On Fri, 28 Aug 2009, Daniel Hill wrote:
I recommended memtest86 xor shunt files around for two reasons... *
so you can get a fast reliable check of "did that fix it?" *
hardware can go flaky for many reasons other than power supplies.
memtest86 passed with current PSU, also how would memtest86 fix my
problem?
I have no expectation that memtest86 would fix anything... it is
merely a test.
The fact that it past that test makes me suspect it is not the power
supply to the mother board that is the problem. (Unless you CPU is a
multi-core)
I recommended the test for one reason... the fundamental rule of Good
Engineering is when fixing hardware (or any problem) you need a good
fast and reliable test of when it's broken or not.
You original problem description didn't sound like a good fast
reliable test... but a complex mishmash of system activities that
could be a fault in the test, or triggering several faults not just
one, or could sometimes work even though all faults are still present.
If you don't have a good fast reliable test, you are subject to wild
superstition. (I changed XXX and it fixed it.... oh no, the problems
back, so my XXX must of broken again! Nope, not likely, more likely is
your test of "fixed" was unreliable and it just happened to work once
or twice when you changed XXX and there was and is nothing wrong with
XXX)
You have a hypothesis that the problem is your PSU, unless you have a
good fast reliable test you will never prove it.
Memtest is Good. It tries to isolate the fault even to the point of
switching of cache for some of it's tests. If memtest works, you can
have confidence in the motherboard, the cache, the cpu, the memory and
the power supply to it.
That still leaves the disk controller, the disk and the power supply
to the disk.
John Carter Phone : (64)(3) 358 6639
Tait Electronics Fax : (64)(3) 359 4632
PO Box 1645 Christchurch Email : [email protected]
New Zealand
