Sign in with
Sign up | Sign in
Your question

my computer drive me mad, plx kindly help

Last response: in Graphics & Displays
Share
April 6, 2005 6:36:57 PM

i hv tried many forums for help and i asked all of my friend that i know is good at computer. but no, the problem isnt solved. therefore i am thinking i need to go to some really pro places like Tom's hardware guide website to seek help. below is my original message
and if u are interested in replies i got. please visit
tweaktown.com forum:
http://forums.tweaktown.com/showthread.php?t=19859
===or===
Steam forum:
http://forums.steampowered.com/forums/showthread.php?s=...

please kindly help me, i am getting mad about this frustrating problem.

**ORGINAL MESSAGE-------:
i know computer better than average computer users but i hv totally no idea wht is wrong with my PC this time....



about a week ago i started to get random crashes as described below. and i thought that its to do with the new driver. i used drivercleaner and re-installed the previous driver, the crash is still there. after many attempts to solve the problem, i decided to format my windows partition and re-in winXP. after a 100% fresh re-install, the crash is STILL there.



that makes me worry about my hardware. however, i think i shall first go through the crash:



1)its random, it occures in games, in browser, in IE, in WMP, and even when the PC is in screensaver mode.

2)when it occures, everything including the cursor freezes. u can only reboot

3) for three to four times, there was a blue screen. once for stop 0x0000007e error and twice for stop 0x0000000a(0x***,0x00000002,0x00000001,0x***) IRQL_NOT_LESS_OR_EQUAL error.searched microsoft.com and with my limited knowledge i can not troubleshoot.

4) for the most recent two crashes, they both occured in the same 3Dgame, the crash caused a loud NOISE in the headphone i was wearing and freaked me out. the noise is like heavy interuption or a ULTRA-HIGH distortion guitar going mad....and no bluescreen , its in blackscreen for the last two crashes.



And for my specs:

AMD64 2800+ 9x200Mhz L2=512

Seagate 80GB SATA 7200 8MB cache

2x256 DDR PC3200 400Mhz non-branded (this i worry the most)

ATI 9550 Extreme 128MB 128bit OC to 9600 by Grandmars(manufacture) -i believe that this card is very stable

Gigabyte nForce3 250 GA-K8NS with AC97 audio and latest BIOS and drivers

400W branded power supply unit

cd/dvd rom and floppy drive



OS:WinXP SP2 with Dx9.0c



my PC used to be very stable and runs tasks smoothly. for once i wanna test the temperature of my cpu under very high workload, i play Half-Life 2, scan virus, BT downlaod, defrag harddisk, play mp3 in WMP all together with the monitoring utility and the temperature raised from 25 to 38, but everything is still running fine and it didnt crash. and the PC is 4 months old and has NEVER crashed unitll a week ago.



please help me to solve this. i have asked eveyone that i know who is good at computer, and i really feel sad about my PC hving such a weird and big problem, and even more sadly, i hv absolutely no idea wht is wrong with it. the crash is getting annoying and i feel helpless....



please share ur opinion/suggestion even if u dont hv much to say, or even telling me wht u think is suspicious. ur help is very much appreciated.



thankyou very much in advance.

**recent note:
i hv used memtest86+ to test my ram, it says "pass 5 error 0". i suppose it means i passed the test 5 times without error, correct me if i am wrong.
April 6, 2005 6:56:24 PM

I highly recommend you trying a new stick of ram, Samsung should do fine, IRQL not less or equal is usually caused by ram, even though i've seen SP2 causing that(on my PC during Rome) So try a SINGLE new stick of ram, and use SP1, and see how that works out for you.

My comp when i first built it was crashing in Rome Total War, and Rome only, so led me to believe maybe next game patch will fix it, but no it didn't, it was SP2, just using SP1 fixed my problem.
April 6, 2005 7:06:16 PM

I had a similar experience that was caused by a generic power supply.

Once I upgraded to a decent power supply, my problems went away and my system became stable as a rock.

What PSU are you running? Name brand? What are the rails rated at (amps)?

