• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Spontaneous problem - severe video curruption with no reason...

Nazo

Nazo

2[H]4U
Joined
Apr 2, 2002
Messages
3,672
Well, this morning, I got up, turned on my PC and lo and behold, the video is INSANELY currupted. Even the bios has a few strange looking pixels here and there (mostly the text looks weird.) The only single thing I could think of was the fact that I had left the air conditioner running moderately higher than usual overnight and when I booted up the chasis temperature read 11C (normally the lowest I've seen was 14C or so.) However, I turned off the air conditioner, started up rthdrbl and atitool, ran them a while, then completely shut down and cut power to the power supply, waited a while, then booted back up and the problem was still there. Those two programs should have warmed up the video card, so I don't think it could be temperature unless it has a more long term effect. The thing that makes so little sense is that last night everything was just fine and, rather than working the video card, it has been sitting idle for ~7 hours. The most unusual thing I did was play an older game (Freespace) which looks like crap, but is easily well beneath the abilities of my video card despite the fact that it doesn't get along with modern cards (hence the reason it looks like crap.) I have changed no bios settings, I did not raise the overclocking (in fact, I set it back to bios defaults after nothing else worked, shut off again, and waited, but still the same problem.) Well, I'm out of ideas. Is there something wrong with it? I can't imagine how the video card will just spontaneously break when not even in use! Yesterday there was a bad storm and the lightning blew a transformer (not on the same street) so the power was cut off, but when the storm was completely gone and the power restored, I didn't have one single problem whatsoever that whole night... I don't know if I can afford a new video card. I really can't afford to have to buy one already, especially since there is no guarantee the next one won't just do the same -- whatever that is... So I'm hoping someone has some idea what could cause this.

BTW, I did check to ensure the hsf was turned on (actually, I can boot without curruption even if the fan is unplugged, the HSF is that good) and the memory heatsinks were in place, not one moved or fallen off.

Oh well, this isn't the first time my computer has spontaneously done something strange when I turn it off for the night. It's getting so that I'm afraid to turn it off anymore... It seems like every time some weird impossible thing goes wrong...
 
Take your video card out and with a cloth whipe off its agp connections and plug it back in and might fix the problem.
 
1. reseat the card.
2. make sure the card is on its own molex.
3. if you happen to have an antec p160, disconnect the power to the front display. (that was my fun problem this weekend, which exhibited the same behaviour.)


If none of that works, I hope you're still under warranty.
 
None of the above.

Sorry, I forgot to mention. My computer is insane. Seriously, there is something actually mentally wrong with it. It doesn't follow logic when it should... Anyway, the problem magically dissapeared when I got back this afternoon. I've always had this sort of thing, but never before with the video. Usually harddrive or other stuff when it spontaneously messed stuff up... Anyway, I guess if I had to come up with something it would be the temperature. It definitely warmed up by the time I got back. And 11C was rather low.
 
You know, I'm still wondering whether the problem was truly due to the temperature or what. Especially I'm curious how having the temperature too low could so thoroughly affect it...

Also, I forgot to mention, but, when I ran ATITool, it said no artifacts found... You should have seen just how horrible that 3D square looked... Another thought occurs to me. Might my monitor have been the issue? I've never ever seen a monitor have quite such an effect, but, I can't help but wonder. Especially I'm worried, because this monitor is getting a bit old and at certain low resolutions with low refresh rates it can make the most horrid high frequency sound that gives me one serious headache.

Anyway, if no one has any ideas, ok. I wouldn't be surprised if it was just simply another one of those cases where my system decides to defy all logic and reason with weird senseless crap (though usually what it loves to do is mess with things like my FSB and multiplier settings.)
 
Hmm, left on airconditioning at higher than normal overnight.

Possible condensation damage?
 
I worry about stuff like that sometimes. The thing is, wouldn't that continue even after a few hours?

