TechSpot - PC Technology News and Analysis

My computer was doing the spontaneous reboot thing, then it just died. I replaced the power supply with the Antec dual rail 12V that you see in my signature, and everything seemed to be fine. When I was playing Madden 05 earlier this morning, the graphics started to go a little goofy and then the game just shut off, but the computer remained on. I then turned on HL2 to see what happened. After about five minutes the computer rebooted in its entirety. I got the whole "windows has recovered from a serious error" and the website that it took me to stated there was a stop error with a device driver(?) although it does not tell me which.

The CPU temp stays between 30-35 degrees celsius when idle and goes up to 50 degrees celsius when doing high graphics stuff (numbers from ASUS Probe). I'm not overclocking the video card or the CPU. The power supply I just installed has a pretty slow fan, it only hits about 1200-1300 RPM max, goes down to 700-800 when sitting around idle. Maybe because it's a bigger fan, moves more air? The temperature of the CPU hasn't changed at all with this new PSU, so I don't believe it's overheating. I dont know how hot the video card gets. The voltages seem right on with ASUS Probe.

I attached the minidumps because I'm at the end of my rope. I thought I had the problem licked with a new PSU, but that's not the case. Can someone peek at these minidumps and give me some advice?

Attached Files

minidumps.zip (42.5 KB, 2 views)

One minidump crashes at your videocard driver. The other minidump crashes with memory corruption. You may have a ram problem with either your physical ram, or your videocard ram.

Try updating your videocard drivers.

Go HERE and follow the instructions for testing your ram etc.

If you could zip 5 or 6 minidumps together, it would give us a better idea of the culprit.

Please let us know the results.

Regards Howard

Howard,

These are the four minidumps that preceded replacing my PSU and if you add these to the two that I attached, you have six.

I was using Omega drivers, so I uninstalled them, and installed the newest ATI Radeon drivers from ATI. I also installed a fan just beneath my AGP slot, it's an Antec PCI fan card with two fans that are pulling heat away from the card. I've not unseated my memory yet, nor have I run memtest (I have to figure it out) since I figured you'd maybe give me more feedback on these minidumps.

Thanks for taking your valuable time. I've learned tons on here from everyone, and have found that building PCs is a learning process.

Attached Files

moreminidumps.zip (87.5 KB, 1 views)

An Update here,

I ran memtest86 and there were errors in the memory right at the first pass of tests. I have four Kingston DDR3200 modules, each 512 MB in size, sitting in DIMM A1/B1 and DIMM A2/B2. I have two older ones (about 18 months) in A1/B1 and two that are about six months old in A2/B2. I removed all the memory modules, put the newer ones in the DIMM A1/B1 slots, and the test ran without issue. I only let it run once, since the first time the errors were picked up right away and I'm trying to get this situation resolved before my little ones wake up.

I removed those, put in one of the old ones in the first slot, and left the other older one out. Test ran through one pass without problems. I then removed it, and put the other older one in its place. That is currently running, and I'm about 50% of the way done with the test and no errors.

So, it would seem to me that somewhere there were memory errors. Was it the combination of all four that caused the problem? The mobo should handle it without a problem. Is it a memory SLOT that is bad? I have not tried modules in each slot, but if that were the case I would assume that it is one of the DIMM A2/B2 slots that is bad then. I was not sure if the Mobo would boot up with memory in the 2 position alone.

nope, 2700, not 3200. my bad. don't know if that makes a difference or not.

Is you ram comaptible with your mobo? Check with your mobo manufacturers website for compatible types of ram. Also, check your mobo manual for ram configurations/speeds etc.

Normally if memtest86+ gives you errors, then your ram is bad. It`s just a case of finding out which stick is bad.

Your latest set of minidumps are crashing with various errors. This also points to faulty ram.

A couple of your minidumps crash at your Zonealarm firewall. Try uninstalling Zonealarm, and using the free Sygate firewall instead. You can get Sygate from HERE

Regards Howard

I'll give the SYGATE firewall a try. I like ZoneAlarm, but if it's giving a problem then I'll just ditch it. I had McAfee and hated it.

I had errors pop up with MemTest when I had all four modules in on the first go, but when I tested each module invididually in the DIMM A1 slot they all passed the first test. Why is that? I would assume if they failed after the first pass with all of them in, the bad stick should have failed with the first pass as well.

I had an interesting thing happen when I turned on my desktop for my son just now. I got an error with one of the executables for the video card, it said click OK to continue and cancel to debug. I think it was ATIEXXX(?). He's playing now so I'll have to look into it when he gets off. It didn't seem to affect the boot though. I played Madden 05 last night for almost 45 minutes without any problems then I loaded up Half-Life 2 and played that without any issues (had a little lag with HL 2 that I don't remember when I first got it and played it). Is it possible for viruses or trojans to adversely affect the RAM??

Sometimes ram sticks will work perfectly well on their own, but cause problems when being run with other ram modules. This has to be some kind of compatibility issue.

Also some systems don`t seem to like having all the dimm slots full of ram. Why that is I`m not sure.

I know viruses/trojans can reside in memory, and cause all kinds of problems.

If you suspect a virus etc, go and read the how to threads in the security and the web forum.

Regards Howard