HD diagnostics; how to determine if an HD is faulty or not

[W@lly]

I've got a pentium 4 PC, socket 478 with an IDE Seagate 160Gb hard drive.
The client said that it would turn off, and on my part have noticed several BSODS, but the common one was DRIVER: IRQL Less or equal...


This pc has no extra video card and is using its integrated one, its an AsRock P4V8M motherboard.

First did swap the 2 sticks of RAM and did run Memtest+ for about 5:35 hours with no error...

I did hear some hissing noises that comes when the harddrive is accessed, at first I suspected that it was the optical dvd drive, but when I had removed it from the same ATA ribbon cable, the hissing noises still came.

So, thinking it was a failing HD, I ran a bunch of different HD diagnostics tools.

I have used Victoria, MHDD, Seagate sea-tools GUI version but all went fine and completed without errors.

I have also replaced the ATA ribbon cables and swapped from IDE1 to IDE2. it still did the noises and BSODs randomly.
While in windows installed G Smart Control and did a short test for 2mins went OK and when did try the extended test it always did not finish the test due to the BSODs.

So I said that maybe Windows was the problem and thus
confirmed with client to do a complete reinstall [it was with Windows 7 Ultimate 32-bit version] and did the same OS and another from mine since it still did BSOD.

I thought maybe the OS is too new for the motherboard since it has 2.40Ghz and 1Gb RAM, and decided to try Windows XP pro then, but still it did the BSOD whenever it did go to screen-saver mode or attached a usb pen-drive...

So what to try from now on? I will try to do a reinstallation onto another HD, this time a SATA hard drive and see what turns out.

What do you use to stress and bench mark a PC after windows installation? I used passmark Burnin software for now, any other recommendations?

Thanks.

 

[ShinyTech]

Try running Drive Fitness Test outside of windows

 

[FremontPC]

Look at the actual SMART parameters in gsmartcontrol (or Crystal Disk info or HDDScan):

Reallocated Sectors
Sectors pending reallocation
Reallocation Events
CRC error rate

Also, check the mobo for bad (swollen or leaking) caps.

I always pop the case open on these older systems and do a bad cap check with the customer present. If that looks OK, I check the SMART report when I get it back to the bench. No sense in doing a self-test if SMART lists funky sectors in it's report.

 

[Cambridge PC Support]

Have you tried swapping the PSU?

 

[W@lly]

Have you tried swapping the PSU?

Yes I did replace the PSU as well, still same results.

Will first lookup the SMART reports info data from Crystal HD and Gsmart control since I'm not familiar with meanings, I can repost here to help me out.

 

[FremontPC]

Good move, Wally! We'll keep an eye out.

 

[gikstar]

when did try the extended test it always did not finish the test due to the BSODs

Have you tried running the long test with the drive attached to your test-bench computer? If it bluescreens there your best off replacing it, which it sounds like you are doing by installing a sata drive.

It could be the motherboard is giving out. Have you tried running PC Check to test all of the hardware parts?

 

[PCX]

You should never try and do a diagnostic within the OS of the computer in question. Otherwise, you run in the the same exact issue as you have described. My suggestion is to get a Parted Magic disk or have a separate Linux Box to run gsmartcontrol and run both the short and extended tests. Before running the tests though, you should check for bad sectors. If there are numerous bad sectors (more than 20) than there really is no since in doing the short or long SMART test.

 

[W@lly]

So I have told you that I opted for another drive, a SATA one but unfortunately, the SATA drive was a western digital 500GB [16Mb Cache] Caviar Blue SATA 3, and this motherboard P4VM8 did not recognize it, no matter what I tried.. E.g. setting the jumper to make drive read at 1.5Gbps did not work, and tried both SATA ports, update BIOS, set to no-RAID, and disabled IDE controller to make sure it autodetects it, still no luck...

However back to the original subject, I opted for another HD, this time a WD 160Gb PATA. While doing the Windows 7 Ultimate installtion, it did a BSOD. And when it did finally install, it still gave random BSODs at the fresh installation!

So I took out the motherboard, and checked for some caps bulging, I think there is one a bit bulged, though not sure if its the cause for the problem....

Finally replaced the motherboard for an equivalent one, another AsRock motherboard, P4i65G model, and currently I am installing Windws 7 Ultimate on the customer's Seagate HD and found out that its the culprit for the hissing noise.