BTW, I have heatsinks on all the ram chips and the GPU has one heck of a nice cooler (once I forgot to plug in the fan but it still got into windows and could run games at factory settings with many artifacts, but no crashes/reboots -- not bad for no fan, eh?) It occurs to me that this might have a bit of a strong effect on those chips.

Anyway, I'm going to do some thorough testing (probably Far Cry) soon. Besides, I'd like to try out this new soundcard and mouse... So far, I think there are no permanent effects though. Whatever it was seems to be basically gone for now. I think that, more than anything else, I'd just prefer to avoid that in the future. (I worry about when it turns winter. I usually prefer to leave my window open and let my room get nice and cool.)
 
Maybe you're psu is being weird. ThermalTake psu's aren't the most reliable in the world. What does MBM (or equivilent) say about your voltages?
 
They always seemed to be holding pretty stable, but, I didn't look at the voltages at that time. I suppose it could be possible that if the PSU got too cold it would have weird effects. One thing that occured to me to wonder was if I had accidentally upped the AGP voltage or something stupid like that (I checked, I definitely did not) so it seems like it could maybe be possible. However, it would most likely affect something more like the CPU first I would think. For some reason I've had troubles with different voltages on this thing. Any lower or higher and it just doesn't seem to hardly work (it's funny because I see people saying they are running 1.8 even 1.9 volt on this same chip at the same speeds -- before I broke the FSB so had 2.5GHz -- yet I can't seem to set any higher or lower at all hardly.) What this means is that I think it would get terribly unstable and probably crash/reboot if the voltages were moving around too much. All the voltages stay within 0.1 of where they should be but the 12 and that might be more due to all the junk I have running on the 12V line. Of course, I can't be 100% sure since I didn't think to look at voltages at that time. It's still beneath 13V (usually beneath 12.5 with a few rare spikes) at the highest I've actually seen while watching.
 
I had this problem before (the display was all corrupted in the BIOS, etc).. had to replace the mobo.
 
Temps too low from air-conditioning? Unpossible unless you have some kind of heavy duty commercial refrigeration and even then it would only be from condensation. I think temps being too low are the least of your worries..
 
It doesn't seem like condensation would be a factor. You get condensation from hot air + cold surface, not hot surface + cold air. Am I overlooking something?
 
Well, if yeah your computer case is cold, and the air around it is hot, the case would start to "sweat", just like leaving a refrigerated coke can on the counter. If you are cooling the outside of the case just as well as the inside it would not be a problem, but I'm assuming you are ducting most of the cold air into the case.
 
ZenOps said:
Well, if yeah your computer case is cold, and the air around it is hot, the case would start to "sweat", just like leaving a refrigerated coke can on the counter. If you are cooling the outside of the case just as well as the inside it would not be a problem, but I'm assuming you are ducting most of the cold air into the case.
Click to expand...
No, he just said he had the air temp down too low....room temps too low for a computer would be unliveable imo.
 
Well, in this case, cold surface, cold air. Remember, this thing is cooled so well that even with the fan off on it, I can still get into windows. It's copper after all. Since I never played any games during the entire time, it wouldn't have been the least bit warm except after I tried to warm it up (and even then I didn't leave it long since I was afraid I could just end up doing permanent damage and I hadn't noticed any obvious improvements.) Also, the PC had just been off, so it wasn't warm from use. All I know for sure is that I turned off my AC, shut off the PC, then went to class, and when I came back, the problem was gone. I haven't seen it again so far.

Anyway, I figured temperatures being too low wasn't very likely, but, it's the only single thing I could come up with to explain such a curious occurance. None of this really makes sense IMO. I just remembered I had recently switched the crappy thermal tape on my memory heatsinks on the video card with artic silver epoxy, but it was well past the time needed to be completely 100% set (I think you are supposed to give it near to a day for it to become truly permanent, before that it's darned close though.) It had been two days I think. Anyway, it does occur to me to wonder though if the temperature could have had some kind of effect on the still semi-new compound though.

