Did heat kill your hardware?
Moderators: NeilBlanchard, Ralf Hutter, sthayashi, Lawrence Lee
-
- Posts: 199
- Joined: Wed Nov 10, 2004 1:44 pm
- Location: Sweden
Did heat kill your hardware?
Silence is good and all that, but it does make your hardware run hotter. Have you killed your hardware with heat? Please post your stories.
Also, which hardware is most sensitive to heat, i.e. when you push your HDD or PSU into 50-60 deg Celsius, what goes first?
------------------------
Recently my HDD crashed, and I suspect heat. It was sitting suspended in a 5.25 bay, with no fan or natural circulation around it. Had to buy a new one, and then the DVD laser in my combo drive died too. Hopefully my new setup will be easier on the stuff, HDD now sitting on a ghetto suspension right above a 120 intake fan.
/datapappan
Also, which hardware is most sensitive to heat, i.e. when you push your HDD or PSU into 50-60 deg Celsius, what goes first?
------------------------
Recently my HDD crashed, and I suspect heat. It was sitting suspended in a 5.25 bay, with no fan or natural circulation around it. Had to buy a new one, and then the DVD laser in my combo drive died too. Hopefully my new setup will be easier on the stuff, HDD now sitting on a ghetto suspension right above a 120 intake fan.
/datapappan
Some type of air circulation around the hard drive is probably a good idea. I noticed in my Antec P180 case that if I take the side panel off the case, the hard drive temperatures will rise dramatically, like 10-15C if the hard drive is busy.
As for failures, I have never had one, even in spite of the fan on my Zalman VF-700 sometimes halting due to a lousy power connector. I saw a peak of 116C (!!) reported one time before the game finally crashed, returning me to my regular windows screen where I saw the temps off the top of the chart in rivatuner, with the "find peak" option showing that temperature. Still, that was at least two months ago and my video card still works fine, as best as I can tell.
As for failures, I have never had one, even in spite of the fan on my Zalman VF-700 sometimes halting due to a lousy power connector. I saw a peak of 116C (!!) reported one time before the game finally crashed, returning me to my regular windows screen where I saw the temps off the top of the chart in rivatuner, with the "find peak" option showing that temperature. Still, that was at least two months ago and my video card still works fine, as best as I can tell.
-
- Moderator
- Posts: 7681
- Joined: Mon Dec 09, 2002 7:11 pm
- Location: Maynard, MA, Eaarth
- Contact:
-
- Posts: 239
- Joined: Sat Mar 29, 2008 7:05 pm
- Location: Toronto Ontario
I'm embarrassed to say that I do. I tend to upgrade relatively frequently so component life isn't all that important to me. Hard drives however are one of those things you don't need to upgrade and will always have use in one way or another. You don't want data loss due to lack of airflow. Another solution would be to back up your hard drives in RAID or even clone it and swap it out every 2 or 3 years. At the rate storage space price is coming down, it's quite affordable to swap it out every now and then.NeilBlanchard wrote:Nobody here keeps their computer quiet at the risk of longevity.
-
- Moderator
- Posts: 7681
- Joined: Mon Dec 09, 2002 7:11 pm
- Location: Maynard, MA, Eaarth
- Contact:
-
- Posts: 239
- Joined: Sat Mar 29, 2008 7:05 pm
- Location: Toronto Ontario
-
- Moderator
- Posts: 7681
- Joined: Mon Dec 09, 2002 7:11 pm
- Location: Maynard, MA, Eaarth
- Contact:
-
- Posts: 199
- Joined: Wed Nov 10, 2004 1:44 pm
- Location: Sweden
I too have the purpose of quietness as well as coolness.
Case in point was, probably I missed out on hard drive cooling whilst tinkering with case setups - I can't know for sure, since my speedfan log went in the "fire". Did heat kill it? It's been used för three and a half years, maybe it's time had come. I will definitely get a proper back-up device soon, at the latest when we get ourselves a digital camera (did I hear stone age?)
So, +116 C on a GPU before shut-off.. How high are your hard drives running? (I had my CPU go to + 78 C befor shut-down).
/ datapappan
Case in point was, probably I missed out on hard drive cooling whilst tinkering with case setups - I can't know for sure, since my speedfan log went in the "fire". Did heat kill it? It's been used för three and a half years, maybe it's time had come. I will definitely get a proper back-up device soon, at the latest when we get ourselves a digital camera (did I hear stone age?)
So, +116 C on a GPU before shut-off.. How high are your hard drives running? (I had my CPU go to + 78 C befor shut-down).
/ datapappan
After switching from a Super Tornado ATX PSU to a fanless PicoPSU, case temps went up significantly without the airflow and one of the drives in the system generated a smart error because of the temp of the drive. It generated an error at 55*C - touching the drive it was very hot. The other drive in the system doesn't appear to monitor drive temps, so I don't know how hot that one got, but I'm sure it was similar since the drives sit right next to each other.
The drives weren't set up with any real airflow before and would consistently run around 50*C without issue for at least a year now.
