[Ham-Computers] Erratic Boot from CD

[nn6o]

Hi Jim,

I'm posting this to Ham-Computers and CC'ing you as the info may be useful to
others - hope it's OK.


Synopsis:
As a follow-up to Jim's previous posts about the boot problems he's
experiencing, he PM'd (personal e-mail) me additional details about the
history and issues with his Dell Inspiron 5160 laptop.  It's was recently
serviced at a shop for overheating and a possible drive failure.  Below are my
thoughts.


My thoughts:
Overheating laptops is fairly common - hair, dust, and other "crap" get
trapped in the heat sink fins/fan, and, depending on the design, also in the
interior of the laptop.  Better designs have the fans drawing air through a
duct so dust doesn't get drawn into the interior of the laptop itself, but
this is not always possible.  Easiest way to clean out the ducts is with a can
of compress air.  Use short (1 second) bursts of air through each duct in the
laptop (while powered off).  Blow air through each duct until the "exhaust" is
clear.  Most of the time, it's not necessary to take apart a laptop to do
this, but if you have easy access to remove the keyboard, doing so doesn't
hurt - again use short bursts of air to blow out dust/debris.

In the case of the Inspiron 5160, a little bit of research shows that it and a
couple other similar Dell laptops had design problems in relation to the
cooling system.  In fact, for these laptops (1150, 5100, 5150, 5160), there
was a class-action lawsuit filed against Dell that was settled
(http://www.lieffcabraser.com/cases.php?CaseID=93).   Similar claim in Canada
also settled (http://www.rochongenova.com/cases/dell.aspx).  Both of these
settled within the past 2 years, so it might be worth pursuing a reimbursement
claim.

I don't think there is any particular problem with the BIOS, but based on the
model history though, the laptop itself could be failing due to overheating
problems in general.  The Pentium-4 "Netburst" architecture was the hottest
running processor core Intel built (and was also the reason it never made it
past 3.2GHz).  So if the system had a cooling problem, then heat related
failures would not be uncommon.

As for the hard drive problems, the best software to test a drive is the
vendor-specific drive diagnostic utility - e.g. Western Digital's Data Life
Guard Tools, Seagate's SeaTools, Hitachi's Drive Fitness Test (DFT), etc.  Not
every vendor has a drive diagnostic available for download (such as Toshiba),
but other vendor's utilities often work with 3rd party drives with some
caveats (such as no lower level tests).  Most all of these utils incorporate
two universal tests based on the S.M.A.R.T. drive analysis standard.  They
will also show SMART status and possibly other parameters.  Use the diag tool
to run a "Quick/Short" SMART test.  If no errors show up with the "quick"
tests and/or drive problems are suspected, then run the
Comprehensive/Complete/Thorough/Long/whatever test.  At the end of the test, a
report is displayed - if problems are found, seriously consider replacing the
drive as most drives are near end-of-life once SMART errors occur.  Another
great diag tool is/was Spinrite by Gibson Research.  It's still available, but
it's fairly limited these days due to the translation layer built between the
physical geometry of the drive and the controller - only the drive vendor's
tools can bypass this and usually on the "factory" tools.  Spinrite will still
find and mark bad sectors, but if a drive has so many bad sectors that it's
out of "spare" sectors (every drive has hundreds, if not thousands, of spare
sectors), then the drive isn't worth saving as the number of bad sectors will
still continue to grow as the drive ages.

About S.M.A.R.T. - all modern drives have a built-in diagnostic program that
constantly monitors the drive for problem trends.  Many years ago, Compaq (now
HP) noticed that most drive failures were preceded by specific symptoms.  By
monitoring for several symptoms or a trending specific symptom, a drive
failure could be accurately be detected a majority of the time before the
drive fully failed.  This eventually became the Smart Monitoring And Reporting
Technology (SMART) which is standard in all hard drives (and SSD's) today.  If
a drive's SMART analysis detects an impending failure, it sets a flag and
notifies the computer it's attached to - it's then up to the computer to
report the impending failure to the user.  On most PC's, this occurs during
the POST (Power-On Self Test), that quick test the computer goes through when
you first turn it on.  When the BIOS asks if the drive is OK, the drive will
report a problem and the BIOS should generate a 17xx error before booting into
the operating system.  Unfortunately, some system vendors (such as Dell) are
idiots and have SMART reporting *disabled* by default.  Why?  Who knows - may
be they don't want the customer to know a drive is failing so they don't have
to deal with support calls.  Regardless, if SMART reporting is disabled, then
the user will never know the drive is failing until it actually fails.  So, I
suggest that you go into your system's BIOS settings and make sure that SMART
reporting is enabled, esp if you have a Dell system.  On the next cold-boot,
it should check the drive and report an error if the drive is failing.  BTW,
SMART can be disabled on the drive itself, but this is rarely set
intentionally.

So, as far as this laptop is concerned, it's definitely time to retire it.
Due to the cooling issues, cold-solder joints and failed components could now
be an issue - it might be working one minute, but fail the next.  A used
replacement drive in the 40GB range would be about $30.  You can find a nice
Intel "Core" or "Core2" based laptop these days for about $250 used and it
should be capable of running Windows 7 without a problem (1GB RAM or more
recommended).  If you do intend on running Win7, I highly suggest getting a
laptop using an Intel based chipset as Win7 support is better.  AMD/ATI's and
Nvidia's support of Windows 7 on older chipsets can be rather lacking.  We're
talking core-logic here, not the Video - that's a different, but similar
issue.

I hope I've answered most of Jim's questions.  There was a question about
WiFi, but he solved that with a higher-power adapter.  If anyone has any Q's,
please feel free to post them.  I hope people find the info here useful.


73,

  - Aaron, NN6O
    {nn6o}@arrl.net