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark03: <font color=red>11,506</b>
Related resources
April 6, 2005 7:15:06 PM

the PSU is a local brand in Hong Kong, and its well-trusted by many users. it costs me 25USD and i dont know much about specs, all i got is this: (from box)
model number 400D
+3.3v_____+5V_____+12v_____+-5v_____-12v
20A_______30A_____12A______0.5A______0.5A

standby 5VSB:2A
3.3V&5.5V output:180W
peak power:400W
April 6, 2005 7:16:21 PM

i would love to try but i hvnt got another ram available
and the memtest86+ programme, is it trust-worthy?
April 6, 2005 7:28:01 PM

I was just thinking that it could be incompatibility issues. And that psu's 12V rail is kinda low.
April 6, 2005 7:45:14 PM

I was thinking the same thing. 12a on the 12v doesn't cut it, especially if it's a generic brand.

The easiest way to test is to try a buddy's (better) PSU.

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark03: <font color=red>11,506</b>
April 6, 2005 7:47:36 PM

but as i mentioned, during very high workload(latest 3d games, defrag HDD, virus scan, play music), it didnt crash. and for four months no matter how long i keep my PC on(sometimes few days), it didnt crash. imo the PSU is less likely to cause such problem if compared to ram/hdd/cpu overheat etc.

however, please kindly let me know more about PSU, their specs, and how to determind wht is "enough" and whts not. i am a total noob when it comes to power and OC :p 

and thanks for the quick replies everyone.
April 6, 2005 7:58:24 PM

in uk i know very little ppl who is good at computer(most dont even own/is familiar with one....)

so sadly i dont think changing PSU is possible.
if its possible for me to heavy-load the system(increase in power needed?) to see if there is any problem?

this might be a noob question, but well...i am a noob with PSU and OC....so....forgive my stupid questoin if it happens to be
April 6, 2005 8:03:35 PM

Any component can make a system unstable, and of all components, the PSU is the most likely to die over time.

If it's during high workloads then it's likely either overheating, or the PSU.

If it was memory you'd see it crash all the time. If it was the HDD you'd see it crash all the time.

If your case/CPU temps are good, it's time to take a long, hard look at the PSU. Get some motherboard monitoring software... find out how many volts it's putting out on the 3.3, 5, and 12v rails under load.

I don't know why so many people are stubborn about considering the PSU as a source of problems. They always take it for granted that PSU's are infallible, when the opposite is true.

For a *great* PSU primer, look at Firingsquad's PSU article HERE: <A HREF="http://www.firingsquad.com/guides/power_supply/" target="_new">http://www.firingsquad.com/guides/power_supply/&lt;/A>

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark03: <font color=red>11,506</b>
April 6, 2005 8:07:42 PM

CPU temp(near idle): 25
usual average: around 30
medium load: 35
heavy load: 36~40

Power supplied to :
CPU:5.375V
Core(wht is core?):1.5V
AGP:1.5V
Aux:1.5V

i am using nvidia nTune(nVmonitor) utility
April 6, 2005 8:41:59 PM

Your CPU temps are great, not a CPU overheating issue obviously.

The voltages you need are the 3.3v, 5v, and 12v rail voltages - not the specific component voltages.

If you can't find them in Ntune, install a utility like "motherboard monitor".

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark03: <font color=red>11,506</b>
a b U Graphics card
April 6, 2005 9:23:47 PM

<A HREF="http://www.majorgeeks.com/download.php?det=311" target="_new">MBM5 clicky</A>

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 6, 2005 9:26:34 PM

i am quoting one of the reply from another person:
Holy cow 400 "peak watts" is a VERY GENEROUS rating for such a horrible unit.

If the 3.3v and 5v peak is 180 W and the 12v rail is only 12 A that means you're real peak is 300 watts (peak power).

A = amps = amount of current drawn.

Power = current * voltage. Power for the 12v is 12 V * 12 A = 144 watts.

The specs for that unit might be OK for the system if it were a quality part but it obviously isn't (a quality part would have given an honest rating of 300 Watts or less). Look for a higher quality PSU to put in there...