Oh well, I guess I'll just use standby mode when not worried about thunderstorms or anything. ^_^
 
Cold = good. Do not be afraid of lack of heat. The colder your shit, the higher overclocks you get, and the longer lifespan of your hardware.


But anyway, we have no F'ing clue what your video card is. A radeon 9600 ?
 
Lol, it IS in my signature... Yes, it's the Radeon 9600 Pro. I guess the only thing about it that might not be perfectly clear is the fact that it is made by ATI, not by another company (such as Sapphire.)

Anyway, I know cold is good for the lifespan of the hardware and all that, I just wonder if it could affect operations when you hit numbers like 11C. In particular, I was wondering if it could be related to that relatively new epoxy. I really have no other ideas to explain this as these are the only things I can think of having changed recently at all.
 
One thing I didn't see mentioned was what sort of MONITOR you are using? an LCD monitor responds slower at cold temps. until it warms up it may show some corruption....I've seen laptops do this quite often when cold. I doubt there is a problem in the tower that would be affected by cool air unless it has condensation it the case.
 
It's a relatively old CRT. Actually, I mentioned before I was wondering if it was possible this thing is finally beginning to truly show it's age. In particular, the low resolution thing. At 800x600@60Hz, it makes a noise that gives me a splitting headache. It is starting to do that at 640 too I think, but usually I avoid 640 at all possible costs. I'm now using Powerstrip, so even on those older games that used 800x600@60Hz *cough* Diablo II *cough* I can force a proper refresh rate now. Of course, I always wondered why low resolution and low refresh as it's usually high resolution at high refresh that has issues... Anyway, as you can see it has it's own share of issues. One of these days I have to get an LCD, but, the resolution thing is an issue, so I guess it would have to be a big one so that I can turn off the ugly resize option and still see it when at lower resolutions. I can't stand that resize they do and, besides, it defeats the purpose of buying something of which the chief selling point is it's clarity.

BTW, my ancient total peice of crap laptop doesn't do that. Then again, the LCD is the worst ever made I swear. It has a refresh rate of about 10Hz, or so it feels like. Even really old platformers tend to tear and blur.
 
AH HAAAA and older CRT and newer vid card can cause probs for sure....when the computer is cold in essence the card will run MUCH more efficiently COLD and may be slightly overclocking the monitor by a couple of hertz...as it warms up it slows down slightly due to the heat...and stabalizes....try a newer monitor and I doubt there will be any problems....60hz on 800x600 is pretty old stuff.
 
Not THAT old... 800x600@60 was what that particular incredibly stupid game insists on doing. It has no options for higher. However, I'm now using Powerstrip to force it to use a decent refresh rate (and to prevent giving me a headache...) I think it actually can do 800x600 @ 120. The monitor can go up to 1280x1024@60 (or 1280x960@72, don't ask me why,) but, I am "overclocking" it a little by using 1440x1050 (about 2x DVD, for maximum DVD/video watching quality, but, still darned close to 4:3 to prevent ratio problems -- 2x DVD being 1440x960.) Anyway, bear in mind that these issued occured on bootup in the BIOS even, which is 640 if I'm not mistaken. (I'm pretty sure VGA text mode is based on 640, but not 100% certain...) Not to mention the little windows boot logo which uses a slightly lower resolution if I heard correctly. Besides, I would imagine that the video card should handle matching refresh rates pretty darned well or else the VSync options (both set to always on) would be next to useless...
 
ya at 640x480 it will be considerably faster I recommend getting RIVATUNER and pre-program all the refresh rates in it so it will run YOUR presets by default with VSYNC enabled. Also if at all possible ensure you have actual MONITOR drivers installed and that it is not seeing it as just a plug&play monitor...believe it or not it will make a huge difference in stability.
 
Gruntled Employee said:
ya at 640x480 it will be considerably faster I recommend getting RIVATUNER and pre-program all the refresh rates in it so it will run YOUR presets by default with VSYNC enabled. Also if at all possible ensure you have actual MONITOR drivers installed and that it is not seeing it as just a plug&play monitor...believe it or not it will make a huge difference in stability.
Click to expand...

