When i play GW for 15min to an hour, my computer will eventually shut itself off. i'm thinking it's a overheating problem, but i'll give you my comp specs anyway, to see if anything is lacking.
intel P4 2.81Ghz AGP 8x
1024MB duel channel DDR RAM
SB Audigy Gamer 5.1
GeForce 4 ti 4200 128MB
80GB hard drive 7,200rpm
330W (i think) power supply
1 fan, still one fan slot free at the back
if you have any suggestions as to what it might be, or how i can reduce my cpu and overall computer temperature, that'd be great. thanks
EDIT:
my motherboard is a gigabyte GA-8S655FX
Help, comp shutting itself off..
dstock
Darkmane
1. Open up your pc case and make sure its not clogged with dust. (in the heatsinks for the cpu and gpu (graphics card) If it is.. get yourself a can of air and go to town. (be sure not to hold the can too close to components)
2. Grab a copy of Motherboard Monitor, I did not see that you stated what motherboard you have... but motherboard monitor should be able to give you the temperature readings of your cpu. If it happens on a regualar basis.. you can check the temp before it normally drops you.
3. If its your GPU (graphics processor) see if the drivers offer a GPU temperature reading you can also put on your desktop.
Lets determine if it is heat first. Before we go too far )
2. Grab a copy of Motherboard Monitor, I did not see that you stated what motherboard you have... but motherboard monitor should be able to give you the temperature readings of your cpu. If it happens on a regualar basis.. you can check the temp before it normally drops you.
3. If its your GPU (graphics processor) see if the drivers offer a GPU temperature reading you can also put on your desktop.
Lets determine if it is heat first. Before we go too far )
dstock
alright i'll try, thanks, and my motherboard is listed above in the edit
Darkmane
According to the MBM support list, your motherboard is supported.
GA8 S655FX(-L) SIS950/ITE8705F
SIS950/ITE8705F 1
SIS950/ITE8705F 2
http://mbm.livewiredev.com/
Hope this helps.
GA8 S655FX(-L) SIS950/ITE8705F
SIS950/ITE8705F 1
SIS950/ITE8705F 2
http://mbm.livewiredev.com/
Hope this helps.
Lunarbunny
Most likely a heating problem, and I don't see any real power hogs there, so I think 330W is fine. (Otherwise I might recommend a better PSU)
I overkilled on my cooling
. I ought to put a diagram on how much air is coming through (although a slot fan for my video card would be nice)
If you are overheating, or you're not sure, run the computer with the case open. It will let a lot more air in, so you will be able to diagnose if inadequate cooling is the problem.
---
My mobo came with simple monitoring software, but that's just me.
I overkilled on my cooling
. I ought to put a diagram on how much air is coming through (although a slot fan for my video card would be nice)If you are overheating, or you're not sure, run the computer with the case open. It will let a lot more air in, so you will be able to diagnose if inadequate cooling is the problem.
---
My mobo came with simple monitoring software, but that's just me.
Lifire
The watts do not matter for a PSU. If it's not a quality PSU, it can easily cause random reboots. I had a 500 watt POS and it was only rated at an actual 300. What it says isn't what it is unless you're buying name brand (enermax and antec, for example).
Lunarbunny
True. I got lucky and got a good one with my case, but sometimes they'll hand off PSU's as higher wattage, but if you read the small print on a cheap B.S. PSU they'll say something like "Do not place loads of higher than xxxW" which is lower than the supposed wattage.
---
The old Compaq I have has a PSU that states "145W" but in the fine print it says no more than 90W.
---
The old Compaq I have has a PSU that states "145W" but in the fine print it says no more than 90W.
dstock
anything else i could do to maintain a cool temperature?
dstock
i got MBM, and doing normal things, not gaming, my CPU temperature jumps around from 21 C all the way to 64 or 70 C, while my case is around 25-40 C. This is with the case closed. what can i do to get these temps down. also, i think when playing GW, my vid card transfers some of the work to the processor, and that's why my comp shuts itself off. computers only shut themselves off when it's the processor overheating, not the vidcard. so what can i do?