------------
if its that bad i might consider getting a new unit, is there any brand that u recommend?

and for the voltages, i cant find it in ntune, but i cant find a utility that provides me with such information. would u name one for me please?
a b U Graphics card
April 6, 2005 9:32:00 PM

Sparkle Power (will have FSP in the model number), Fortron Source, Antec and Enermax all make great PSUs. You can typically get the FSP and Sparkle Power PSUs for less than the others. Here is a recent <A HREF="http://www.tomshardware.com/howto/20050228/index.html" target="_new">THG Power Supply Review</A>.

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 6, 2005 10:03:35 PM

someone has another story, quote:
i had this exact same problem before!!

it was a hardware conflict, turned out to be my crappy realtek sound card, replaced it with an audigy, problem solved

unfortunately you are using onboard audio, so buy a decent sound card, disable your onboard audio device in bios, and i bet it goes away

for now to test the theory, try disabling your onboard sound, see if it still locks up, if it does then i could be off, but it is definately a piece of hardware causing this!
--------
wht do u guys think? sorry but please understand that i hv very little surplus to spend on computer at the moment, i will only get a new and decent PSU if i really need to. and plus, i had no problem before so i think i shall consider buying a PSU later. advice needed.
April 6, 2005 10:18:27 PM

average vlaues
+3.3 : 3.28V
+5.0 : 2.91V
+12 : 11.54V
-12 : -3.45 (high:1.63V low: -8.75V)
-5 : 0

since i know little about these values, please tell me if there is anything wrong
thanks for your help :) 
a b U Graphics card
April 6, 2005 11:55:40 PM

Your voltages need to be within 5% of the rating of the rail. Your 5V rail SUCKS ASS! The PSU is your problem. If your 5V line is consistently at those levels then DON"T run the PC until you replace that PSU...you will end up frying components.

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 12:10:48 AM

but the program that i DLed detected my CPU temperature as -10C....the highest value is 1C....while the nTune, at the same time, detected my CPU temp as 28C
a b U Graphics card
April 7, 2005 12:31:00 AM

The temps are probably from you not having the correct sensors selected. I've never had problems with MBM5 misreading the voltages coming from the PSU - especially not that large of a margin. I don't think nTune nTune will detect the voltages on the 3.3V, 5V and 12V lines, but you should be able to monitor the voltages in the BIOS. Monitor the voltages in the BIOS for 5-10 mins and see how much they are fluctuating and get a feel for average V on each line. Be aware of sudden spikes up or down. Let us know what you abserve.

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 10:14:08 AM

okay i hv tried using Ram A, and it crashed, for once.
and i tried to use only Ram B. didnt crash, YET.

everything was fine with Ram A for 4 hours, then the PC crashed. and for Ram B, no problem for up to 9 hours now.

is it too soon to say its the ram thats cuasing the problem?
April 7, 2005 12:14:13 PM

i think i will get a new PSU
2.9 on the +5 is far too low....according all articles i read
a b U Graphics card
April 7, 2005 1:09:47 PM

You NEED to get a new PSU, if 2.9V is the correct reading. Retest RAM A when you get the new PSU...the crashes may just go away. I seriously would <b>NOT</b> run that PC until I had a good PSU installed. You have a <b>serious</b> chance of killing other components in your system and - at a minimum - shortening their lifespan.

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 3:14:23 PM

To see if it's one of your RAM sticks, try running memory testing software.

Here's a link to a memory tester (I've never tried this one, I usually use memtest86... memtest86 is a DOS program, this one's probably easier to use)

<A HREF="http://hcidesign.com/memtest/download.html" target="_new">http://hcidesign.com/memtest/download.html&lt;/A>

If your memory is bad, this program will report an error.