However, I have since added a case fan and can easily bring drive temps down to mid 30*C level by cracking open the drive bay face plate. Right now with the faceplate closed it is sitting at mid 40*C levels. I think I'm going to leave the face plate cracked open or drill a few holes in the face plate to let some air through past the drives.
Moral of the story - At a minimum, keep those hard drives below 50*C. In theory cooler is better with hard drives, but I don't know at which point going cooler makes a significant difference.
The drives weren't set up with any real airflow before and would consistently run around 50*C without issue for at least a year now.
However, I have since added a case fan and can easily bring drive temps down to mid 30*C level by cracking open the drive bay face plate. Right now with the faceplate closed it is sitting at mid 40*C levels. I think I'm going to leave the face plate cracked open or drill a few holes in the face plate to let some air through past the drives.
Moral of the story - At a minimum, keep those hard drives below 50*C. In theory cooler is better with hard drives, but I don't know at which point going cooler makes a significant difference.
That matches my experience and feelings. My HD temps are normally around 37-43C from idle to sustained max thrashing, but SMART data shows a lifetime peak of 53C - probably back when I ran with the side of the case off. Even a little airflow over the drives makes a big difference, but with no airflow, they get quite hot. At least speaking of 3.5" drives. I'm guessing anyone with 2.5" drives doesn't have to worry as much about heat.
I burned up an Athlon64 CPU by mounting the heatsink rotated 180 degrees so that it was not making good contact with the die.
Only other parts I remember having problems with were really cheap PSUs supplied with cheap cases. (And yes, I have long since started buying decent cases [for low end systems, still have a couple Antec 1650's in service which I am still satisfied with] and better PSUs.)
Only other parts I remember having problems with were really cheap PSUs supplied with cheap cases. (And yes, I have long since started buying decent cases [for low end systems, still have a couple Antec 1650's in service which I am still satisfied with] and better PSUs.)
I've had a few things die from heat.
A disk that died in a crappy HP case (SMART said it had gotten to 55C).
A motherboard VRM (actually a capacitor in the VRM) that was perennially overheated trying to feed an overclocked Pentium 830D (close to 150W). The capacitor literally exploded, as electrolytics are prone to do after a few months of being very hot.
And a GPU (it isn't dead, but it produces artifacts at 85C). This last is of the most interest: when it was new, it ran perfectly at 105C, and I did run it that hot a few times during benchmarks and tests. It didn't start to show problems until almost a year later. (Its throttling temperature is 130C.)
A disk that died in a crappy HP case (SMART said it had gotten to 55C).
A motherboard VRM (actually a capacitor in the VRM) that was perennially overheated trying to feed an overclocked Pentium 830D (close to 150W). The capacitor literally exploded, as electrolytics are prone to do after a few months of being very hot.
And a GPU (it isn't dead, but it produces artifacts at 85C). This last is of the most interest: when it was new, it ran perfectly at 105C, and I did run it that hot a few times during benchmarks and tests. It didn't start to show problems until almost a year later. (Its throttling temperature is 130C.)
Just killed a motherboard due to long term high temperatures over north bridge and south bridge. Actually the south bridge failed first, creating problems with most devices connected to the PC (mostly USB). This was a heatpipe cooled motherboard, and i was watercooling the CPU and GPU so just used the provided minifan over the VRM heatsink. This was not enough, obviously, as the rest (especially the NB) didnt get any airflow.
My next motherboard (building it right now) will have VRM and chipset completely watercooled in their own loop. I wont trust motherboard manufacturer claims about their "uber cooling heatpipe" thingies anymore.
My next motherboard (building it right now) will have VRM and chipset completely watercooled in their own loop. I wont trust motherboard manufacturer claims about their "uber cooling heatpipe" thingies anymore.
I do. I do it all the time, largely because most of my computer hardware is old "junk" anyway. I've killed plenty of hardware, although I really don't know how much of it was due to heat. For example, I think one of the old 2.5" drives I killed with a homebrew enclosure was due to excessive pressure rather than overheating.NeilBlanchard wrote:Nobody here keeps their computer quiet at the risk of longevity.
I've killed a number of power supplies and motherboards, although again I really can't tell whether it was due to overheating. In the case of the power supplies, I believe some of them died to overheating because I left them on too long without checking for dust and the intake grill was totally covered.
As far as I can tell, I've never managed to kill a CPU due to overheating. I've killed some by mishandling them and/or foolishly trying to stick one into the wrong motherboard. Well, maybe I melted a CPU or two due to improper heatsink application...I'm not sure.
Optical drives? They never seem to last, with me. Lots of them have gone bad on me, and I really don't know why. They just stop reading and/or writing reliably, being unable to boot up a liveCD.
Now 3.5" drives are an interesting story. I used to kill a LOT of 3.5" drives due to overheating. But this was before I was into quiet computing, and before I had the slightest clue that they needed any cooling at all. But after I had the clue that they needed at least some minimal airflow, not a single 3.5" drive has failed on me. Yet.
So. In my experience, motherboards and power supplies may be the most vulnerable to failure.