But, in VGA text mode, it uses 60Hz. This thing can go up to 140 or so at that resolution!!! Also, as I said, I'm already using a program that controls the refresh rates. Oh, and this is an ATI card. I don't know just to what extent a program made for nVidai will work on ATI, but I'd rather stick with programs made for ATI instead... In other words, even if the video card was using a higher refresh than 60Hz, it would still not be 133% more just because of a slightly cold room!

As for the drivers, well, that can't affect the BIOS however, I can assure you it's not running as a plug and play monitor. Using powerstrip, I took the driver for my monitor (Optiquest Q71 btw) and just added two new resolutions on top (1440x1050 & 1440x960.) Regardless, I've been using it with this driver for well over a year with no problems (in fact, across two video cards counting this one) and, again, it still doesn't affect the BIOS since it is a windows driver and the BIOS uses only plain VGA test mode (not one of those fancy GUI BIOS setups which still usually just use 640x480x4 -- still VGA.)
 
Sorry, for some reason I thought you had a NVIDIA card....as for your BIOS screen...it starts that up in 320xsomething res. not 640x480...

only thing I can think of is that ATI cards sometimes have flashable BIOS I have flashed 3 older ATI cards and it worked quite well...maybe see if they have a new BIOS update for your card..if that doesn't work then a MOBO BIOS update might, although if it works for the most part I wouldn't try this unless there is absolutely no other choice.
 
Gruntled Employee said:
maybe see if they have a new BIOS update for your card..if that doesn't work then a MOBO BIOS update might, although if it works for the most part I wouldn't try this unless there is absolutely no other choice.
Click to expand...

I'd have to be pretty desperate to do that. There are so many things that can go wrong with that it's not even funny. Besides, bear in mind I've had this card almost a year with no problems before that -- or after for that matter... While the mb is relatively new, I've had it well over a month at this point and the only problem with it I caused and is unrelated to the video in any way (messed up the stinking southbridge.)

Oh well, if no one can figure out just exactly what happened, I guess it's no big deal. I posted the original message just after class before I came back home, so I didn't know at the time that my PC had yet again magically fixed itself. I sure do wish it would quit messing with me like this though.
 
Option B)....Don't shut it off.....just shut off monitor....Mine NEVER gets turned off unless I gotta make a hardware change.
 
ZenOps said:
Hmm, left on airconditioning at higher than normal overnight.

Possible condensation damage?
Click to expand...
condinsation affecting the monitor given the no artifacts report on the video.

How old is the monitor
 
jacuzz1 said:
condinsation affecting the monitor given the no artifacts report on the video.

How old is the monitor
Click to expand...

Got me. Old. This was a hand-me-down and I've had it so long I really couldn't say.

The thing is, I've seen the effects of weird signal things, but I've never seen anything quite like that. Besides, why would the text mode just have weird little colorization on parts of the text like that, but no lines or anything?
 
Well, it happened again. Very similar occurance too. It stormed in the middle of the night, and, in my 3/4 sleep state, I rememered to shut down everything and unplug the modem, but I forgot to unplug the PC. I get up in the morning, turn it on, and the exact same thing has reoccured. At one point earlier the problem seemed to be gone, I went into windows, ran ATITool just to be 100% sure, and bam, it just screws up again. Again low temperatures, everything was just about the same. Except that ATITool was reporting errors. (I do wonder though, last time I ran ATITool for only seconds or so as I was afraid my card was tearing up.)

Well, all I know to do right now is just leave the PC off for a long time again and turn off the AC. This is rather annoying to say the least. What on earth is causing this? Just the card getting too much voltage in it's capacitors or something? Or is it perhaps related to the motherboard or something? Because, I never had this problem when I was using this same card on my previous MB/CPU combination. Could the AGP voltage perhaps be screwing up?

I'm running out of ideas.
 
