strange issues (black screen, blue screen, and freezes)

[Wingo101]

I noticed that your 8800GT is overclocked... I suggest you lower it to the standard 8800GT speed, or even underclock it using RivaTuner to make sure it's not your Video Card

Even though it is factory overclocked, it can still have an issue with bad memory. Also, check your GPU core temperature before & after lowering the clock speeds.

 

[Mussels]

I noticed that your 8800GT is overclocked... I suggest you lower it to the standard 8800GT speed, or even underclock it using RivaTuner to make sure it's not your Video Card

Even though it is factory overclocked, it can still have an issue with bad memory. Also, check your GPU core temperature before & after lowering the clock speeds.

good catch.


Sometimes the drivers conflict with stock OC'd cards, they tweak things to make performance better in the drivers, and it lowers the maximum OC on that card.

its possible that this has happened, and lowered it below the "stock" of that card.

 

[keakar]

START >> control panel >> administrative tools >> event viewer>>

application = programs
system = most of the time is hardware

thanks, here are the errors/warnings i see:

- Google updater failed to start (not working because i disabled it)(appears with every boot up)
- TCP/IP has reached its security limit placed on concurrent TCP connect attemps (appears about once a day sometimes more)
- Unknown error on CMDre 00000001 00000000 00000000 00000001 00000001(appears just once)
- The Lbd boot-start or system-start driver failed to load (appears 4 or 5 times)
- clock isnt syncronised because it cant get a usable time stamp from providers (appears 5 or 6 times)

 

[Mussels]

- Unknown error on CMDre 00000001 00000000 00000000 00000001 00000001(appears just once)
- The Lbd boot-start or system-start driver failed to load (appears 4 or 5 times)
- clock isnt syncronised because it cant get a usable time stamp from providers (appears 5 or 6 times)


^ those three are the worrying three.

How the hell can your clock be off?

 

[Wingo101]

- Unknown error on CMDre 00000001 00000000 00000000 00000001 00000001(appears just once)

That sounds like a hardware issue....

Did you try lowering the clock of the GPU?


Is your system overclocked in any other way? Maybe just reset your BIOS to default settings to make sure.

 

[keakar]

did you notice any fluctuations in your voltages whilst running 3d mark?

no, the voltages stayed steady the whole time

it froze up a few hours after the 3d test and i saw the funny grafic blocks on the screen again so thinking i had a bad driver issue i uninstalled my video driver and went back to the old 182 driver and it seamed ok at first but now its acting up again and i made sure it hasnt updated the driver by itself

-
How the hell can your clock be off?

welcome to my world lol

or should i say my twilight zone nu nu nu nu, nu nu nu nu....


@ wingo101 i am purely running on default auto detect settings (i confirmed they are correct voltages), i dont overclock at all so its not from me changing settings to cause stability issues

as for the video overclock, sounds like its worth a try but rivatuner tells me it wont work with my drivers (i tried it with driver 182, then 184, and then i reinstalled the latest 195 driver) and it still tells me i need a newer version of rivatuner but i have 224c and that IS the latest one isnt it?

rivatuner was showing 256bit g92 512mb video ram so at least i know the video ram is fully showing, not that it means much

i found an EVGA video monitor overclock utility that i can use to monitor and underclock the video card but how do i know what to set it too?

here is where my card is set at as purchased:

core clock - 650
shader clock - 1620
memory clock - 950

and gpu temp is 40c at idle


Edit:
i saw many of the 9800 cards at newegg were set to 550 core clock so i set my core clock to 550 and my shader clock to 1370

i left the memory alone since it was already lower then where most of the 9800 cards were set

 

[keakar]

ok i cant say turning the card down helped or not because most of the latest freeze ups were coming just after starting the computer up cold after being off all night.

the problem is getting it to start off with the new settings on boot up so i know if its the default "superclocked" settings thats causing issues or not.

i tried both the "evga precision" utility and "rivatuner" to change the default settings on the card to see if that changed the defaults but it still keeps starting with the default "superclocked" settings, so how do i change the default clock settings so the card starts with the lower clocks?


latest freeze info:
this morning i got a series of 5 lines or rows of "dashes" (-----) accross the screen with each a different color. they were a yellow row, blue row, purple row, pink row, and a green row.

