I do see where you're coming from, Ian. BTDT>BTS! The most memorable
occasion involved excessive vibration of a brand spanking new IDE (PATA
to the younger generation) HDD and the shite audio quality pressing
created by Microsoft of their official windows 98 install CD, coupled
with a 24 speed CDROM which, unfortunately rather lived up just too well
to its model name, "Vibrant" all exacerbated by the PC being built into a
taller than usual "Desktop Tower" case made from cheap tin with an overly
well endowed drive bay stack that had an uncountable number of resonance
modes.
The first attempts to install windows all failed, eventually showing up
as a few bad sectors on the brand new drive when it finally occurred to
me to run the HDD diagnostics. Assuming the drive had been faulty to
begin with, I waited for my customer to RMA his purchase and return with
the replacement drive so I could have another attempt.
I'm pretty certain that I ran the HDD diagnostics on the replacement
before trying to install windows again so that when it failed yet again,
I was able to see that the attempt had caused two or three bad sectors to
become evident. It was only after this had happened that it occured to me
that the badly balanced Microsoft supplied installation CD had been the
source of a severe enough vibration to actually cause the HDD to create
bad sectors.
I burnt an image copy of the windows 98 CD and repeated the installation
using the much better balanced for high speed CD-R copy, eventually
completing the installation successfully, proving my thesis in the
process.
In the end, I got around the problem of CD induced vibration by
strapping the HDD with large plastic cable ties to the underside of the
PSU using four small stick on rubber feet on the top side of the HDD as
protectors/dampers since this was the one location that seemed least
effected by vibration from the drive bay cage (when you've only got
lemons, make lemonade).
Other times that I've seen HDDs suffer bad sectors and corruption was
when the 4 pin Molex contacts had become loose causing rather nasty
transients on the 5 and 12 volt connections, a condition the HDD
controller hadn't been designed to cope with (they do, however, cope just
fine with the rather more leisurely loss of supply voltage in the event
of mains power loss - instant voltage dips are just too much of a problem
to deal with).
The other, pretty rare scenario you're referring to, involves a failing
contact in one or more of the 20 or so active signal pins in the 40 way
interface connector. Since, unlike the SATA case, there isn't any ECC
coding of the data and commands flowing across the 16 bit wide data path,
not only is it possible for data transfers to become corrupted in each
direction, more importantly command words are also vulnerable, leading to
the possibility of a read command being interpreted as a write command or
else a sector address becoming corrupted and data being written to
randomly wrong sector addresses. Thankfully, properly connected ribbon
cables rarely cause such problems but when they do, all Hell lets loose.
As I've already mentioned, this sort of data/command corruption cannot
arise undetected in the case of SATA. In fact, no harm comes of literally
hot swapping a SATA connected drive with itself (i.e, you can disconnect
the SATA cable from the drive and then reconnect it several seconds later
without complaint from the OS virtually regardless of the data traffic
flow activity at the time).
Much longer and windows will pop up a warning message and wait patiently
for you to decide how to proceed, or at least that *was* the case with
win2k and even winXP before this stopped being true with win7 and
possibly Vista - I can't recall when MS decided to try and disguise the
PC as a "Magic Box" rather than a piece of 'High Tech Equipment' electing
instead to make it behave like a wounded animal that knew it was best to
fake the appearance of being in rude good health and thus not admit that
there was anything wrong - a strategy that produced a behavioural
response more akin to that of a PC suffering from a massive overload of
malware than one trying to deal with a simple 'technical hitch', the
reporting of which would have given the game away that the 'Consumer'
didn't possess a "Magic Box" after all but merely a collection of
extremely high tech parts held together by an OS written by a two bit
company that doesn't care one bit for its customers.
In this, I speak from bitter experience dealing with a win7 box that
showed all the hallmarks of a 'Mystery Virus' with which I spent almost a
whole week tracking down, a problem that would have become swiftly
apparent in the case of win2k and XP.
Only then did I finally realise Microsoft's true purpose in their
roadmap to perverting the original nature of the PC as a general purpose
personal computer to the level of a "Magic Box" designed to extract as
much revenue from the hapless consumer via the perverted business of IPR
ownership by large corporations capable of influencing Democratic
governments into enacting punitive laws against any transgressions that
would interfere with the cashflow arising from copyright protected
materials and ideas such corporate entities had perverted into a modern
day cash cow revenue stream.
It was Microsoft who first demonstrated how IPR could be used as a
licence to print money two decades ago and the rest of their partners in
crime have simply followed the money stream. It should have come as no
surprise that this was where it was all going to end.
Apologies for the rant. The point I was originally trying to make is
that SATA disks simply don't show the same symptoms of 'bad data
connectors' as their IDE predecessors. Bad data connectors will show
symptoms but they're not quite so inventively mysterious as those we used
to see with IDE drives, essentially being one of non-fatal
unresponsiveness, often cured (even if only temporarily) by a swift slap
to the side of the PC case.
The power connector, otoh, has exactly the same potential for data
corruption and bad sector generation as the IDE drives but, unlike the
data connector, the SATA power connector is a much more secure fit
designed to facilitate safe hot plugging/unplugging using higher quality
materials than the low quality 4 pin Molex connectors which can so easily
succumb to intermittent contact with the potential to produce some really
savage power supply transients so SATA drives generally only show
problems with the DATA cables and never, IME, with their power cable
connectors.
So, to address your conjecture, the problem most definitely had nothing
whatsoever to do with 'Bad Hardware' and everything to do with 'bad iso
image files' as I thought I'd made apparent in my OP. However, perhaps I
didn't stress this fact enough and left far too much as an exercise in
"Reading Between The Lines". If so, my apologies to one and all.
--
Johnny B Good