|
Hi,
I've had this problem for a couple of months where my computer would just spontaneously shut down and restart without posting, requiring me to restart it manually another time. Sometimes it will happen 1-3 times in a day, sometimes nothing happens for 2 weeks or so. It has now been 1 week since the last event.
I have a feeling this could be caused by my power supply, Corsair 500w builder series. I read a lot of review about this PSU, and noticed that it wasn't held in as high regard as the previous Corsair PSUs had been, because the warranty was 2 years instead of 5, and the parts have been made in China instead of Japan. Plus, much more people were reporting problems with it in comparison to the older Corsair PSUs.
I sometimes hear a small vibrating and whistling sounds from inside my case, but I haven't been able to confirm if it comes from my PSU. However, there is something peculiar — every time when there is this whistling sound, when I touch my computer case, the sound goes off, and goes back on again when I stop touching it. I suspect this could have something to do with static electricity.
Note that the whistling sound isn't always there, only half of the time I use my computer.
|
It's desktop right? In that case maybe you're fan(s) are broken, or not connected. Or maybe you just got so much dust inside the computer that it's making you're whole computer warm(which makes it shut down) Also, try clean you're computer for dust with a vacuum cleaner(CAREFULLY).
|
On September 16 2011 16:41 tonning wrote:
It's desktop right? In that case maybe you're fan(s) are broken, or not connected. Or maybe you just got so much dust inside the computer that it's making you're whole computer warm(which makes it shut down)
There is hardly any dust, because I clean the interior and dust filters 1-2 times a month. And if the fans were broken, I would have noticed it already... They're running fine.
|
On September 16 2011 16:43 DNB wrote:Show nested quote +On September 16 2011 16:41 tonning wrote:
It's desktop right? In that case maybe you're fan(s) are broken, or not connected. Or maybe you just got so much dust inside the computer that it's making you're whole computer warm(which makes it shut down) There is hardly any dust, because I clean the interior and dust filters 1-2 times a month. And if the fans were broken, I would have noticed it already... They're running fine.
It might just be that you have disconnected one of you're fans then because this just sounds really like it, because of the shut down which indicates to a warm computer.
|
On September 16 2011 16:48 tonning wrote:Show nested quote +On September 16 2011 16:43 DNB wrote:On September 16 2011 16:41 tonning wrote:
It's desktop right? In that case maybe you're fan(s) are broken, or not connected. Or maybe you just got so much dust inside the computer that it's making you're whole computer warm(which makes it shut down) There is hardly any dust, because I clean the interior and dust filters 1-2 times a month. And if the fans were broken, I would have noticed it already... They're running fine. It might just be that you have disconnected one of you're fans then because this just sounds really like it, because of the shut down which indicates to a warm computer.
Well, let me check, because I have only 2 fans in my case (excluding component fans)
Edit: No fan is disconnected, so overheating is hardly the problem. My coretemp shows <40C temperatures all the time.
|
I have no idea if it's not overheating, but if it is try check ur CPU and if it needs some new thermal paste, last thing to check if its not overheating 
|
How would I check if it needs new thermal paste? Even now coretemp shows 20 celsius in my CPU, and I've only had my PC for half years.
|
On September 16 2011 17:09 DNB wrote:
How would I check if it needs new thermal paste? Even now coretemp shows 20 celsius in my CPU, and I've only had my PC for half years.
Take of the old one and take on some new. It was only the last thing to check for tho. Did you set up you're computer yourself or did you buy a spesific brand?
|
On September 16 2011 17:18 tonning wrote:Show nested quote +On September 16 2011 17:09 DNB wrote:
How would I check if it needs new thermal paste? Even now coretemp shows 20 celsius in my CPU, and I've only had my PC for half years. Take of the old one and take on some new. It was only the last thing to check for tho. Did you set up you're computer yourself or did you buy a spesific brand?
We put it together with my friend who has been doing it before. CPU is Athlon II x3 with the stock cooler, no clocking whatsoever.