i could move my mouse all i wanted to but it couldnt interact with anything and the mouse buttons didnt do anything. the screen had a very slight flicker every 2 or 3 seconds like a slow screen refresh might look. ctrl+alt+del did nothing.

at the time all this was going on my mulimeter tested voltages are:

+3.3 = 3.42
+5 = 5.12
+12 = 12.20


i restarted computer and now the computers working fine (as it usually does after a restart) so i still have no clue what the problem is but the grafical errors make me lean towards some sort of video card issue.

the hardest thing for me to understand is if it is a video card issue then how can that cause my computer to freeze up. shouldnt it still work but just not display correctly?

 

[overclocking101]

you know I think it's 1 of 3 things after reading this entire thread.

1: PSU, I looked up the specs and it really is a weak psu man looks nice on paper but it's rated at 550W MAX power and at 20A. and it looks like with your system it's been working around the max for the amount of time you have had it in wich isnt good for a psu.
2: GPU sounds like something could be very wrong with your gpu, can you take a picture of the backside of the gpu, the newer non refrence 8800gt cards were flawed and had serious overheating/ artifact issues (I know I have one and it does it badly the same thing you are describing).
3: your motherboard, maybe the pci-e slot has something wrong and isnt supplying enough power to the gpu (dunno but it is possible) do you have another pci-e slot you can check it in?? or is that a single pci-e slot board

 

[keakar]

you know I think it's 1 of 3 things after reading this entire thread.

1: PSU, I looked up the specs and it really is a weak psu man looks nice on paper but it's rated at 550W MAX power and at 20A. and it looks like with your system it's been working around the max for the amount of time you have had it in wich isnt good for a psu.
2: GPU sounds like something could be very wrong with your gpu, can you take a picture of the backside of the gpu, the newer non refrence 8800gt cards were flawed and had serious overheating/ artifact issues (I know I have one and it does it badly the same thing you are describing).
3: your motherboard, maybe the pci-e slot has something wrong and isnt supplying enough power to the gpu (dunno but it is possible) do you have another pci-e slot you can check it in?? or is that a single pci-e slot board

#1 - i really think the PSU is enough for my system but im no expert so i cant argue this point except to say it NEVER drops voltage when the system is running games or stress testing and is solid as a rock in that reguard. i was under the impression my rig only needed 350-400 watts to handle it so thats why i got a 550 watt psu.

#2 - umm, not sure what your asking for but there is no discoloration or any visual clues on the back of the card. i yanked the stock cooler of as soon as i got it because of the heat issues they had with the 8800 cards so i never ran it with those high temp stock coolers, with the acelero s2 cooler im using the highest it ever recorded the gpu temp was 50c and most of the time it sits at 40c idle. i can put my finger on the back of the card and while its hot its not hot enough where i cant hold my finger on it which is not much of a test but im thinking it would be hotter then that if it was overheating.

#3 - my board doesnt have SLI or crossfire so its only got 1 pci-e x16 slot. (see it here)

 

[Mussels]

1. PSU's weaken over time. getting a PSU that "works fine" at first, it may well get worse later on. From what you're saying its fine for now, but that doesnt mean it cant have erratic behaviour (for example, the 5V SB rail has weakened so using too many USB devices could make it crash randomly - i've seen that before)

2. GPU temp is one of many temps. is the REST of the card still as cool, or cooler, than the stock cooler would have it? (ram, VRM's, etc)

 

[keakar]

1. PSU's weaken over time. getting a PSU that "works fine" at first, it may well get worse later on. From what you're saying its fine for now, but that doesnt mean it cant have erratic behaviour (for example, the 5V SB rail has weakened so using too many USB devices could make it crash randomly - i've seen that before)

2. GPU temp is one of many temps. is the REST of the card still as cool, or cooler, than the stock cooler would have it? (ram, VRM's, etc)

well im no expert or anything but its got great airflow over the card and all the ram has copper finned heatsinks added and the vrms are in the direct path of the fan but the config is so strange you cant get a heat sink on them. the factory heatsink didnt cool them at all, they werent in contact with anything and the airflow was directed away from them because part of the heatsink blocked the air and diverted it elsewhere. it always bothered me the way they did that but thats the way they made it.

thinking i might need a new card i asked for opinions and i was told over in the video card forum that the best choice to replace my 8800 would be a 5770 so assuming it comes to that point, what would that do to my power needs?

im not just going to be buying stuff but im shopping for opinions on a replacement card and a replacement psu so if and when we finally figure out this issue i will have done my homework and know what to go buy.

i was thinking a 750 watt would be a good replacement for my psu but if i upgrade the video card, weather its now or in the future, will that be the right psu for it?

