Graphics problems and lock ups

JustHalf

I'm having some problems. I'm running on:
* Win2K SP1
* 1GHz @ 1333MHz
* 256Mb PC133 SDRAM
* GeForce2MX SDR (32Mb) w/12.4 drivers
* ABit KT7A-RAID (latest BIOS & drivers)

Every so often (usally about 5-10 mins after boot) I get weird mouse cursors. Some are split, some have lines of random noise.

I'm also having lockups in Quake 3 and Counterstrike. The game will freeze, the sound will repeat over and over (then stop) and I can Ctrl+Alt+Delete to get the mouse pointer back, but that's locked.

Help?

davoc

not too sure now, but it could be the fact that you have it overclocked, quite a sizeable overclock too by the looks.
that can sometimes mess with pci cards- such as your sound card- and because of that your sound card might be running too far out of its spec for it to be to be stable.
give it a go at 1Ghz and see how she goes from there.

Leverz

Overheating, I would say, plain and simple.

BNC

<font face="Verdana, Arial" size="2">Originally posted by davoc:
not too sure now, but it could be the fact that you have it overclocked, quite a sizeable overclock too by the looks.
that can sometimes mess with pci cards- such as your sound card- and because of that your sound card might be running too far out of its spec for it to be to be stable.
give it a go at 1Ghz and see how she goes from there.</font>

I'd agree. Whats the temp telling you. Did the problems happen after you had done anything or just all of a sudden? What is the fsb running at? Any cooling on the video card?

Stonemason

Could be your ram doesnt like being forced so hard i had a very simlar problem recently which became especialy noticble in hitman 47 but also happend in cs.All i did was drop the speed abit while playing these games and i had no problems (i suffered no loss in fps still rock steady at 99 in cs )

Stone
ps if your getting alot of page faults this was also a sympton

Supercell

Looks like overheating to me too, out of curiosity,what do most people here use to measure temps inside the pc?

BNC

I have a kt7a mobo. It has a couple of thermal sensors on it. You can do temp tests from the BIOS or there are a couple of monitors that run from windows. Motherboard Monitor is one, VIA HM is another. The former is pretty good, you can set alerts if your fans or temp are out of specified limits etc.

JustHalf

Still happens at a gig (tried that already). I think it might have something to do with one or more of:

* Overheating GFX core
* Power supplied to chip
* Power supplied across AGP bus

The PCI bus speed is 33MHz, so I assume the AGP is at the default also (it's the default settings for a 266FSB Athlon).

Stonemason

Huuum if your running at 266fbs i wouldnt be suprised as those t/birds are 200fbs.Having said that Gerry knows more on these matters than i so if he figured it could take it id have to say i dont know.Just wondering what powersuply your useing i would recomend at least a good 300w.


Stone

BNC

No, the newer t/birds are running on a 266fsb. If you want you can check the little number on the processor, afaik, if it ends with a B its a 200, if its C its a 266.

Stonemason

Think we had this discusion before and most people agreed that with the unlocked Axia chips you could push them to 266 even though they are suposed to be 200.


Stone

_CreeD_

The FSb won't be an issue. The CPU won't really be bothered by it (eg. a 1.2Ghz(100Mhz x 12) cpu will be as stable when run as 133x9 - providing you choose the 133/33 PCI setting and your ram can handle it.
Which leads me to the important bit, the KT7a has a load of ram tweak options. I found when I had my PQI 133 ram I got lockups all the time with these settings maxed, turning them off and running at the same speed it was fine (Though the system was slower). Upgrading all the ram to Crucial fixed it all up.

BNC

Yes you can do that no problem. I have a 1.2 200fsb running at 1.4 266. What i said was that the newer chips are 266fsb not 200

JustHalf

Right, I'll try dropping the RAM settings (maxed ATM) until I get a new PSU, then pump them up again (just in case the PSU made a difference, if not, drop it down).

I'd say the lock-ups and the graphical gliches are caused by two seperate issues. I think the gliches are due to excessive core heat, but I have a crapply shaped card (no room for a blue orb, for example).

core

I get lots of lockups for some strange reason also in lots and lots of games. hmmmmmmm.

"Hey come back here kid!"

Illkillya

it sounds like overheating for JustHalf, but Core - i'd say its your drivers. especially if you have a voodoo3 which has a lot of problems if you don't uninstall your drivers before upgrading

Gerry

right, lets get this straight: The chip doesn't care what fsb it is running on, it will happily run on whatever fsb the motherboard can reliably run.

Hopefully backing off the ram settings will fix it dave, I told you they can cause occasional crashes, and to back them off if this was the case.

Your ram is not being overclocked, since it is pc133 ram running at 133mhz.

As for the power supply, well since running the machine at 1ghz does not help, I don't think the psu is at fault. I also don't think a 1 gig would be overheating with a thermoengine cooler. I ran a 1 gig at 1.4ghz @ 1.85 volts on the same model of 235 watt psu. However, maybe your one just isn't up to scratch.

Apart from that, you may be having irq conflicts. Try running the machine without any pci cards, and put the sblive back in slot 2.

I dunno if your geforce2 mx is overheating, make sure it isn't overclocked for the moment.

JustHalf

Not CPU core, GPU core! The graphics glitches! And I know what speed my RAM is meant for, thank you very much! It's only overclocked at 140/280DDR FSB (other setting I was playing around with).

I didn't think of the RAM though. Thanks.

Gerry

yeah, I was just pointing that out since a number of people suggested it might be overheating. Give it a good run of prime95, to see if it is the graphics card.

_CreeD_

Hmmm, moved my system to a new case last month and started having lockups (New one wasnt ventilated aswell as the old). I checked the motherboard monitor after each crash and noticed the ambient temp was 29 and the cpu-43. Which should've been fine, but was still ~2 over the old readings. SO I went mad trying to get better cool arflow around the Cpu. Lowering the CPU clock speed also helped, BUT.....
The long and the short of it is it was fine (At full power) when I dropped my GTS pro clock speed by 5Mhz. While the CPU was in fact warmer it was not the culprit, but was contributing to the Vid card now being too hot to run at it's old stable speed.

I'd say your problem might well be the Vidcard alone, lower it's clock speed.

Gerry

Thats interesting, because I reckon my geforce gts is getting flaky also. I have had to raise the io voltage to keep it stable at default speed, now the main reason why it was crashing was that it is running on a 112mhz agp bus instead of 66, so I backed off the cpu speed a little. I think with enough cooling of the core it might run on a higher agp bus...