By the way, I'm curious to know that if my CPU temperatures are fine, why would the problem be in my thermal paste? The computer can shut down even when there is no strain at all i.e. when it's idle..
|
On September 16 2011 17:22 DNB wrote:Show nested quote +On September 16 2011 17:18 tonning wrote:On September 16 2011 17:09 DNB wrote:
How would I check if it needs new thermal paste? Even now coretemp shows 20 celsius in my CPU, and I've only had my PC for half years. Take of the old one and take on some new. It was only the last thing to check for tho. Did you set up you're computer yourself or did you buy a spesific brand? We put it together with my friend who has been doing it before. CPU is Athlon II x3 with the stock cooler, no clocking whatsoever. By the way, I'm curious to know that if my CPU temperatures are fine, why would the problem be in my thermal paste? The computer can shut down even when there is no strain at all i.e. when it's idle..
Did you test it with a stress test as Prime95? 20 idle and 40 max are really low. It can't go under your room temperatur and 40 is very low unless you have some monster cooling. Run prime95 for a while and see what you temperatures are. I wouldn't change the thermal paste unless you have really high temps.
As for the vibrating sounds, it could be fans, psu or hdd. Make sure all your fans are spinning correctly and if it's the hdd there are different ways to lessen the vibrations. If it's the psu it's harder and probably easier to replace it.
Probably not related to your restarting issue though.
|
On September 16 2011 17:22 DNB wrote:Show nested quote +On September 16 2011 17:18 tonning wrote:On September 16 2011 17:09 DNB wrote:
How would I check if it needs new thermal paste? Even now coretemp shows 20 celsius in my CPU, and I've only had my PC for half years. Take of the old one and take on some new. It was only the last thing to check for tho. Did you set up you're computer yourself or did you buy a spesific brand? We put it together with my friend who has been doing it before. CPU is Athlon II x3 with the stock cooler, no clocking whatsoever. By the way, I'm curious to know that if my CPU temperatures are fine, why would the problem be in my thermal paste? The computer can shut down even when there is no strain at all i.e. when it's idle..
No clocking set, allright.
Without the thermal paste the cpu can overheat but you're cpu is fine, I was just asking to make sure. Maybe you have gotten a trojan horse.
|
Alright I got Prime95 going and now I can hear a silent but constant buzzing voice somewhere near my DVD-case and HDD. Also, when I touch the case, it goes silent.
Also, what kind of a Trojan Horse would cause spontaneous shutdowns? Never heard of those... I would at least imagine they would shut your PC down regularly instead of first twice a day and then a 2 week's pause...
|
And the temperatures? You can use HWmonitor for temperaturs and other readings if you don't have a good program. Could help identify some problems with voltages if it's your psu.
Can you open up the side of your pc? It's easier to identify were the sound are coming from. Prime95 shouldn't stress your hdd so possible a fan issue. For hdds you can try this:
http://www.silentpcreview.com/forums/viewtopic.php?t=8240
If it's a fan, turn off your computer and clean them off first of all, preferably with compressed air. Make sure all screws are secured and that no cables are obstructing them from spinning. Check both cpu and case fans. Some stock cpu coolers using clips have been known to losen somewhat.
That might solve your sounds issue, but unless your temperatures are incorrect it shouldn't help with the restarts unfortunatly.
Could you list your full computer specs? Random restarts can be a lot of things so you just have to go through everything. It's not a bad idea to run a anti-virus/malware program, test you memory (memtest86) and so on. If you can borrow a psu and test it, it would eliminate that factor.
|
On September 16 2011 17:59 gruff wrote:And the temperatures? You can use HWmonitor for temperaturs and other readings if you don't have a good program. Could help identify some problems with voltages if it's your psu. Can you open up the side of your pc? It's easier to identify were the sound are coming from. Prime95 shouldn't stress your hdd so possible a fan issue. For hdds you can try this: http://www.silentpcreview.com/forums/viewtopic.php?t=8240If it's a fan, turn off your computer and clean them off first of all, preferably with compressed air. Make sure all screws are secured and that no cables are obstructing them from spinning. Check both cpu and case fans. Some stock cpu coolers using clips have been known to losen somewhat. That might solve your sounds issue, but unless your temperatures are incorrect it shouldn't help with the restarts unfortunatly. Could you list your full computer specs?
