putting his thermometer in the middle area of the case, a bit closer to the case fan but in an area far from the PCI cards, and well aerated, it generates 25. His CPU reading in BIOS states 24-25 or above. That would mean we both have flawed arguments, just not entirely true. Yes it can be below case temp, but in rare cases, AND never more than 4º. If it was, something is definitly not right. So in conclusion his temps hover at or 1º below case temp, then go up higher than Case temps in heavy load.

    Nice debate though, it was fun yelling at him constantly to get the damn thermo in the right place, since he first put it just near the CPU, while it touched the chipset heatsink, talk about cheating lol.

    Anyway case closed.

    --
    Meow
  22. The chip cant be lower than the rooms air temp, case temp is usually higher than airtemp, and if the case fans blow directly on the hsf, then the cpu can theoretically be below case temp, but I said room temp so its moot.

    :wink: The Cash Left In My Pocket,The BEST Benchmark :wink:
  23. how much cooler is the Thoroughbred 2200+ going to be than my Palomino 1900+.....it's getting too warm and hanging the system again....so instead of getting an aluminium case, i'd go for a .13 chip (colder and a mini-upgrade all in one)
  24. Well the room temp is usually less high than case temp, as everything inside is heat. Even with case fans, (he put the thermo near an exhaust one, so no air is being blown in to bias the thermo temp) it still shouldn't be much lower. One thing I noticed, is when he had the thermo near the CPU, the temp was 27. Oddly enough it may be more accurate, but then again the HS of the chipset was underneath it. It does however bring the question that maybe the P3's internal sensor is malfunctioning, and actually the P3's temp is above 30? In any case, I think my conclusion was fair and right, it exists that chip temp can be below case temp , but not more than 3-4º, on RETAIL AIRCOOLING. Otherwise something is not working, and it is to be blamed on very flawed temp reading.

    --
    Meow
  25. Quote:
    Well the room temp is usually less high than case temp, as everything inside is heat. Even with case fans, (he put the thermo near an exhaust one.....

    stop obsessing about your friends "better rig".

    "<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
  26. Sure, a 733MHZ is better in Meltdown's world, than an XP1600.

    --
    Meow
  27. Quote:
    and if the case fans blow directly on the hsf, then the cpu can theoretically be below case temp

    Yes, it’s a fact, actually that’s the principle of the <b>air ducts</b> in CPU cooling.


    <b>(<font color=yellow>as good as it looks</font color=yellow>)</b>
  28. CPU temps can never be below the case temps with air cooling, with the following exceptions:

    1) Air ducting. Taking cooler air and dumping it directly on a CPU. The CPU temp may them be lower than the ambient case temp, but not lower than the cooler air that you are ducting in.

    2) Evaporative. In theory if his heats sink is wet, it is possible to go a degree or two below ambient as the water (or other liquid) absorbs the heat and evaporates off. That's why humans sweat.

    3) Peltier or other device.

    The same holds true for water cooling. Using a rad you will never get below the ambient temp that surrounds the rad. Using a evaporative tower (bong) you can get a degree or two lower. Exceptions are if the water is pressurized.

    As for Labdog, I think what he was trying to say is that his metal case was a degree lower than the ambient temp inside the case. I think that he'll find that it is not lower than the ambient temp OUTSIDE the case though.

    As for Eden's friend, it is much more likely that he is getting an incorrect reading from Asus Probe. Have him touch his thermomitor off the heatsink. The CPU will be a little bit warmer than the heatsink.

    --------------
    Knowan likes you. Knowan is your friend. Knowan thinks you're great.
  29. I think when he took the temp with the cheated method, it may have been put in a position touching the HSF, however there is no way to touch the CPU, as it is entirely covered by the HSF. The temp then was 27. I fully agree the Probe was wrong, I mean seriously, 17ºC, then 13º? Lol give me a break, water cooling at its worst could give such temps, but in no way could a P3 CuMine at 1.6V do that with the Retail Intel cooler.

    I was thinking maybe that fan to fan method would have indeed provided a much cooler CPU than case, but from the pics he sent, it doesn't look that way at all, he'd have to put the case fan on the case side, which would have been removed and therefore not blowing on the HSF, since he needed to take off the case to take pics.

    --
    Meow
  30. Quote:
    Sure, a 733MHZ is better in Meltdown's world, than an XP1600.

    ok, sounds like he's getting more out of his 733 than you with your XP1600 your only mission critical app is Unreal tournament.

    "<b>AMD/VIA!</b>...you are <i>still</i> the weakest link, good bye!"
  31. What exactly is he getting more of? Crappy gaming on a TNT2?
    Or is simply cuz he has lower temps, which don't matter as long as a system works stably?

    Sorry Melty you're trying too much now...

    --
    Meow
Ask a new question

Read More

CPUs Cooling Cases