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark03: <font color=red>11,506</b>
a b U Graphics card
April 7, 2005 3:18:24 PM

Even if some of his memory is showing erros at this point it could be due to his PSU...5V rail at 2.91V

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 3:33:34 PM

ok sorry for my stupidness but
i got some new readings for u guys
i am using the Gigabyte nForce3 250 mobo, model number GA-K8NS

okay, here comes the confusing bit.
in MBM5 i changed the "voltage configuration" to "GA-8KNXP" coz there is NO "GA-K8NS" available and it, by looking at it, is the closest option.

using the new config, i get: (all are average values)
+3.3V @ 3.3
+5.0V @ 5.52
+12V @ 12.14

yes i know i might had choosen the wrong config, sorry about that guys. and how can i ensure that the "GA-8KNXP" is the right one?

i went through the help doc in MBM5 after one of my friend telling me that with +5@2.9V. its very lucky for me to even boot the computer.

besides, i ran prime95 for FOUR hours, during the FOUR hours, CPU is constantly at 100% usage. and no, it didnt crash.

so.................any clue ? Ram B has been doing fine for 15 hours now, would u say that is lone enough for a conclusion?
April 7, 2005 3:39:59 PM

just to add information
during the four hours of CPU stress test
CPU temp went up from 25C to 45C
a b U Graphics card
April 7, 2005 3:41:53 PM

15 hours tells me that stick of RAM is good. I'm still concerned by your 5V rail - your average is still outside the 10% range. And if that is the average, then you're going to have spikes that are above that voltage. What are the high/low readings? I still say the first thing I would do is replace that PSU...

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 3:49:45 PM

since i changed it on the fly
the initial 2.9 reading might hv quite a big effect on the average.
i am restarting MBM5 and i will monitor for say....15 mins and post the high/low/avg

however, first of all, how do i ensure the config is RIGHT for my mobo?
April 7, 2005 3:52:58 PM

hmmm....during this time do u mind telling me wht values are expected for the +5V rail? should the difference between high and low be smaller better? and expected high and low please?
a b U Graphics card
April 7, 2005 4:05:29 PM

All of the voltages are supposed to be within 10% of the rated value, but I usually run into stability issues if the variance is greater than 5%. Using the 5% standard here are the numbers:
1. 5V rail should be between 4.75V and 5.25V
2. 3.3V rail should be between 3.15V and 3.45V
3. 12V rail should be between 11.4V and 12.6V


__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 4:08:59 PM

okay, for 16mins of various tasks undergoing
readings: (rail@average[high/low])
+3.3V@3.28[3.31/3.23]
+5.0V@5.40[6.82/2.11]
+12V@12.16[12.22/12.03]
number of readings during the 16 mins: 552

i did various tasks like insert a dvd/browse website/download 3d mark etc during the period.

please tell me how do u find the readings.
April 7, 2005 4:12:49 PM

okay, eveyrthing seems fine, BUT the +5V rail
+3.3V rail has 2% variation to my calculation
+12v rail has 1.8% variation
but the +5V rail has sth wrong
could it be the config?
i am using GA-K8NS and the config is GA-8KNXP
a b U Graphics card
April 7, 2005 4:13:35 PM

That PSU is hosed! Your 5V rail on your PSU is causing your problems.

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 4:24:03 PM

understandind how much the readings can change if a wrong config is selected, i strongly think that the thing we should work on is try to get the config right.

as far as i know, if the +5V rail ever drops down as much as the readings i provided shown, there should be a big problem, either a crash or even damage in my PC components.

and after i changed to some other setting hoping it to work.
the +12V rail has a reading of 7.6 high 9 low 5 avg 7
a b U Graphics card
April 7, 2005 5:00:39 PM

If you don't believe those readings, then do what I said before: reboot, go into BIOS and manually monitor the volts in the system health section.

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 5:10:39 PM

yes i tried it, and i cant find the readings....why is it so? i dont know, i went through every single page, and no, there is no voltage. i went to gigabyte's website for their faq, and no, there isnt an answer.

i found out sth rediculous though. if u want to enable SMART for ur hdd, u will hv to go to bios, and press Ctrl+F1 to enable the "very" advance option in BIOS. and no, its not stated in the BIOS nor the user manual....

all the BIOS says is:
+3.3V-----OK
+12V------OK
nth is mentioned about the others.

i hate gigabyte
April 7, 2005 5:49:50 PM

Too Lazy to look through all the threads here but have you ran the SeaTools to check you hard drive for having issues. I had the same HDD you had with the same problem which later on I got BSDs and then HDD died day later.