Lunarbunny
Maybe the heat sink is not contacting properly? Make sure it's on the socket the correct direction (I put mine on backwards once a couple years ago ). If everything seems ok with the heatsink, you may want to try getting some thermal grease to increase the contact with the heatsink. If that reading is true, that's one hell of an increase 
. It's possible that your thermometer is whacked, but let's assume that it's correct ATM.
You could try opening the case and running a desk fan at it if you have one.
There are two temperature scenarios that automatically shut off your computer:
Rapid heat increase for an certain amount of time
or
Temperature threshold (I think the MSI mobos are set to something like 93 degrees C. Not sure about yours, but it probably has a setting in BIOS setup)
). If everything seems ok with the heatsink, you may want to try getting some thermal grease to increase the contact with the heatsink. If that reading is true, that's one hell of an increase . It's possible that your thermometer is whacked, but let's assume that it's correct ATM.You could try opening the case and running a desk fan at it if you have one.
There are two temperature scenarios that automatically shut off your computer:
Rapid heat increase for an certain amount of time
or
Temperature threshold (I think the MSI mobos are set to something like 93 degrees C. Not sure about yours, but it probably has a setting in BIOS setup)
Sin
I think it's the PSU.
Currently you are marginally powered, sure the rating is "meaningless" but only to a point. Its about total power delivery if you have multiple devices on at once. Now if one those devices is marginal, meaning, using just a slight bit more current than usual, then it will be stealing voltage from something else--usually the next weakest device. So say your graphics card is starting to use more energy via it's fan, well that might make it try to draw from another device, but if your total output is maxed, well it's gonna try to draw on the power supply and the system will just shut down. The 15 minutes to an hour is what makes me think this as it probably is different depending on the ambient temperature in the room, so it would change based on the demands on the fans.
If it were me I'd get a higher rated power supply just to raise the marginal maximum threshhold and see if that solves the problem. It is my understanding that P4 and above processors really require a 400 watt power supply, so I'd get one of those or better.
Currently you are marginally powered, sure the rating is "meaningless" but only to a point. Its about total power delivery if you have multiple devices on at once. Now if one those devices is marginal, meaning, using just a slight bit more current than usual, then it will be stealing voltage from something else--usually the next weakest device. So say your graphics card is starting to use more energy via it's fan, well that might make it try to draw from another device, but if your total output is maxed, well it's gonna try to draw on the power supply and the system will just shut down. The 15 minutes to an hour is what makes me think this as it probably is different depending on the ambient temperature in the room, so it would change based on the demands on the fans.
If it were me I'd get a higher rated power supply just to raise the marginal maximum threshhold and see if that solves the problem. It is my understanding that P4 and above processors really require a 400 watt power supply, so I'd get one of those or better.
Virtuoso
Is it restarting or shutting itself off?
-Virt
-Virt
dstock
Quote:
|
Originally Posted by Virtuoso
Is it restarting or shutting itself off?
-Virt |
kizayaen
Seems that could well provide a solution, although I really hate to tell a person to spend money on hardware they may or may not need. Still, the points raised so far are all good, and the combination of a new power supply and a better ventilated fan may well cause your problem to go away.
Sin
A little something I learned...
You know how people want the air to flow through the case? Well, if you put a front and rear fan set, amazingly it becomes a totally regulated flow, they sort of syncrhonize and it doesn't really adjust to ambient conditions as well. Anyway, to make this short. If you can, forget the front fans, put 3 in back if you have to but don't put anything blocking the input Amazingly you get a higher flow rate. I can try to explain it but just say it has to do with the difference in pushed and pulled air in combination with changes to barometric pressure.
A car will run differently in the mountains than at sea level because of the difference in the air. With fans operating at 6000 rpm this effect is amplified. Barometric pressure can change the throughput and thus the thermal dynamics.
Hopefully this was helpful.
You know how people want the air to flow through the case? Well, if you put a front and rear fan set, amazingly it becomes a totally regulated flow, they sort of syncrhonize and it doesn't really adjust to ambient conditions as well. Anyway, to make this short. If you can, forget the front fans, put 3 in back if you have to but don't put anything blocking the input Amazingly you get a higher flow rate. I can try to explain it but just say it has to do with the difference in pushed and pulled air in combination with changes to barometric pressure.