My specs are:
GPU: Radeon HD 5850 xtreme
CPU: Athlon II x3 445
PSU: Corsair CX500
HDD: Some Samsung 500GB one
Motherboard: Asus M4A77TD
Memory: Corsair 4GB 1600 or 1333 (not sure)
DVD player: Some standard LG one
I'm running Prime95 along with HWmonitor, here is a pic of the temps and voltages, someone more knowledgeable could tell if there's anything peculiar with them.. :
![[image loading]](http://i.imgur.com/UjrkE.png)
|
You can check in the bios what memory timings and frequency you have or download something like CPU-z. As I said it can be a lot of things and you have to eliminate the different factors.
1) Virus/malware
2) Reinstall all device drivers and maybe remove programs you installed when the problem started to happen. Alternative making a clean install of windows, but I'd save that if nothing else works.
3) Make sure temperatures and voltages are in an acceptable range during load. Check you memory (memtest86), run some disk diagnostic tool, check you gpu temperaturs, especially if it restarts during games.
5) Clean your computer and make sure ever everything sits well. Possibly re-seat your hardware.
6) Try getting a hold of a second psu and see if it works with that one. Same with the motherboard.
I'm sure there are more things that you can try that others might fill in.
Edit: Your temperaturs looks fine. The +5 v line looks strange though at 3.66. I'm no expert on voltages and HWmonitor so I suppose it could be a software issue or some sort of regulation. Given that you have these problems though it's quite suspect.
|
On September 16 2011 18:26 gruff wrote:
You can check in the bios what memory timings and frequency you have or download something like CPU-z. As I said it can be a lot of things and you have to eliminate the different factors.
1) Virus/malware
2) Reinstall all device drivers and maybe remove programs you installed when the problem started to happen. Alternative making a clean install of windows, but I'd save that if nothing else works.
3) Make sure temperatures and voltages are in an acceptable range during load. Check you memory (memtest86), run some disk diagnostic tool, check you gpu temperaturs, especially if it restarts during games.
5) Clean your computer and make sure ever everything sits well. Possibly re-seat your hardware.
6) Try getting a hold of a second psu and see if it works with that one. Same with the motherboard.
I'm sure there are more things that you can try that others might fill in.
Edit: Your temperaturs looks fine. The +5 v line looks strange though at 3.66. I'm no expert on voltages and HWmonitor so I suppose it could be a software issue or some sort of regulation. Given that you have these problems though it's quite suspect.
I have absolutely no idea about voltages, how they work in a computer and what they indicate in case they are off, like the 3.66 you mentioned.
With viruses I have tried my best by scanning with spybot and adaware regularly.
|
I do believe after thinking about if for a moment that the +3.3 and +5 "share" their total load. Do the +5 value go up to 5 V if you turn off prime95?
Sorry I can't be of more assistance. I don't want to mislead you into buying something you don't need. Hopefully there are others that can interpet the results. You do want all rails to deliver what they say though. +12 should be 12, +5 be 5 and so on.
|
What you're hearing could be coil whine as well. As for where does that come from, I'd wager on the PSU, that was the case here (well, the reason, actually). Check w/ a friend's one.
|
On September 16 2011 18:55 gruff wrote:
I do believe after thinking about if for a moment that the +3.3 and +5 "share" what their total load. Do the +5 value go up to 5 V if you turn off prime95?
Sorry I can't be of more assistance. I don't want to mislead you into buying something you don't need. Hopefully there are others that can interpet the results. You do want all rails to deliver what they say though. +12 should be 12, +5 be 5 and so on.