 

[Mussels]

5770's are fairly power efficient actually, so if it runs an 8800GT, it will run one of those.

Wattage is not your concern - quality and amperage on the 12V rail is.

(a 450W corsair is as good, or better, than a 600W generic, for example)

 

[keakar]

latest freeze up notes:
it seams to happen almost every day at start up and once or twice during the rest of the day. cold starts seams to be a very common time for it to freeze up now so im sure overheating isnt causing the issue here and underclocking of the video card didnt help at all.

i checked my voltages in bios section and it shows:

Vcore - 1.202v holding steady
DDR18v - 1.952v holding steady
+3.3 - this keeps fluctuating constantly from 3.344 to 3.360 to 3.376 to 3.392 to 3.408
+12v - this keeps fluctuating constantly from 11.98 to 12.048

a volt meter shows the +12v going slowly between 12.19v and 12.20v at idle and 12.13v to 12.15v when running video games.

there are objects no longer being generated in my video game as well because in the middle of a task force i noticed my navigational display map doesnt show any ships around me when im surrounded by over 30 of them and im not sure how much of that has been going on or for how long but i just now noticed it.

i ran 3d mark again on high detail settings and this time i put the volt meter on the wiring harness and the voltage was 12.13v to 12.15v and at the end of test #8 it dropped to 12.01v for a second then returned to the 12.13v to 12.15v and at the end of test #17 it went to 12.08v 3 or 4 times but it always returned to the 12.13v to 12.15v (it runs 17 out of 17 tests)

so "Mussels" let me ask you this, would YOU feel good about replacing the psu at this point? im just at the point now where i dont know what else to do except to try replacing something and im thinking the psu looks to be the most likely choice to replace.

 

[Mussels]

try testing the rails when the PC is just powering up from a cold boot.

PSU's are well known to have cold boot issues that clear up as they heat up.

voltages tested in the BIOS are useless, you need a multimeter/voltmeter. you tested 12V, so test the rest.

the one test you mentioned of the 12V rail going between 12.01 to 12.15 is fairly high a change in my opinion - as you saw its not enough to crash the PC, but if the PSU works worse when its too cold or too hot, that could explain your problem


Personally i would replace the PSU, simply because even if turns out its not the problem, its a good investment that can be reused in your future PC's as well. Just make sure to buy something decent these days (not a generic, and not a coolermaster)

 

[keakar]

try testing the rails when the PC is just powering up from a cold boot

ok, i have the meter all set up and i will try a few cold starts tomorrow to see what i can find but im also taking your advice and i just ordered this psu:

CORSAIR CMPSU-750TX 750W ATX12V / EPS12V
http://www.newegg.com/Product/Product.aspx?Item=N82E16817139006

its $90 after mir and it has 60amps 12v single rail

i sure hope i dont still need to get a video card after i replace the psu

 

[TheMailMan78]

my computer specs to the left are up to date

this system has been running flawlessly for a year now and i keep it clean, well cooled, and dust free

right out of the blue lately my computer started freezing up randomely, sometimes when its just sitting idle it will make a "bong" sound and then the screen goes black and the computer is frozen with no way to get in but restarting it.

most of the freeze ups come when multi tasking but it has done it sitting idle once or twice.

a couple of times when i started the computer it gave me the blue screen that says windows had a problem and needs to restart but doesnt say anything more.

sometimes while the computer seams to be working ok, when i start the media player or games it sometimes give me the blue screen but it only says windows had a problem and needs to restart but doesnt say anything more.

just one time the blue screen did say that the nvidia driver had an issue so i updated that but its still giving the random freezes and so far i havent seen any more blue screens yet.

i cant remember how you get to read the error report to see whats going on for sure so can someone remind me how to do it?

