Subscribe
to the RSS feed
Originally Added to Website: 14 Sep 2009
Last
updated: 11 Aug 2014
Unknown change - didn't find any when comparing to the stored copy.
This guide was written for a system that's getting power, but is having
hardware problems of some sort (and may have problems booting into
Windows).
I'm primarily interested in BSOD analysis, so this guide is written
from that viewpoint.
This may not be easy to do with a minor problem that's not easily
reproducible - but it's the only way that I've seen so far.
Just
FYI, the most common failure that I see is with video cards -
especially those used in gaming. Next comes RAM, then hard
drives, and then the motherboard. If it's a motherboard
problem
it can be very difficult to diagnose, so be prepared to take a lot of
time! Good luck!
I have a similar blog entry here (dated 24 Feb 2012): http://www.sysnative.com/forums/entry.php/2-How-to-Strip-Your-System-Down-in-Preparation-for-Hardware-Troubleshooting
H/W
Diagnostics:
Please start by running these bootable
hardware diagnostics:
http://www.carrona.org/memdiag.html
(read the details at the link)
http://www.carrona.org/hddiag.html
(read the details at the link) - Test ALL
of
the hard drives.
Also, please run one of these free, independent online malware scans to
ensure that your current protection hasn't been compromised: http://www.carrona.org/malware.html
(read the details at the link)
There are also free, bootable
antivirus disks at this link: http://www.carrona.org/malware.html#rescue
Run the additional free diagnostics here (if able to boot into
Windows): http://www.carrona.org/addldiag.html
Run SiSoftSandra freeware (if able to boot into Windows): http://www.sisoftware.net//?d=dload&f=sware_dl_3264&l=en&a=
Use the Ultimate Boot CD (UBCD) to run additional tests (this is a
bootable disk, so booting to Windows isn't necessary): http://www.ultimatebootcd.com/
Disassembly:
Unplug the system from the wall, and ensure that you remain grounded to
the metal chassis throughout the procedure.
Then, remove everything from the system leaving only these components:
Case (and remove it if you've got a
non-conductive
mat you can use for testing (and no kids or cats to get into it)).
PSU (Power Supply Unit)
CPU, cooler and fan
1 stick of RAM (presumed to be good
since you ran
the MemTest86+ utility earlier)
Video card (only one - this includes any
onboard
video)
Monitor
Keyboard (notice I didn't mention the
mouse - you'll
have to use keyboard commands for the first 2 steps!)
Floppy drive OR
CD
drive (not
both)
On the very first boot, disable any onboard devices (like the USB
ports) in the system BIOS.
NOTE:
At some point (if this is a hardware failure) you'll find the error is
caused by a particular piece of hardware. While this is most likely the
cause of your problems - please remember that it can be a combination
of things that have caused this. As such, a single device replacement
won't necessarily fix the problem.
Boot
Troubleshooting:
So, on to the troubleshooting! Just follow these steps until you get
the error. Then do it again to double check. Then replace the device
that caused it and all should be well with the system (see caveats
above).
- Try and boot from a DOS
floppy/CD. See if
the problem occurs.
- If it works, then add the mouse and
try
again. See if the problem occurs.
- If that works, then add the hard drive
and try
again. See if the problem occurs
- If that works, then start trying to
get Windows to
load. See if the problem occurs
Generally, this sequence should at least
allow your
system to boot up. If not, post what you've done on the forums and
someone will be able to help you isolate the problem.
Boot
Devices Troubleshooing:
This involves replacing each of the devices that was left attached to
the system. Obviously, you can't replace everything, but the
more
you do replace (and test) the more likely you are to find the
problem. If you had no trouble booting, then you can leave
this
to last - checking the Device Troubleshooting section first.
Here's my suggestions/comments:
Case - as a last ditch effort, ensure
that you test
outside of the case on a non-conductive mat.
PSU (Power Supply Unit) - get a PSU
tester (about
$10 US) - use HWInfo if you can get into Windows (free from
here:
http://www.hwinfo.com/
). The stripdown procedure will (generally) reduce the load
on
the PSU - so it should be obvious if this problem is caused by the load
on the PSU. OCCT (one of the video tests from above) has a
PSU
test.
CPU, cooler and fan - run lot's of CPU
stress
tests. Monitor temps and ensure that the fan/cooler is
working an
not clogged with hair, dust, etc.
1 stick of RAM (presumed to be good
since you ran
the MemTest86+ utility earlier)
Video card (only one - this includes any
onboard
video). Replace it. If you don't have a spare,
purchase one
from a store that allows returns.
Monitor - try another one.
Don't forget that
the cable can also be to blame!
Keyboard (notice I didn't mention the
mouse - you'll
have to use keyboard commands!) Get a spare, it's good to
have
one on hand
Floppy drive OR
CD
drive (not
both) - If you don't have a spare, either remove the
device from
the system or purchase one from a store that allows returns.
As an aside, it's always a good idea to have a spare network card on
hand. They run about $15 (US) from most major retailers - and
there's just no substitute if your network card dies while you're
working on your system!
Device
Troubleshooting:
Once that's done, start by enabling the
onboard
devices (one at a time) and rebooting. Check for the error and then
enable the next one.
After all the onboard devices are done,
then start
with the devices that plug into the mobo inside of the case. Do these
one at a time also.
Then start plugging in the external
devices, one by
one, until they're all in.
So, you've found the bad device and have
replaced
it. Either this makes the problem go away or it doesn't. If it fixed
it, great! If not, let's start the troubleshooting all over and try it
again. Since the case is stripped, the easiest thing is to run through
this hardware troubleshooter again.