I had the SAME EXACT problem yesterday. Turned out that the main power supply cable was loose, my 3.3v was at 2.8!! Make sure your 12V and main power connector is seated properly.

My post from yesterday
 
Nah, in fact, one of the first things I did was pull the power cable and hit the power switch to get rid of almost all the power (the computer seems to store just a tiny bit -- enough to run about 3/4 of a second or so.) I did this multiple times in fact. So, I know it wasn't loose since I plugged it back in just afterwards and didn't plug it in loosely.

Anyway, I was wondering about the voltages. They weren't perfect, the 3.3 was at 3.4, the 5 was at 4.97, and the 12 was at 12.4. While not perfect, the 12 was the only one that was truly far off, and really that's not SO bad.

EDIT: Btw, I made certain the BIOS was set to the default correct AGP voltage (1.5, the lowest it can even be set to) and even changed the AGP frequency from auto to 66. After that, I even gave 50 a shot instead. Still no good. Nothing seems to work. Interestingly, the linux framebuffer things seems ok -- I think. It's not very graphical.
 
Those seem a bit high. My mobo always stays within 5% of the voltages, albiet a bit lower.. 12v is at 11.52, 3.3 @ 3.29, 5v @ 4.83.. Perhaps the high 12v settings is causing the problem? Do you have any power supplies lying around to check??

Also, just having the 3.3 down to 2.8 created massive video corruption on my system that was not apparent when I powered down the PC. After 5-10 minutes in windows 2d kept freezing and nearly all the text on the screen was corrupted. I know a 12v @ 12.4 is not proportional to the voltage anomoly in my system, but it may be the cause. Luckily, in my case it was simply a loose cable.
 
No, I have no spare PSUs. Besides, that's about normal for it I believe. Anyway, that's less than a 4% variance. Sure, it sounds like a lot, but you have to remember that 0.4 isn't such a high variance for a 12V line. I saw worse when I had a crappy PSU before this. I'm pretty sure I read somewhere that it's normally only a problem for most components if it varies by more han 10%. Certainly 2.8 is more than 10% in fact. (About 15%.)

Anyway, I haven't exactly memorized the numbers, but, I'm almost 100% certain that the PSU isn't acting any differently than ever before. Also, wouldn't variance along the 12V line be FAR more likely to affect the CPU? And the VCore is as rock solid at 1.63 as it usually is (this mb overvolts by a lot.)

EDIT: An addition to my previous mention of the framebuffering. I just ran QTParted (partition magic clone) and see no graphical glitches in it. After running in that rescue disc a bit seems to be when it suddenly fixes. Since it's working ok atm, I'm going to try running it minimally (no atitool or anything) for a while. I already set the clocks back to factory just to be sure when it first messed up this morning, so that won't be a problem. Last time I turned on the AC after it worked and it sudenly messed up again, so I guess I'll not turn it on for a while.

The 12V line is still at that and no video curruption.

EDIT2: Just noticed MBM5 reports two 12V numbers. +12 is at 11.92V, -12 is at 12.4V. Are these actually two seperate lines? If +12V is the ground that -12V uses, then it shouldn't really be going over 11.92V since it can't be more than it's ground.



EDIT3: And here we go again. Currupted once more. Ran for at least an hour without problems, I try to check my e-mail and bam. Not even especially cold. The mb went up to 20 and it was 18 without curruption a bit ago.


Now working again. The thing that worries me the most is, can I still try to sell this card if it might actually have some minor damage causing this? I suspect now that it has to have been due to the storm though. Each time there was a storm.

Last edit: Ok, problem solved. The GPU is apparently screwed up. I played around with the heatsink a little and, every time I did, the results were a little bit different. However, in the end, I just can't get rid of the curruption. Well, now, lucky me, I'm stuck with an ancient PCI Rage Pro card (eg the chip that made S3 look good...) Until I can upgrade. *sigh*
 
You must log in or register to reply here.
Back
Top