I bought an IBM Deskstar hard drive in October 2000 after many hours of reading and research on various websites such as storagereview.com, tomshardware.com and others. The IBM Deskstar model DTLA-307045, had a reasonable cost per gigabyte and was a top performing drive in terms of transfer rate and low noise level. Or so I thought.
Two months later, in December 2000, the drive made some scratchy sounds. At this time, I was using multiple boot partitions and it was the TEST partition that was bad. I was able to boot the MAIN and GAMES partitions just fine.
I ran Scandisk with Surface Scan on TEST. Scandisk got to 13% complete (cluster 244,974 out of 1,789,755) when the noises started and the computer locked up.
Next, I downloaded Drive Fitness Test (DFT) from IBM and ran it. DFT locked up while detecting PCI IDE controller 0. Data Advisor from Ontrack ran okay and ran with some hiccups but otherwise passed. After a while, I determined C:\WINDOWS\USER.DAT was corrupted. I wiped the partition and restored it from a saved drive image.
Two months later, in February 2001, the noises returned. This time, Data Advisor's surface scan resulted in the "deathscratch" noise. I tried "wipe and restore" which resulted in "Attempt to restore image failed - Error 4". FORMAT C: from the DOS prompt also resulted in the deathscratch.
I attached the IBM "Deathstar" hard drive to the Primary IDE controller on the motherboard and ran IBM's DFT. This is what I got.
Analysing error logs (deathscratch sound)
"One or more bad sectors found. Please run Erase Disk to
repair the drive. NOTE: Erase Disk will DELETE ALL the data
on the disk drive. Return code 0x70"
"Serial # YMFM2855
Disposition 0x70 - Defective Device
Failure code 0x77
Test date 18-02-2001
Technical Result Code (TRC) for RMA 7077BD88"
I decide to bite the bullet and run Erase Disk.
July 2001 heard a beep and click-click during playback of an MPEG file. Did not hear it again when I played the file a second time.
September 2001 squeaks and beeps during a cut-and-paste in Picture Publisher.
DFT completes Quick and Advanced tests successfully.
November 2001 another beep/squeak during MPEG playback. No other problems.
February 2002 deathscratch returns once more. I start backing up to CDRs. For the first time since I have had the computer, some CDR have errors due to buffer under-run.
DFT recommends Erase Disk again. I get another hard drive to facilitate the backup and relegate the Deathstar to secondary storage only (video captures and downloads).
July 2002 a whisper-quiet deathscratch followed by a beep.
November 2002 up to now, the deathscratch would stop after a few cycles (click, click, scratch, scratch, scratch, scratch, scratch). Now, it doesn't stop until the computer is reset or the power is switched off.
DFT finds bad sectors again. "TRC for RMA: 7077ED8E"
I (finally) am fed up with the drive and decide to send it back to IBM. The 3-year warranty period will expire in August 2003. But I don't want to send the drive off until after the Christmas mail rush.
Deathstar Power On Deathscratch.Why wait so long before sending it back? The deathscratch has been such a huge problem that there is class-action litigation against IBM. Other unlucky Deskstar owners have returned their drives and reported receiving a refurbished model in return. If IBM is simply running DFT's Erase Disk and sending the drives out, I thought that if that's the case, I may as well run DFT's Erase Disk myself.
- Hard drives are getting cheaper.
- IBM's packaging requirements for returning hard drives is strict.
- It may cost more than it's worth
- But the bad sectors may be multiplying...
January 1, 2003: IBM gets out of the hard drive business and sells the hard drive division to Hitachi.
March 9, 2003: I finally get around to sending the d*** drive back. A Return Merchandise Authorization or RMA is requested from Hitachi Global Storage Technologies.
March 11, 2003: The hard drive is packed up and shipped via Purolator. Besides the Purolator packing slip, I also have to fill out an FCC Form 740 (Statement regarding the importation of radio frequency devices capable of causing harmful interference) and a Proforma invoice. Both forms require description of the goods including device model numbers and serial numbers and country of manufacture.
March 18, 2003: Hard drive is delivered to Hitachi in California.
March 21, 2003: I checked the status of the RMA - drive received, replacement not sent out.
I sent an email requesting that the replacement drive is sent via USPS (United States Postal Service) in order to avoid brokerage fees from UPS. Christina replies that Hitachi will cover all shipping charges.
According to CBC Marketplace, brokerage is included in shipping charge when the package is shipped by air. When shipped by ground, brokerage fee is incurred by the recipient).
March 27, 2003: received notice that my replacement drive has been shipped.
March 31, 2003: received my replacement drive. The package was dropped off on my doorstep since no one was home at the time. Is that dumb, or what? Luckily, my neighbour rescued it.
It was shipped from Malaysia of all places. It is sealed in a silver, anti-static bag complete with dessicant pouch. The hard drive and its shipping carton are both labeled "SERVICEABLE USED PART" and in spite of the MAR 03 sticker, the drive certainly looks used as I feared [photo: 60 kb] .
April 14, 2003: The drive is working so far but it has not been put through any heavy use. I'm keeping my fingers crossed.
July 4, 2003: I installed the drive into another computer to perform a backup. A single squeak was followed shortly after by the infamous deathscratch. Holding the drive vertically seemed to lessen the severity and frequency of the scratching sounds ("scratch, silence, scratch" rather than "scratch, scratch, scratch"). I shut down the computer and let the drive rest for a few hours. When I tried using the drive again, it seemed to work normally and as a test, I copied a few gigabytes from one partition to another with no problems whatsoever. But I still had doubts.
July 6, 2003: I ran Drive Fitness Test (DFT 2.10 Build 13, Feb. 8, 2001). The Quick Test gave two deathscratches duing the "Basic" test section, then completed successfully. The Advanced Test yielded the same result.
How can DFT report the drive is functioning properly when it makes these sounds? I downloaded the newest version of DFT (version 3.40 Build 03, Feb, 17, 2003) and ran it. Apart from a discrepancy in the cache size (reported 1916KB on one page and 1962KB on another page), the Quick and Advanced Tests in DFT v3.40 again yielded the same double deathscratch and result code 0x00.
July 9, 2003: With DFT v3.40, I used Corrupted Sector Repair, Erase Boot Sector and Erase Disk. No problems reported but the drive still makes scratching sounds during the Quick Test.
Hmm, strange. When I used Erase Disk before, the scratches disappeared. I decided to use Erase Disk in DFT 2.10 but I had no success. DEATHSCRATCH LIVES ON!!!
![[photo: 60 kb]](rma2.jpg){kind=link}