<i><font color=red>Only an overclocker can make a computer into a convectional oven.</i></font color=red>
April 7, 2005 5:55:15 PM

Dont trust the software and BIOS readings for your voltage rails for your software/mobo can read them wrong. The best way is to take a voltmeter or multimeter to check the rails that way.

<i><font color=red>Only an overclocker can make a computer into a convectional oven.</i></font color=red>
April 7, 2005 7:46:09 PM

okay things are getting even more weird
i took Ram A out of slot 1

put Ram B into slot 1
crashed at the winxp loading screen

i put Ram A into slot 1
crashed at the winxp loading screen

i put Ram B into slot 2
crashed at/just before the winxp loading screen

i put Ram A into slot3
crashed at/just before the winxp loading screen

i put Ram B into slot3
crashed at/just before the winxp loading screen

i put Ram A in slot 1 and Ram B in slot2
crashed at the winxp loading screen

didnt try any more combinations

so far, the only one will work is Ram A in slot1 with nth else
and Ram B alone in slot 1 once runs fine for 4 hours before crashing
but Ram A alone in slot 1 runs PERFECTLY fine for 20+hours with no crash.

the conclusion is....?
my slot2 and slot3 is screwed and Ram B is faulty as well???
i am not saying that this isnt possible but this is unlikely to happen....wht do u think guys?
April 7, 2005 8:38:05 PM

All of your RAM tests are worthless until you get a working PSU, if your 5v rail is that bad.

The PSU is probably causing the crashing.

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark05: <font color=red>5,275</b>
a b U Graphics card
April 7, 2005 9:38:29 PM

Your system is saying the 3.3V and the 12V rails are OK - not the 5V rail. If you don't believe me about the voltages or trust the software, then do what Add said and break out the multimeter to test the volts. Seems kinda odd that the only rail that is giving you "false" readings in the software is the 5V... :eek: 

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 7, 2005 10:19:05 PM

no if i change the config to sth else
the +12V will drop to low:4.7 and high 8

i will try to get a multimeter asap
a b U Graphics card
April 7, 2005 10:48:50 PM

To be on the safe side I would minimize using the computer until you're sure the power is good...

__________________________________________________
<font color=red>You're a boil on the arse of progress - don't make me squeeze you!</font color=red>
April 8, 2005 12:13:00 PM

but there must be some problem about the ram isnt it
Ram A in slot 1 has NEVER crashed, for up to 35 hours now

i will still get the multimeter but i suspect the ram more.
April 8, 2005 2:26:28 PM

Dude, if the ram is bad it's probably because your PSU has killed it.

For god's sakes, replace that POS PSU before it kills your whole system!

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark05: <font color=red>5,275</b>
April 8, 2005 3:22:05 PM

why not a single body will listen to me?
the MBM5 is outdated and does NOT support my mobo. and detecting my voltage wrongly. plus, using a software to detect voltage is inaccurate anyway.

i got a multimeter today, and the results are as follow:
+3.3v 3.21 +-0.02
+5V 5.17 +-0.01
+12V 12.24 +-0.02
each of them is observed for about two minutes, with the COMPUTER RUNNING.

and about the weird Ram problem, seems that i hv to accept that slot2, slot3 and Ram B died at the same time for no reason. please let me know wht u think.
April 8, 2005 4:13:51 PM

Good, now you can eliminate the possibility of the PSU causing your troubles (although you should probably check it under load just to be sure)

As for as memory slots, yes, I've seen them "go bad". In most cases they're shorting out because of dust and simply need a good cleaning with an air gun, but in rare cases they simply die. It's rare, but it can happen.

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark05: <font color=red>5,275</b>
April 8, 2005 4:26:54 PM

but isnt it extremely rare that 3 rare things happened together? at the same time?

and i might get some new ram to test it
i am still thinking of getting 1GB ram
or should i get one 512 only?
will that be compatable with my current working ram?
both will be DDR400 and PC3200
April 8, 2005 4:45:14 PM

What 3 things, exactly?

________________
<b>Geforce <font color=red>6800 Ultra</b></font color=red>
<b>AthlonXP <font color=red>3200+</b></font color=red> <i>(Barton 2500+ o/c 400 FSB)</i>
<b>3dMark05: <font color=red>5,275</b>
!