Stress Testing Individual Components?
DY357LX
Playing since day 1. Still can't Comm.England Join Date: 2002-10-27 Member: 1651Members, Constellation
Playing since day 1. Still can't Comm.England Join Date: 2002-10-27 Member: 1651Members, Constellation
in Off-Topic
<div class="IPBDescription">Urge To Kick Pc Rising.</div> Over the past 2/3 weeks my pc has been randomly restarting itself.
Last night I was enjoying a game of CS: S with some friends and it
restarted again prompting me to decide "i've had enough".
So I backed-up my important files and formated the drive this morning.
After getting all the cricital updates, DirectX, ATI Drivers (4.12) etc
I re-installed Steam and whilst waiting for it to fully update I decided
to go play a level or 2 on Black Hawk Down....... 2 minutes into the first
level and my damn pc restarts itself.
It's driving me crazy because I can't figure out whats causing it.
I've ran with onboard (removing the SoundBlaster Live) and it sounds
the same but will still randomly restart.
Is there a way of stress testing each component seperately so I
can narrow down where the problem is stemming from and replace it?
Last night I was enjoying a game of CS: S with some friends and it
restarted again prompting me to decide "i've had enough".
So I backed-up my important files and formated the drive this morning.
After getting all the cricital updates, DirectX, ATI Drivers (4.12) etc
I re-installed Steam and whilst waiting for it to fully update I decided
to go play a level or 2 on Black Hawk Down....... 2 minutes into the first
level and my damn pc restarts itself.
It's driving me crazy because I can't figure out whats causing it.
I've ran with onboard (removing the SoundBlaster Live) and it sounds
the same but will still randomly restart.
Is there a way of stress testing each component seperately so I
can narrow down where the problem is stemming from and replace it?
Comments
1) Your power supply is inadequate for your system.
2) Your memory is bad.
3) Your computer is overheating.
1) Your power supply is inadequate for your system.
2) Your memory is bad.
3) Your computer is overheating. <!--QuoteEnd--> </td></tr></table><div class='postcolor'> <!--QuoteEEnd-->
The Power Supply is fairly new and is 480w GoldenSilent
with a 12 cm fan on it.
I'll try running on 1 stick of ram and alternative the stick
every now and then.
All the temperatures seem to be normal, there's 2 fans in
there (3 including the one on the PSU). It's a very slight
possibility because the system has been running with the
side panel off for weeks now and it's only 5 degree's in
this part of the UK at the moment. (Not sure of room temp.)
According to the LCD display on the front of my pc, the CPU
Temperature is at 21 degree's C.
When my PC kept crashing, however, I found out that it was the RAM by using <a href='http://www.memtest86.com/' target='_blank'>memtest86</a>. It's a nice and simple program, so give it a whirl.
There's some good news and some bad news.
Good: I've greatly narrowed down the source of my pc problems.
The RAM is corrupt. I ran the RAM tester and it resulted in 28 errors!
Bad: Well...er.. I kinda mixed that in with the "Good" but now I need
to sell a kidney to buy some ram. But at least now I can buy some
DDR 400. (Thats pc 3200 right?)
Thanks for the posts guys. Much appreciated.
Yes.
Bad: Well...er.. I kinda mixed that in with the "Good" but now I need
to sell a kidney to buy some ram. But at least now I can buy some
DDR 400. (Thats pc 3200 right?)<!--QuoteEnd--></td></tr></table><div class='postcolor'><!--QuoteEEnd-->
More good news: Don't whip out the cash just yet.
Does your bios allow you to change memory timings and memory/memory controller performance options of less descript character(e.g. performance mode: default/standard/turbo/leetshizzle etc.)? The first thing I would do is to take a look at what can be done to improve stabillity as it is. Some RAM simply doesn't like particular motherboards wery well and you may have to loosen the settings a bit.
Things you can do:
Up the DRAM timings to something nice and loose, like 3-8-8-10. Run memtest86, did it help any? If it did, try tighter timings until you start getting memory errors again, try and find the points where you start getting errors and back off a bit. When you think you got it, run memtest86 overnight or something to really make sure.
If you have very low latency modules some boards memory controllers will just not like it. in particular, 2-2-2... timings, try 2-3-2 and see if that works any better.
Make sure that you are not running leetshizzle/ultra/turbo/whatever mode which on many systems that have this sort of (usually memory latency/memory controller latency enhancing) feature is not stable with most sticks of RAM. Back down to standard or default or whatever and see if it helps.
You can try upping the DRAM('memory chip voltage') voltage a notch or two and see if that helps.
If you at any time do anything stupid in bios that causes your computer not to boot(setting tow low latencies, too high/low FSB speeds or whatever) then you will need to either:
if your motherboard supports it shut of all power(pull wall socket plug/flip offswitch on the PSU), wait for a few seconds and then boot the system up, it will beep tauntingly at you and tell you that 'OVERCLOKING FAILED OR CPU IS UNWORKABLE. DEFAULT SETTINGS RELOADED' or some similar amusing error message typically in all caps engrish;
alternately, you will have to shut of power to the motherboard(wall plug/PSU switch), pull out your motherboard manual and locate the jumper for clearing CMOS. Cycle power and reseat the CMOS jumper into it's original position. Some motherboard manufacturers love to put fan headers right next to the clear CMOS jumper, I've even seen fan headers on BOTH sides with even spacing between all pins(9 all in all in a neat row), just 'SYS FAN#' or a 'J#'(in the old days, jumpers were seldom marked with a usefull inscription like 'CLR CMOS') in wicked tiny print next to each set of 3 pins. In general just don't open your computer to do anything when you are tired or significantly distracted.
More bad news: The cause of those memory errors is likely your RAM but it may not be, since the processor and memory controller(on the motherboard if not an athlon64/opteron, integrated on the processor if it is) are greatly involved in memory access. These in turn are affected by your supply voltage, which is affected by both PSU and motherboard voltage regulation. If a PSU is telling you it is x watts of pure leet power, there is usually fine print somewhere telling you the bad news, that this is only under the artifical condition that you sum maximum power from each power line when used alone, which is off course pointless because that will never happen and your real maximum power is significantly lower; or that the maximum power only applies when the PSU is at near ambient temperatures, and in a computer case it will take probably take in ~25-40 degree celcius hot air from above the processor and graphics card and expell 40-60C hot air so that's far from attainable in the real world; It may also be a peak power usage, sustainable for only short periods of time; it may also be an accurate statement, but that doesn't prevent the PSU from having horrible voltage regulation or you recieving an unusually bad unit. The electrolytic capacitors may also 'leak' in some PSUs and on some motherboards, this will cause their capacitance to become significantly lower and they can't filter the current within acceptable toleranses any more.
The processor might cause these problems by overheating, by not being able to run at the speed it is currently running at(this can be caused by many things, most of which are improbable, a bad CPU getting through quality assurance(pressumably rare), voltage spikes in the line current have caused it some damage(very unlikely if there is a line conditioner and proper grounded socket), the CPU might not be recieving a clean, stable DC voltage from the motherboard due to problems with voltage regulation or may get problems due to electromagnetic interference(unlikely even if you have some nasty EMI sources near your computer)). You may live in an area where voltage regulation is bad, the wall socket voltage often dipping below the PSU's tolerances('brown out'), again, unlikely unless you see long periods of crash free computer usage intersperced with short periods of constant crashing.
etc.
SUYF.
Its prob the video card, I had the same exact problem with my geforce fx 5600. My bets are that its corrupted memory.
Go with a saphire radeon 9800 if your looking for something good with a reasonable price, I use it and I've never been happier.