A car will run differently in the mountains than at sea level because of the difference in the air. With fans operating at 6000 rpm this effect is amplified. Barometric pressure can change the throughput and thus the thermal dynamics.
Hopefully this was helpful.
dstock
how will a 350W p/s hold me out instead of a 420W? i'd save myself about $90 (CDN) if i went with the 350W
Sin
no, man, get a 400 or better, I mean it, you might be okay but 20 watts isn't gonna raise the marginal (minimum) threshhold that much from 330. I woud say you need 5 watts or more for a good marginal increase, so just for sake of having something to work with: 20 watts / 6 devices = 3.33 watts of average "reserve" to go to each device if all were powered up at once. You need at least 5 watts I would say. so 400 (+70 or 10+ watts more per device available) would definitely do the trick. Understand I am only guessing, not some electrical engineer, but it just seems very obvious that this is the problem to me.
Lifire
Quote:
|
Originally Posted by Sin
A little something I learned...
You know how people want the air to flow through the case? Well, if you put a front and rear fan set, amazingly it becomes a totally regulated flow, they sort of syncrhonize and it doesn't really adjust to ambient conditions as well. Anyway, to make this short. If you can, forget the front fans, put 3 in back if you have to but don't put anything blocking the input Amazingly you get a higher flow rate. I can try to explain it but just say it has to do with the difference in pushed and pulled air in combination with changes to barometric pressure. A car will run differently in the mountains than at sea level because of the difference in the air. With fans operating at 6000 rpm this effect is amplified. Barometric pressure can change the throughput and thus the thermal dynamics. Hopefully this was helpful. |
Sin
That's cool, I just shared what i discovered with the way the flow works. It seems the fans draw easier than they blow air as far as pulling through a slot. We lowered the temperature by 7 degrees just removing the front fan. I didn't expect it either. It could be case specific but it made sense when you thought about a room with warm air pumping in to it the air has to evacuate at a similar rate for it not to acummulate in increasing temperature, if it evacuates at a faster rate it actually could contribute to the flow, but wait it can't as the air is pumped in, see? In a closed environment like that the flow is going to go toward the exhausting fan and so long as no restriction the air will leave as quickly as the fan will allow. The larger fan in front was restricting the flow. Every system I've done this to (4) the result has been 5 - 7 degree temperature reduction. They don't have top fans though so that could make a big difference obviously. Yea I tend to rely on hands-on experience above anything else.
And I only do this when there appears to be a temperature problem. I just mentioned it to this guy cause he was thinking he might have a temperature problem. I tend to believe the whole thing is the power supply, in fact the PSU may be working "marginally."
And I only do this when there appears to be a temperature problem. I just mentioned it to this guy cause he was thinking he might have a temperature problem. I tend to believe the whole thing is the power supply, in fact the PSU may be working "marginally."
dstock
with the temperature monitoring program i downloaded, it says my cpu and case temperatures are changing as much as 40 degrees in 5 seconds, back anf forth. is this ea problem with my motherboard temperature sensor? if so, what can i do about it?
Zeppo
I am actually thinking its a power problem. The same sort of thing was going on with my friends computer. HE would play a game for 5 - 15 minutes, and then it would restart. It ended up being the power supply. He upgraded and it hasn't happened since. So my guess would be power supply.
dstock
Quote:
|
Originally Posted by dstock
with the temperature monitoring program i downloaded, it says my cpu and case temperatures are changing as much as 40 degrees in 5 seconds, back anf forth. is this ea problem with my motherboard temperature sensor? if so, what can i do about it?
|
Darkmane
I am not ready to say its the power supply yet... although of course it is a possibility. What kind of Heatsink / fan do you have on your processor? Did you put it together yourself? Is there a fan on the northbridge or just a heatsink ... bleh let me go look it up. bbs. I agree with checking out the heat issue by taking the side panel off and stickin a fan near it to see if it stays on. IMO if it was the power supply it would have totally failed on you by now with all the shutdown and restarts.