Only thing I've ever had die on me through overheating was a graphics card, and that was through my own stupidity. I was trying to flog more horsepower out of a cheap passively-cooled card with a frankly silly overclock. I ostensibly got it running well and was quite pleased with myself, but the card fell over dead within a couple of months
HDD is the first to go, and i have lost a hard drive to heat before. But i dont really think it was my fault. The A/C in our dorms went out for like a month, and it got up over 40c in the room, and the HD got up to around 60c for a prolong period of time.
Since then though i switched to 2.5" drives, which dont overheat. My next drive will be a velociraptor enclosed in a scythe 2.5" quiet drive.
Since then though i switched to 2.5" drives, which dont overheat. My next drive will be a velociraptor enclosed in a scythe 2.5" quiet drive.
In mucking about with old towers from yard sales, I've encountered a few with dead hard drives crammed tight below the floppy drive in the small 3.5 inch bay common in old cases.
Had an IBM 390E notebook that I upgraded into oblivion. Put a PII 366 in place of the original 333, maxed the ram to 2x128. The cpu, ram and hdd were all in close proximity. The machine would lock up after 20-30 minutes. I eventually traced it to one of the sodimms being heat sensitive, but the extensive testing it took to find that out killed the notebook.
Heat kills through poor design.
Had an IBM 390E notebook that I upgraded into oblivion. Put a PII 366 in place of the original 333, maxed the ram to 2x128. The cpu, ram and hdd were all in close proximity. The machine would lock up after 20-30 minutes. I eventually traced it to one of the sodimms being heat sensitive, but the extensive testing it took to find that out killed the notebook.
Heat kills through poor design.
The only component ive ever had die on me due to heat was a HD. Not sure how hot it was.. but it was uncomfortable to touch while sitting in my case.
But overall when I worked at making my components quieter -- they also got cooler. No longer does my cpu run at a steady 70c... 45-50 is where it sits now. HDs now run in the low 30s instead of 40ish.
Not sure about the GPU.
But overall when I worked at making my components quieter -- they also got cooler. No longer does my cpu run at a steady 70c... 45-50 is where it sits now. HDs now run in the low 30s instead of 40ish.
Not sure about the GPU.
-
- Friend of SPCR
- Posts: 5
- Joined: Wed Jul 14, 2004 7:49 am
- Location: MN
-
- Patron of SPCR
- Posts: 744
- Joined: Tue Mar 04, 2008 4:05 am
- Location: London
- Contact:
The closest ive come to disaster was years ago i took off my CPU cooler (stock AMD cooler) while i was giving the system a good clean.
Put everything back in, turned on the system, loaded windows and just loaded up UT2004 when out of the corner of my eye i saw a tube of thermal paste. And UNOPENED tube of thermal paste.
Probably took about a quarter of a second to hit the power switch. Not sure how hot the CPU got, but i do know it wouldnt have lasted too long without any paste on the stock cooler when gaming.
Put everything back in, turned on the system, loaded windows and just loaded up UT2004 when out of the corner of my eye i saw a tube of thermal paste. And UNOPENED tube of thermal paste.
Probably took about a quarter of a second to hit the power switch. Not sure how hot the CPU got, but i do know it wouldnt have lasted too long without any paste on the stock cooler when gaming.
-
- Posts: 87
- Joined: Sun Apr 29, 2007 1:20 am
-
- Posts: 239
- Joined: Sat Mar 29, 2008 7:05 pm
- Location: Toronto Ontario
I "do". Except that I don't consider high temps to be a risk for longevity. Nor do I attribute any of my hardware failures to heat.NeilBlanchard wrote:Nobody here keeps their computer quiet at the risk of longevity.
Before I switched to this computer, I had a computer with Athlon XP 1800+ palomino that idled ~60 Celsius. It also has a ghetto-passive video card, I simply unplugged the fan from the radeon 7500. I fan swapped the DTK psu fan with 80mm Nexus, and tried to silence its god-awful 80GB Samgung V40 hard drive by using double-sided tape to wrap it in foam and glue it on the bottom of the case. It's still working after 6 years. I'm most surprised that the cheap psu is still working after the fan swap. I was kind of expecting it to burn by now, but somehow it just keeps on working.
And to answer to OPs question; No, I don't believe I have killed any hardware with heat. Anything below 75 C is fine and dandy (120 C for graphics).
dont know about that. a friend of mine bought a PC from some online company that manufacturers and sells them.FartingBob wrote:The closest ive come to disaster was years ago i took off my CPU cooler (stock AMD cooler) while i was giving the system a good clean.
Put everything back in, turned on the system, loaded windows and just loaded up UT2004 when out of the corner of my eye i saw a tube of thermal paste. And UNOPENED tube of thermal paste.
Probably took about a quarter of a second to hit the power switch. Not sure how hot the CPU got, but i do know it wouldnt have lasted too long without any paste on the stock cooler when gaming.
decided to try overclocking it (c2d e6600), but couldnt OC it at all. i asked him about temps, turns out it was idling at 60C. i told him to check the mounting of the heatsink.
turns out they had shipped the computer out without applying thermal paste.