Biostar P4M900-M4 slowed way down and i'm not sure if it's the mobo or HDD that is the issue. Memory has been ruled out with a 24 hr test with memtest 86+ and no errors. I started getting random BSOD's to the point windows crashed with errors on the HDD that I had to recover from. Since the HDD has been wiped with fresh install of OS, latest drivers, etc but still slow and now BSOD's are starting to return. This is my second mobo, first one went bad under warranty and this one is V6.2. OS is Windows XP Pro SP3, HDD is Western Digital WD5001AALS-OOL3B2 SATA, memory is Crucial 1GB DDR2PC2-5300 unbuffered X2. Video card is ASUS Nvidia GeForce 8400 GS 512MB. The HDD acts like it's working it's self to death, when running applications or even games there will be a split second pause where things seem to hang and that's when the drive starts running. I've used WD Diags on the drive and it checks okay, this is driving me nuts, any ideas would be greatly appreciated. Thanks in advance.
I'd like to ask some basic questions, and request you run some standard stress tests. Some of this will certainly feel like repetition, but please give it a try so we can go down a known path.
1) Clear CMOS and reload factory BIOS defaults. (Pull plug from wall, remove battery from mobo, press case power on switch a few times (to drain psu), go grab a cup of coffee, put the battery back in, plug back in the wall, boot up and go straight to BIOS. Load and save factory defaults ONLY changing something if its absolutely necessary to get your PC to run (like integrated vs PCI graphics or RAID).
2) Download Prime95, Furmark, and CPUID's Hardware Monitor. Run Prime95 for an hour with "Detect Rounding Errors" checked, and watching temps with HWM. Do not let cpu temp go over 75C. Run Furmark and do not let the gpu exceed 95C. Then run both simultaneously. Let us know what happens. Note that this test will stress memory more than memtest86. Download links are here:
Thanks for your quick reply, I downloaded the programs today and will run the tests tomorrow. One thing I forgot to mention in my post is that I completely remove power from the computer when I shut it down and have the bios set to restore power after power loss that way I shut monitor, speakers, etc all off at the same time. If I shut the computer off now, when I turn it back on, all drive lights come on as normal but no post, no beeps. I have to hit reset 3-4 times before I get a post beep and a post. The computer has to be off for a period of time though. If I turn it off then back on in a few minutes it doesn't do it and posts normal. I've tried another known good power supply, same thing. I'll let you know about the tests. Thanks
All my equipment is attached to power strips that I switch off and power is supplied through a 10KVA isolation transformer. I've been leaving the PC run 24/7 at this time as I'm afraid I won't be able to get it to boot back up. Sorry, I over looked your other questions. CPU: Intel SL6WU 512K 3.0GHz HT PSU: MGE Model WIN-600PS 600W. The mobo I purchased from Newegg 7/30/2009 and it went bad at the end of the year and was RMA Biostar under warranty I think the first part of this year, can't find the paperwork on it so I can't be exact. So far on the tests I've ran Prime 95 and it ran 10 hrs with no problems. I've ran Furmark stability test with xtreme burning mode, displacement mapping, and post FX with no porblems but haven't ran the two together yet. No heating issues at all, not even close, and no lockups. Note: The first mobo went bad via my fault, I was running a 3.2GHz 1G HT Extreme processor in it and at that time didn't realize that the mobo didn't support a unlocked processor. I should have checked. I've not overclocked, not ever, so that's not an issue. Thanks
First, if you do decide to shut the PC down, just make sure to let Windows exit gracefully before turning off the power strip switch.
Second, one of the problems folks who "turn off the power switch" sometimes have is premature failure of the CMOS battery. I don't know what your PC usage pattern is, but if the periods of "Off" are far longer than "On", the battery is at least cycling more than most users would see.
So a corrupt CMOS/bad battery could be a factor. I'd clear CMOS, restore defaults, change the BIOS parameters that need to be changed to get back to where you were, and leave the PC plugged in for a while.
Third, I'd suggest you check the power and data (both ends) cables to your disk(s). This week we had a user with a "slow" PC who found a bent pin (IDE) on a connector to a disk he wasn't using. Disconnecting that drive (and later straightening that pin) solved the problem. The theory is some form of I/O error was sensed there and caused a lot of disk controller overhead, slowing everything down.
Fourth, if you have been turning off the PC by fliiping the power strip switch, you likely have damaged data on the disk. These broken clusters could concievably be causing considerable overhead. Chkdisk followed by defrag may help. On chkdisk:
Thanks again for your reply. First off this PC is used extensively every day so any off time or power removed time is not real long, maybe 8 hrs at the very most but more like 5-6 hrs. The CMOS battery is currently holding at 3.37V. Any time I remove power is only after shutdown is complete and everything is off and the HDD is parked. But here's the thing though, even if I just shutdown and leave power applied and the PC sits with it's standby voltages applied, when I hit the power switch to turn it on it does the same thing, drive lights, no post beep, no post. Also when I'm running applications or playing WoW and the HDD starts reading and writing everything gets jerky for those milliseconds that the drive is running. As far as the CMOS goes, I did reset it, powered down and removed power and then the battery, pressed the power switch several times to drain any transient voltage. Reinstalled the battery and powered the PC back up and reset the CMOS to defaults again. I did all this before I ran the tests you asked me to run. I also have ran chkdisk and also WD Diags on the drive and both showed no errors. All the cables is the first thing that I checked in the beginning when I started having problems but the HDD is SATA so it really doesn't have pins as per say, more like tabs. I really appreciate you taking your time to discuss this with me, this thing is driving me nuts!! You've been far more informative that Biostar tech support was. This problem also affects the audio some but not real bad but when windows is loading and you have your windows start music, it cuts in and out (chops) as windows loads. I've also removed and reinstalled the buss mastering drivers, idk, I'm stumped. Thanks.
OK, the basic stress tests show passing grades. Clearing CMOS and reloading defaults didn't help, eliminating corrupt/incorrect BIOS settings. The same symptoms occur with a known good psu. The graphics card does not seem to be involved. The OS has been freshly installed, and the same symptoms appear.
This leaves only three candidates, the disk, the SATA cable (or the way it is routed), or the mobo. (Yes, I know you have run chkdsk and WD Diag.) If this were my problem, I'd do this:
1) Replace the SATA cable and retest. Assuming symptoms recur . . .
2) Use old and new cable in a different SATA port on the mobo and retest. Assuming symptoms recur . . .
3) Clone the "C" drive to a spare HD and retest. Assuming the symptoms recur.
4) Replace the mobo, ideally with a different brand (Assuming you are now running a cpu clearly supported by the Biostar BIOS).
After telling Biostar the list of things I've done, I don't see how they can deny an RMA on the mobo. If its out of warranty or they decline, I'd put in a Gig or Asus board.
Thanks again for your time and advice. ASUS has always been my choice of board but they didn't make a retro board that I know of for socket 478. I got this board trying to save money and reuse what was still good but now I've been unemployed for over a year and a half and now am disabled so anything new is out of the question. I'll try your last suggestions and pray one of those correct the problem but if push comes to shove, I'll just shoot it. Thanks again.