if its a virus its no big deal since im getting ready to update from xp to windows 7 but i want to be certain its not something hardware related before i do a reformat and install windows 7.

what would be the top suspects to look at here and how do i test them?

im thinking if it were the video card it wouldnt be effecting the system so even thou it was my first idea im guessing its probably not the problem.

its been a long time since i was fixing a computer so whats the best diagnostic test program for dummies. im not into benches or detailed diagnostic figures, i just want a pass or fail kind of test to see if something works or has issues.

thanks for your help in advance guys

I would do a backup and a clean install man. Sometimes its just better to format C: than to figure out the issue. If that doesnt fix it post back and we will figure it out.......together........in th gayest fashion possible.

 

[Mussels]

ok, i have the meter all set up and i will try a few cold starts tomorrow to see what i can find but im also taking your advice and i just ordered this psu:

CORSAIR CMPSU-750TX 750W ATX12V / EPS12V
http://www.newegg.com/Product/Produc...82E16817139006

its $90 after mir and it has 60amps 12v single rail

i sure hope i dont still need to get a video card after i replace the psu

its the same PSU i have. it'll smoosh your PC into little bits with 400W to spare
(oh and your link broke somehow)

 

[keakar]

its the same PSU i have. it'll smoosh your PC into little bits with 400W to spare
(oh and your link broke somehow)

lol i didnt even look at your specs

i see you say its loud at load, it claimed to be quiet so is it very noisey?

and heres a working link for that psu if anyones interested

http://www.newegg.com/Product/Product.aspx?Item=N82E16817139006

 

[Mussels]

lol i didnt even look at your specs

i see you say its loud at load, it claimed to be quiet so is it very noisey?

and heres a working link for that psu if anyones interested

http://www.newegg.com/Product/Product.aspx?Item=N82E16817139006

on my PC with two high end video cards and a heavily OC'd CPU, in ambients of 35C and up the fan kicks in and gets loud.

on your PC, the fan will never go off its lowest speed.

 

[keakar]

updated voltages:

cold start

12v= 12.25-12.28v (fluctuating) / drops to 12.21v after 5 minutes warm up time
5v= 5.13 steady
3v= 3.42 steady

right after i took these readings it froze up on me and i reconfirmed the readings hadnt changed so this is my volts when it freezes

it repeated the grafics errors on the screen with the 5 rows of different colored blocks

 

[keakar]

update:

new psu installed (newegg delivered it in just a day and a half!)

noticed some immediate diffrences with computer sluggishness gone and bootup took a few tries before i reset the defaults, now it boots fine so i guess it was reading things differently now.

video card idle temp is now 42c and temp rises to around 45c when in use where it was at 40c before and went to around 42c when in use so i hope this fixes everything.


lets compare scores with 3D marks 2001

with high detail setting
got a score of 28043 with old psu
got a score of 49112 with new psu

with low detail setting
got a score of 27720 with old psu
got a score of 48334 with new psu on the

so my scores almost doubled with the new psu


the old psu specs sticker wasnt visible until psu was removed from the case but when i got it out i looked at it and saw that it stated it had a

20 amp 3v rail
20 amp 12v rail
50 amp 5v rail

but the website and the specs on the box it came in said it had a

30 amp 3v rail
30 amp 12v rail
54 amp 5v rail

so here is a perfect example of what happens when you dont double check things before you install them. (i made sure it was the correct model number sticker used on the psu)

as a reminder my 8800gt video card has a 26 amp minimum required

important lesson to remember:

had i checked the actual psu itself instead of trusting what the box and website said before installing it i would have originally returned it and bought something else with proper amperages to begin with

 

[Mussels]

so is the system stable now?

 

[keakar]

so far so good so i think that was all it was but im keeping a keen eye on things


and what a strange and long road to get to it


windows acting screwy, video acting screwy, voltages reading ok, man these problems can cause grey hair

i "almost" wish it would have just smoked itself and i would have known right away what it was. but with my luck it would have taken out the motherboard

and your right, this psu is dead silent


i want to personally thank you for hanging around to help me with this issue, as a modderator i know your in high demand around here and your time is limited.

 

[keakar]

well its been a week now trouble free so im sure its fixed now.


thank you to everyone who tried to help