I think by this step I ruled out the problem which was the P4VM8 motherboard.... will update you later guys.

 

[FremontPC]

Wally -

You can always boot of an Ubuntu 9.x CD and use Disk Utility to check the SMART numbers on a drive.

Frequently a handy thing to have around.

 

[W@lly]

Wally -

You can always boot of an Ubuntu 9.x CD and use Disk Utility to check the SMART numbers on a drive.

Frequently a handy thing to have around.

What does that tool do exactly? I know about it yet I have always relied on CrystalDisk, and GSmart Control software for testing HD health status..

 

[YeOldeStonecat]

Age of the motherboard puts it in the bad bulging cap era...
The P4V family was based on the dreaded Via chipset...not the good Intel chipset. That alone is a huge reason to take that motherboard and toss it in the trash.

IRQ less than blue screens are pointing to drivers misbehaving in the RAM. Usually do to corrupted or incorrect driver, but sometimes due to bad hardware that they're controlling.

I doubt it's a faulty HDD issue.

 

[Magic Miguel]

If it's a drive smaller than 3TB I use Hitachi's Drive Fitness Test. I assure you and anyone else that it will find out if a drive is bad or not, regardless of if it was made by Hitachi or not. I'm a little disappointed that they've stopped development of it since Hitachi was sold off to WD.

Also if you can find the dump files you can throw that into the Microsoft Debug tools and I would say 80% of the time you can find the driver that's causing the problem and you can go from there.

Like what Ye said, I've seen countless problems like this only to find that the motherboard has bulging or busted caps.

 

[ReliefAshama]

With no concern for the data on the drive - boot to systemrescuedisk and use the badblocks tool that's part of the e2fsprogs utility.

 

[W@lly]

This thing is driving me nutz... I replaced the motherboard, swapped PSU and used another HD to make sure. Yet it is still giving me BSOD, mostly PFN list corrupt, NTFS.SYS and IRQL LESS OR EQUAL TO...

Like I already mentioned, I did over 5hours test with Memtest86 v4.10. I did run PC Check and pass mark Burn in Test, they all passed!!

Oh and by the way, I did install both Windows Xp sp3 and Windows 7 Ultimate....

I have noticed that BSODS happen when the PC is turned off for a long time... It then is stable after resetting it and run a bunch of tests, open applications on it, surf the web etc....

I am running Memtest again to make sure the RAM is ok. Regarding memtest, how many passes do you run or what is the minimum time recommended?

 

[kwest]

At my rate my client would be at about 800 for everything you are doing. Offer a data recovery and sell him a new system. No point spending all this time on that old of a machine.

 

[W@lly]

Now found the problem! It was the stick of RAM, although it passed memtest for over 8 hours! I have replaced it and voila, no more BSODS...

Somehow I can remark that this PC had a bit of every problem...

Mobo was changed due to its making noises, found one bulgy capacitor.
HD was replaced since the client's one made noises and put it as a backup drive onto an external casing.
RAM was replaced too, after several BSODS, POST not booting from correctly from HD or reporting that File system is corrupted. And while I believed that OS was corrupted due to all resetting the PC, I tried to install Windows 7 from scratch, from an internal optical drive, from a USB optical drive where it could not load, guess what, the RAM did not let these!

Phew, I must say that this thread is not about the HD, but a headache of a fault-finding PC!

 

[FremontPC]

"What does that tool do exactly? I know about it yet I have always relied on CrystalDisk, and GSmart Control software for testing HD health status.."

Ubuntu Live CD's are useful if you don't want to install anything on the hard drive before you see the SMART attribute report, or if you just can't boot from that drive. Also allows you to access the files on the drive if you just can't boot Windows, you can access the internet (assuming Ubuntu has the right drivers for your hardware) to download tools and so on. It's just good to have one on hand. It doesn't have to be Ubuntu. You can probably go through the threads about Linux distros on this site and try a few different Live CD's, find the one that works best for you.

Good job, finding the faulty RAM. How many times have I heard about sticks passing all tests, but replacing the RAM yields the cure.

 

[joydivision]

I prefer to run the diagnostics from the hard drive manufacturer, they are mostly all on the Ultimate Boot CD. Crystal Disk is handy for a quick diagnoses e.g if I have no reason at all to suspect a hard drive problem I will use Cystal as apart of my general routine checks, if that is all clear I won't bother with diagnostic scans.