If you are getting a SMART error then the drive is usually bad. The
controller on the motherboard could be bad as well, espcially since you
have had so many bad drives, which is either bad luck or your drive is
bad now because your motherboard controller is bad or there may be a bad
connector on your power supply. I've had a bad power supply or two in
only a few of my customer's computers, the bad power supplies would just
chew up hard drives and motherboards but would pass any voltagemeter
test I used to diagnose the power supplies with the power supply tester
which is for sale at most online e-retailers. So don't rule that out.
Try a different but known good power supply first, if possible, and see
if the smart errors go away.

Then confront HP, tell them that the power supply killed your
motherboard and is eating hard drives every 4 months if that's the case.
Beause it is really odd and unlucky to get so many bad drives. I'm
guessing that the second drive may be from an older system and hasn't
ever been replaced before and isn't having problems, so I think there's
either a problem with the power supply connector being used or the power
supply is no good and maybe the motherboard is bad. This is all
hypothetical because I can't actually see your computer and diagnose it
any further than what you have shared here.

I've had bad experiences with Maxtor brand hard drives. You should use
the disk utility that is provided on the hard drive manufacturer's web
site to determine if the hard drive is faulty.

Here are links to the big 4. Maxtor has been a subsidiary of seagate
for a while now, but I'm including it since some PC makers (*HP COUGH
CRAP COUGH*) like to put Maxtor branded drives into their systems still
(sometimes).

Maxtor - 'Maxtor - External Storage, External Drive, Hard Drive'
(http://www.seagate.com/maxtor)
Seagate/Quantum/Maxtor - 'Seagate Technology' (http://www.seagate.com)
Western Digital - 'www.westerndigital.com'
(http://www.westerndigital.com)
Hitachi GST & IBM (Deskstar) Hard Drives - 'Hitachi Global Storage
Technologies' (http://www.hitachigst.com)

You should use a power supply tester that can test all of the
connectors on your power supply and verify that the correct voltage is
being supplied. Also, HP does offer a hardware diagnostic software ISO
that can be burned to bootable CD-R (depends on model) that will
diagnose the hardware and give a usually effective result.

Orientation of the drives doesn't matter if the drives aren't jostled
around while they are spinning. I know that drives don't like to be
mounted at 45 degrees and flat mounting gives the best performance
results. The straight up and down mounted drives usually causes them to
make noises during disk activity that is more than the flat mounted
drives.

The little bit of vibration a hard drive makes during normal operation
shouldn't cause it to loose it connection with either the power
connector or data cable. If this is the case replace the data cable, if
the connector on the power supply does not snugly fit into the power
socket of the hard drive, replace the power supply, don't fart around
and try to "fix" a power supply or you may end up in the hospital. Just
my generic friendly warning, never alter or repair power supplies or
their connectors, just replace them as directed on their affixed
stickers.

 First, those are not hardware diagnostics.  Those are Windows
diagnostics.  It is an HP.  Computer manufacturer is more responsible
- provided comprehensive hardware diagnostics that give numbers and
detailed messages.  Without those numbers and messages, then the few
who can actually answer your question will not post.

 Only one 'disk drive computer' controlled both drives.  If either
drive is defective, then the one computer will report a failure.  Is
the defect in drive two but you keep replacing drive one (because
drive one controls drive two)?

 Second reason for failure are completely and always defective power
supply voltages.  Completely defective power supply can boot a
computer.  The naive then assume the power supply is good.  Power
supply causes strange things to happen once every month.  Does that
sound like your computer? 'Sound' means nothing. Again, the numbers.
Voltage measurements are the only way to identify or eliminate that as
a reason for a failure.

 Heat is not a problem.  Your computer must be in love when a room is
100 degrees F.  A computer that is failing in a 70 degree room does
not have a heat problem.  It has a hardware problem.  How to find a
completely defective machine?  Operate it even in a 100 degree room.
If it fails, a hardware problem probably exists - another reason why
more responsible computer manufacturers provide comprehensive hardware
diagnostics - to find that failure even when heat makes the failure
easier to find..

 Step back.  Is your problem with hardware or software?  Windows does
little to find hardware failures.  Windows works around hardware
failures - does everything possible to ignore those problems.  However
Windows also has system (event) logs so that months after a problem
occurs, then a hardware guy can identify the problem.  What (exactly
letter by letter) does each failure message in those logs say?  Just a
few suggestions that would go after a problem rather than its
symptoms.