Actually, the +5 value is only 2.5V when the computer is idle. I have little knowledge about PSU's. Could one set the right voltages in BIOS should they be imbalanced like right now?
|
The bios settings are for the cpu, memory and such, not the psu. It could be that HWmonitor is simply showing an incorrect value but there should not be any reason (that I'm aware of at least) why it would be at 2.5 V in idle. I think to be certain you'd have to measure the actual voltage with a multimeter (or whatever it's called in english) on the psu. Easiest would be if you had another psu to try with to see if that fix your restart issues. Think about it like you would about a power outlet in your wall. The devices you use are specifically made to use and tolerate a certain voltage so you'd expect to always have the same delivered. If it's not then you can't expect things working as they should. Though I'm not certain how a low voltage on the +5 rail would affect the computer (I think it's mostly for usbports, ps/2, floppy, pci ports) I can't imagine it being good if the readings are correct.
Maybe try contacting corsair support about the psu and the readings and see what they say. Even if the warranty is out they might be able to help you, I've had good experiences with them before. Dunno. Or wait until the real TL tech forum geniuses wakes up.
|
|
|
HWMonitor's Voltage readings should not be trusted. This isn't the first of many times I've seen wacky 3.3V, 5V, and 12V readings (on this forum, myself included, just do a google search you'll see).
Similarly, TMPIN0 and TMPIN1 are grossly wrong or are attempting to read sensors that don't even exist (this is what happens on my pc).
We need some of the other Hardware experts around here to weigh in. It could be the PSU, but HWMonitor is NOT going to tell you that with any degree of certainty.
P.S. If it was really 15V on the 12V rail the PC probably wouldn't even start, much worse things would happen.
|
Well yeah, if his computer was running fine I would probably say it's HWmonitor showing faulty values (my mobo shows 200 C in HW...) but it feels suspect as a failing psu would fit with his symptoms. Which is why I suggested checking with other software to see what they show and preferably checking with multimeter to be certain. Random restarts is a bitch since it can be numerous of other things that causes it.
Personally I would like to write off that possibility before going onto the the next thing.
|
Check the BIOS hardware monitor page. It won't tell you load voltages, but it will tell you whether HWMonitor is showing garbage readings. Most of those are definitely garbage though.
Even if the voltages are fine, it doesn't prove that the PSU isn't causing the problem, but there are a lot of software possibilities here anyway.
|
PSU or GPU, Have you Stress tested your GPU yet ? it may take even longer stress test (i.e 40 minute match in a heavy map with full effects such as shattered temple, or a program called furmark)
|
On September 17 2011 04:42 kuRdRone wrote:
PSU or GPU, Have you Stress tested your GPU yet ? it may take even longer stress test (i.e 40 minute match in a heavy map with full effects such as shattered temple, or a program called furmark)
SC2 is incapable of stressing anything with more power than an HTPC GPU.
|
That's what the furmark is for, what i am refering to with the SC 2 is a broken memory on the GPU
(sorry for english skills- incapable to explain very well)
|
If Furmark doesn't find it, SC2 won't either. And if SC2 is doing screwy stuff with the VRAM, it's most likely software, not hardware.
Anyway, it's happening at random, which doesn't usually imply GPU. That's usually patterned.
|
Yep, I'm pretty sure it's not a GPU problem since it can shutdown even when idle.
I went to the BIOS to see if I could find the voltages and temperatures, but they weren't there. How would I know the actual temperatures and voltages in case coretemp/Hwmonitor or other programs show false values?
|
On September 18 2011 20:28 DNB wrote:
I went to the BIOS to see if I could find the voltages and temperatures, but they weren't there. How would I know the actual temperatures and voltages in case coretemp/Hwmonitor or other programs show false values?
BIOS display is supposed to be Power->HW Monitor Configuration for that board, although if there's no actual display there you could try installing Asus PC Probe II.
None of these are guaranteed to give correct values though. For voltages you can use a multimeter.
|
|
|
|
|
|
EPT