Sin
Right back to the power supply regarding the temperature changes. Reason? The power supply isn't able to sustain power accurately so any number of things can be going wrong, including sensors not working properly--say you have a fan that is starting to wear, therefore drawing more current. That's current that isn't available to another device, like maybe a sensor? If you fans were sensor driven this would look like the sensors were bad, when it's not they or the fan, it's the power supply having so little reserves to buffer currency demand fluctuations. Your computer will also perform better just in having more available when the CPU needs more power.
First get the new power supply and that will rule this out entirely. Why? If you need a new video card for instance, you'll probably need more power too. More ram? More power too. You see? Start with the most essential component to having a computer: The power supply.
I am sure people have had similar expriences with their car when the battery or alternator was acting up. Please change the weakest component in your system pursuant to specs. Again: Intel and AMD recommend having a 400 Watt power supply for todays machines. Please heed this and your problems may be gone entirely.
First get the new power supply and that will rule this out entirely. Why? If you need a new video card for instance, you'll probably need more power too. More ram? More power too. You see? Start with the most essential component to having a computer: The power supply.
I am sure people have had similar expriences with their car when the battery or alternator was acting up. Please change the weakest component in your system pursuant to specs. Again: Intel and AMD recommend having a 400 Watt power supply for todays machines. Please heed this and your problems may be gone entirely.
dstock
ok i have a 420W power supply that i could buy, and i probably will. but to sidetrack from power for a moment. my MBM temperatures for my cpu, then case, have been fluxuating extremely, for instance both going from 5 degrees celcius to 61 degrees celcius in 5 seconds. and i have the side of my computer open while these fluxuations are occuring, and i feel no difference. this leads me to believe that my motherboard TEMPERATURE SENSOR, or thermometer, whichever you call it, is faulty. and if my temperature sonsor is faulty, that could be why my computer shut itself off an unpredictable intervals while playing GW. if the tempterature sensor is awry, what can i do to fix it? new motherboard? or is the broken temperature sensor repairable?
Sin
Alright. I don't know about repairable, it may be bypassable but probably not good to do that. Again it might just be power fluctuation that is causing the sensor to be cooky. I don't know of it being all that common for a temp sensor to get messed up. Lansing Kai Don would know this one for sure. If you get the new PS and still have a problem I'd consider it. I just never seen one of those go bad before and what you describe is classic power supply failure for my experience.
Ask Lansing Kai Don, Mss Drzzt, Loviatar, SSE4 or even ZennZero might have some ideas, as well as a bunch of other people who frequent the tech corner. They'll likely post if they see anything they can add or help you with.
Ask Lansing Kai Don, Mss Drzzt, Loviatar, SSE4 or even ZennZero might have some ideas, as well as a bunch of other people who frequent the tech corner. They'll likely post if they see anything they can add or help you with.
Lunarbunny
There is a band-aid solution if the thermometer is on the fritz. It's slightly dangerous, as your processor will obviously fry if it gets too hot. There should be a setting in your CMOS setup that has a temperature threshold. You can either turn it off if allowed, or turn it up to some absurd number.
To be safe, I'd say switch the PSU first. Much less risk than doing some of these things, and it will most likely be a long-term investment.
To be safe, I'd say switch the PSU first. Much less risk than doing some of these things, and it will most likely be a long-term investment.
dstock
it could be power failure? hmm ok, well i'll try and get that 420W asap and see what happens then, it's just odd. the temperatures go below room temperature, a few degrees above freezing, which is impossible. i never would have thought that could be from p/s failure. the readings are all over tha map. plus, my motherboard's only a year old, and it'd be hard to believe the temp sensor is messed up already. should i also look into that case with all the fan slots, but keep the front ones empty like you suggested sin?
Lunarbunny
Sorry about posting so fast again, but an edit seems silly now.
Any component that's not getting the proper voltage/amperage continuously will most likely go on the fritz. That may be what's happening.
Any component that's not getting the proper voltage/amperage continuously will most likely go on the fritz. That may be what's happening.
