This is a classic example of how to diagnose problems on your PC, see if it helps you before posting!
CuZ

Far Cry locks up after playing for a few minutes. Have uninstalled, cleaned the registry, and reinstalled. Nothing helps. Have all the latest drivers, Drive Cleaner 3 used between driver updates. My desktop settings are set the same as the Far Cry video settings: 1024x768, 32 Bit. I've tried everything and have run out of ideas. Here's my system specs.

Far Cry v.1.31, Build 1378
ASUS A7N8X-E Deluxe with latest BIOS 1013 and
latest nVidia nForce chipset drivers v4.27.
AMD Athlon XP 3200+, 400MH fSB
1 GB DDR Memory configured dual channel
PNY nVidia Verto GeForce FX 5700LE Optima with 128MB DDR memory.
MGE Vigor 450 power supply
Sound Blaster Live value with latest driver v5.12.2.252
Samsung SP1614N ATA 160MB Hard drive using 48 Bit LBA.

I was using the nVidia onboard audio processor and disabled it and installed the Sound Blaster in an attempt to eliminate possible audio processing problems but I still get the same lock up trouble. Any help would be appreciated.

This message has been edited. Last edited by: CuZtds_Lst_Stnd,

Not that this particular driver should matter, but isn't the latest nVidia platform driver at v5.10? See http://www.nvidia.com/object/nforce_udp_winxp_5.10 for details.

Your problem sounds like it could be a power supply or overheating problem, but your power supply seems to be more than enough for your system. I'm also guessing that you have cleaned out the dust bunnies in your system when you had it open to work on the sound card, so overheating shouldn't be the issue.

Are your memory sticks matched for dual channel, and working well? I had a problem while back with a bad memory stick that would only show up running Far Cry; I had to get a DOS-level boot command memory test utility to finally isolate it to the memory stick.

Other things to consider are any overclocking you have - try disabling them and see if that makes a difference.

Sorry, I forgot to note in my system summary that my operating system is Windows 2000 SP4. The nForce drivers are the current one's for Win 2000. I don't think it is overheating because Asus probe shows it running cool and I have tested with the side panel off of the computer case. There is no dust buildup at all. I am not overclocking anything and have never overclocked anything. Although my memory sticks are matched I could remove one to see if it solves the problem. It does seem as if it could be symptomatic of a memory problem because the only way to recover from the lockup is to hit the reset button. I run Unreal Tournament 2003 and Call of Duty UO with no problems but I know that Far Cry is much more processing intensive than these. I'll try removing one memory stick this evening after work and give it a try. Thanks.

Just a bunch of questions....

Are you running your mem and CPU in sync at 400mhz? Running async caused probs for me.

Are you sure your system is stable? (have you run memtest, prime95, etc.)

Are you overclocking? I remember reading someplace that "may" cause probs with FC.

bios settings; Check out, http://www.nforcershq.com/forum/ , go to the asus forum and post. Couple of things off the top of my head are AGP freq locked at 66mhz and turning off fast writes.

Have you checked out the FC tweak guide it's stickyed at the top of this forum.

I have never overclocked and run my system as the bios defaults, as far as memory and cpu speed that is. Most other settings as well.
According to SiSoft Sandra my CPU speed is 2.19GHz, with a FSB of 2x199MHz (398MHz data rate). How change the option for sync or async clocking settings, I thought there was only sync? AGP is a 66MHz Bus running at 2x.
I ran Memtest86+ for three hours with no errors. Am not familiar with prime95 but I'll look it up.
As far as stability, I have no problems with the system other than Far Cry. I can burn DVD's and CD's, watch DVD movies, edit videos, play other games such as Call of Duty and ut2003, and run all the DXDIAGS with no problems. I am running DirectX 9.0c.
I've been thinking of using another smaller hard drive, loading only WIndows 2000 and Far Cry on it to try to eliminate possible conflicts with other software and to see if the hardware runs ok then. I just need to scrounge up a spare drive.
I've read the FC tweak guide and another Far Cry tweak guide at Tweakguides.com. I am not pushing the settings. I use the Far Cry configuration program and just use autoselect. I am not being aggressive on my settings at all during my tests.
I'll check out your link and prime 95. I'll post my results. Thanks for the suggestions.

Scheval - sorry for the not so helpful suggestions, sounds like you've covered all the easy things quite handily, I don't know that there is a simple solution here.