Sin
Dude a good case with good ventallation always is a good thing. Put the front fans in if you want. I only take the front one out with trouble situations and that's been 4 times not like 100. Like I said above, I mentioned it here if yours was one of these. I don't think it is. Get what you can and deck it out, why not? hehehe
The key here is not do too much at once or you'll almost never figure out where the problem is. Lunar has an idea there with adjusting the temp threshhold, if you have to. Let's make sure you got clean, abundant power first. That might be the whole thing.
The key here is not do too much at once or you'll almost never figure out where the problem is. Lunar has an idea there with adjusting the temp threshhold, if you have to. Let's make sure you got clean, abundant power first. That might be the whole thing.
dstock
alright thanks, i'll have the power spply by monday, and then i'll see what happens, and probably come back for more suggestions and questions. thanks alot guys
Sin
I hope that's all it is. Make sure to blow out the dust bunnies while you got it all apart too. Good luck with it man! 
dstock
thanks again, and 1 more quick question, with the case, it comes with 2 front fans, 1 side, 1 top, and 2 empty slots at the back. should i fill those slots for a total of 6 fans, or move the 2 front ones to the 2 back slots??
Sin
There should be some info on the case you are buying showing the flow they designed it for. That is called thermal dynamics. I'd follow their recommentations unless they prove to be inaccurate.
Darkmane
Accoring to the manual for your motherboard, there is a section in your bios that will inform you of possible problems with your power supply. Didja check that section out?
I am not saying that the bios will tell you for sure. And many times the bios has a problem determining actual voltage. But it is something to look at.
I am not saying that the bios will tell you for sure. And many times the bios has a problem determining actual voltage. But it is something to look at.
Sin
There may be something to that, the difficulty is in the bios being supplied from the same faulty power supply. Meaning what it tests may be different than just a marginal draw beyond maximum. That's a nice feature, dont' get me wrong, in fact it will be far more useful with a better power supply in the system. I am just hoping this will clear up the temperature sensor problem he is having. The motherboard is 1 year old. If this doesnt' fix it, he's likely gonna need a new mother board, which something like this doesn't usually occur but with dirty or demanding power draws in the first place.
Mss Drizzt
If you are getting those kind of readings you really must I mean must take the Heatsink of and wipe it clean and put on some new Artic silver 2 paste and reaply. Also get a new fan for the heat sink. It will cost you very little.
Then check the Psu. As above mentioned will be cheap and definetly will not hurt. Also you may need a new battery for your mb.
You can not fix the mb prob unless you really know how to solder and are prepared to take a very big gamble.
Then check the Psu. As above mentioned will be cheap and definetly will not hurt. Also you may need a new battery for your mb.
You can not fix the mb prob unless you really know how to solder and are prepared to take a very big gamble.
Darkmane
Quote:
|
Originally Posted by Mss Drizzt
If you are getting those kind of readings you really must I mean must take the Heatsink of and wipe it clean and put on some new Artic silver 2 paste and reaply. Also get a new fan for the heat sink. It will cost you very little.
Then check the Psu. As above mentioned will be cheap and definetly will not hurt. Also you may need a new battery for your mb. You can not fix the mb prob unless you really know how to solder and are prepared to take a very big gamble. |
Mss Drizzt
Quote:
|
Originally Posted by Darkmane
Well I wanted to say this because personally I think that is what is the problem. Poorly mounted HSF..... IMO. But since everyone is hell bent on possible power supply. I just don't think it is because 'normally' Intel based mobos do not have a problem with power. As they use much lower voltages than AMD. It could be that the HSF just sucks. (HSF could be warped) I lap every HSF for every system I build. I have seen some pretty warped HS's from reputatable vendors. Anyway.. thats why I thought we should determinne if it IS heat .... and powr supply later.
|
I believe you are correct. That mb and cpu do not have a big power draw so it would be much better to check the HSF than to spend money you don't have to. I'm a big beliver of cheap first expensive second. And since reaplying the HSF to the CPU will only help try it first........oh yessss.
dstock
a motherboard battery? whats that?