If you are going to create a small test partition, you might want to consider a Win2K partition first, but then re-doing it as a WinXP partition and see if the OS makes any difference. This is just a stab in the dark though, no real experience that Win2K might be a problem. It's just that most people running Far Cry are doing it on WinXP rigs.

Hey Slug I thought you were right on the nose...PSU and heat. Those are common problems with the A7N8X series MB's. Even if it is a high watt PSU if the output on all the rails isn't stable and in spec then that will cause probs.

Scheval if you can run Prime95 torture test with no probs then it's probably save to say that your system is fine.

I was running async CPU at 266mhz and mem at 333mhz and it caused problems with one game. You didn't say what spec your mem is so I thought I'd mention that. I don't have the exact info on the bios settings (I'm at work) but it's under advanced chipset features. Check out http://www.nforcershq.com/forum and go to the asus section. There is a stickied OC tutorial that thas a lot of info there

Also, have you checked the event log? That might point you in the right direction.

Of course then there's the standard stuff too... run adaware, spybot, defrag, etc.

Thanks for the info. I'll try running Prime95 this evening when I get home from work. My memory is two 512KB DDR PC3200 by Kingston and it ran Memtest86+ for 3 hours error free. I suppose it could be CPU overheating although the CPU fan and heatsink are high quality and ASUS probe shows the fan speed running above the benchmark rpm's. I don't know of anyway to check the CPU temp while Far Cry is running. Although Windows seems to operate every other software fine, at this point I am wondering if something got corrupted during the Norton Ghost drive migration when I upgraded to the 160GB one. I think Far Cry was running better before then so it seems to be the logical possibility. That's why I want to try a clean Windows install. Still, I would think that a corrupt Windows installation would have more symptoms.

I vote for either CPU or vid card hitting too high of a temperature.

The best way to read real time temps is with a thermal probe under the CPU or GPU.

Does it crash if you run at lowest graphics settings? A 5700LE will be very strained to run at 1024x768 and above medium. Lower to 800x600 and low graphics and see if you crash. I bet not.

OK, thanks. I'll try all your suggestions.

Motherboard Monitor (mbm5) is a great free utility for checking temps and PSU output. I have it set up to read the temp from the diode in the CPU itself, the socket temp and the case temp. It can record all the heat and psu readings in various file formats too. One feature I like about it is I can just put the cursor over the icon in the system tray and the temp measure will pop up.

This is the link to the OC tutorial. The "there" I posted last didn't seem to work
http://www.nforcershq.com/forum/viewtopic.php?t=23939

edit: just google motherboard monitor to get the DL links if you interested.

I started to talk about MBM but took it out of my post since you cannot read temps during play only when you exit the game. There is about a 3-4 sec delay until it polls the probe and reads the new temp once you quit FC. Also, you can set the program to display temps in the taskbar all the time without mousing over it.

Hey coops if you double click the sys tray icon and go to "Interval & SysLog" you can record the info. I checked the "enable interval log" box and .txt box, set the "interval log interval time" to 60 seconds and the "maximum interval entries" to 550. Played and it works fine.

By default it puts the file in the ...programs/motherboard monitor 5/log... folder.

edit: I've had too many beers but if you set the interval time to your liking and make sure there is enough "interval entries" it works.

Thanks for the tip. I tried it and the log did not save, I must be doing something wrong. I have version 5.3.7.0.

Ran some tests this evening. System rebooted (halt on all errors in bios) while running 3D Mark03, and while running Hot CPU Tester Pro. I got Illegal Sumout, hardware failure while running Prime95. I was able to monitor CPU and MB temps with ASUS Probe while they were running. All voltages were stable and the temps rose just a few degrees but I still got a failure. Looks like I have a CPU or MB problem that only fails when being stressed by software like Far Cry and those designed to stress the machine but it dosn't seem to be heat related. I think I have some hardware troubleshooting to do. Thanks for the links and the referrals to the tools. At least I now have some diagnostic tools other than Far Cry with which to isolate my trouble. Thanks for all your help.

Scheval - the pleasure is mine. Can't speak for fabulouscoops or pakmanis, but for me it's a pleasure trying to help someone knows what he/she is doing.

It's easy to spout off the typical answers - overheating, drivers, overclocking, PSU, etc., and they answer people's problems about 90% of the time. Your problem is a real challenge and I will be curious to hear what you find.

I ran the game at 800x600 and it still locked up. I'm going to examine my cooling system closely, clean and reinstall the cpu heatsink, and temporarily place a large fan blowing into the case while testing. I will borrow a temp probe to check the accuracy of ASUS probe. Hopefully it is a temp problem and not a CPU replacement.

@scheval

Sounds like your system is rebooting or throwing up errors ONLY when it's under stress, i.e. when running Hot CPU Tester Pro, 3DMark03, Prime95, etc.

Ever considered that it might be due to your system under stress "overrunning" the +12V rail on your Power Supply Unit (PSU)? This is a common cause of "classic" random instability symptoms EXACTLY like the ones you've described, i.e. reboots, lockups, memory errors, fatal exceptions, etc.

Even with temps looking pefectly normal, a stressed system is STILL drawing MORE current (Amps) than usual on the PSU's +12V rail (which powers the majority of your internal hardware, i.e. motherboard, hard disks, CD/DVD drives, etc).

With sufficient hardware installed, your PSU might already be drawing VERY CLOSE to the maximum current its +12V rail is rated for (the "+12V Max Current"). The +12V rail will NORMALLY be stable at this level, but sufficient system stress CAN draw enough EXTRA current to push the +12V rail OVER its maximum current rating, "overrunning" it and causing other voltage rails (+3.3V, +5V, -5V, -12V) to be pulled DOWN, "underrunning" them which typically results in the "classic" random instability symptoms I described earlier.

In your case, I would theorise that Far Cry's advanced graphics are stressing your video card, which causes it to run hotter than usual. Heat lowers the resistance of electrical components, and the lower resistance causes more current to be drawn. If the extra current draw is enough to SLIGHTLY overrun your PSU's +12V rail, instability will result, hence the lockups. Same goes for 3DMark03 but, since it stresses your video card even MORE than Far Cry, the overrun will be more severe, resulting in reboots.

I strongly suspect that installing a PSU with a higher "+12V Max Current" than the one you currently have will most likely cure your problem. I had similar instability problems when I first installed my GeForceFX 5900XT video card. I quickly realised that the most likely cause of the instability was the 5900XT's extra power connection "overrunning" my existing 400W PSU. Once I upgraded to a 650W PSU, my system was suddenly PERFECTLY stable again!

It's surprising how many people overlook PSU requirements when building their own PC, since the PSU is actually the most VITAL component in any PC! After all, without a RELIABLE power supply, NOTHING will work RELIABLY! People lacking electrical training don't realise that installing MORE hardware draws MORE current from the PSU, and sooner or later they'll EXCEED the PSU's maximum limits and start wondering WHY their system has suddenly become unstable!

Tip: When choosing a PSU, DON'T fall for the trap of thinking a higher "Total Power Output Rating" is necessarily better. E.g. 500W is better than 400W. Those ratings can be VERY misleading! Instead, look at the PSU's +12V Max Current rating and think, "the higher the better"! E.g. a 400W PSU with +12V @ 24A is BETTER than a 500W PSU with +12V @ 18A!

Basically, the higher the +12V Max Current, the MORE headroom you have for powering future hardware upgrades. This is why I upgraded from a 400W PSU with +12V @ 16A to a 650W PSU with +12V @ 25A. I know better PSU's are available, but I just wanted a cheap one (QTEC) which would easily deliver all the power I would ever need!

Hope this helps you.

This message has been edited. Last edited by: H8Ball (PC Engineer),

Thanks for the info. Actually I do have an electronics degree but utilize my skills in the central office of the phone company, pc's are just a hobby. The specs on my power supply are:
+3.3V@15A, +5V@35A, +12V@17A, -5V@0.6A, -12V@0.6A, +5VSB@2A but I think my +5 is actually 4.8 something. The others are slightly higher than specs.
I will be testing with MBM 5 and will monitor my voltages as well as the temps. However, I did just find out that the AMD Athlon diode cannot be monitored on an ASUS A7N8X-E deluxe motherboard using MBM 5.
( http://forums.pcper.com/showthread.php?s=&threadid=300388 )
I do have about six USB devices attached to my PC but other than those and the motherboard all I have pulling power is the video card, hard drive, and DVD drive. Do you know if the USB devices pull much power while sitting idle?

Is it possible to rig an ammeter on the +12 voltage line while testing somewhere? I guess it would have to be inside the power supply to measure all the 12V lines. If the 12v line pulls too many amps will the other voltage levels drop as well or just the 12v line? Some of the power supplies have two +12v lines, +12v1 and +12v2 noted in their specs. Do you know if these power supplies have two separate +12v power supplies for different devices? Do you combine the amp ratings in this case to determine if its adequate or not?