i am afraid that these figures are the same for booting and non booting.
the pwr supply that is fitted, is an etech mtp400,
Same skills used for electronics is also how I learned to fix so
many automobiles AND discover why those failures were directly
traceable to management. Cars designed by bean counters who did not
even know how to drive. What happens when all cars of that year are
built with no valve guides? I did not just learn how to do a valve
job. I asked some embarrassing questions (as I do with computers) to
learn those missing valve guides were directly traceable to Henry Ford
and his cost control programs. Learned because one should never just
fix something. One should also know why so that the problem is
solved; symptoms are not cured.
Back to your current problem. If those same voltage readings occur
during no boot, then power supply is not the component causing boot
failure. That excessive voltage on 12 volts during power up would not
harm electronics. But due to a slow response by multimeters, then
possible is a 12 volts well exceeding 13.76. Fact that 12 volts is
excessive on startup means the power supply should be replaced - even
if it does not cause boot failure. But the immediate reason for your
boot failures is elsewhere.
The list of parts that can shutdown a computer is short. Memory,
CPU, power supply system, sound card, video controller, and some
motherboard functions. But in your case, the shutdown is also
occurring (if I remember correctly) before Windows boots (eliminating
sound card) and sometimes before BIOS even executes a memory test
(eliminating memory). We have eliminated power supply system as a
suspect. CPU works just fine once system is booted (eliminates CPU).
Video controller only works in the simplest (Mode 3 or 7) operation
when booting. Video card is possible but not a likely reason for
failure. Only left are some motherboard functions (with a remote
possibility that video controller is the suspect).
Why is the motherboard defective? Did that spiking 12 volts create
a motherboard failure? Well we would know better if the 12 volts
could be monitored with an oscilloscope. That probably will not
happen. You could try any other video controller just to eliminate
that remote possibility. However it appears you have two suspects.
One that may or may not have been causing your boot failures (power
supply) and another that is the only remaining suspect (motherboard).
How to select a power supply? Do not get lost in the hype over
watts. 500 watts is typically gross overkill. But those who don't
know electricity will hype watts and dollars as the only power supply
metric. Those who learned from free markets have better benchmarks.
Any decent product will provide a long list of numeric specs in
writing. Only 1% of us will understand those specs (just like cars).
But if those specs are provided, then the 1% who know can 'blow the
whistle'.
How to sell power supplies that do not conform to minimum
standards? Hide the numbers. Don't provide the long list of specs so
that the 1% have no facts. Then many functions required even in 1970
and required by Intel specs can be forgotten to increase profits.
Some examples of numeric specs that should be in writing:
Specification compliance: ATX 2.03 & ATX12V v1.1
Short circuit protection on all outputs
Over voltage protection
Over power protection
100% hi-pot test
100% burn in, high temperature cycled on/off
PFC harmonics compliance: EN61000-3-2 + A1 + A2
EMI/RFI compliance: CE, CISPR22 & FCC part 15 class B
Safety compliance: VDE, TUV, D, N, S, Fi, UL, C-UL & CB
Hold up time, full load: 16ms. typical
Efficiency; 100-120VAC and full range: >65%
Dielectric withstand, input to frame/ground: 1800VAC, 1sec.
Dielectric withstand, input to output: 1800VAC, 1sec.
Ripple/noise: 1%
MTBF, full load @ 25°C amb.: >100k hrs
Such power supplies should also costs about $60 retail. That does
not say a $60 power supply is sufficient. Such reverse logic does not
apply. It only says the full retail price of a supply containing
required functions would be at least $60.
A more than sufficient supply could be 350 watts. That same supply
may be rated by another manufacturer as 500 watts. And neither
manufacturer was lying. It's called number games. The game works
only when most customers (many who would make recommendations) don't
know technology; do only as a bean counter would do; compare numbers
without knowing what those numbers mean.
Useful are current (amp) numbers for each voltage. Start with
current numbers for the original supply. Based upon your meter
numbers, the +5V (red wire) had a heavier load (compared to its max
current number). Your new supply should meet the same current numbers
listed for 3.3, 5, and 12 volts. And hopefully, the new supply has an
even larger number for the 5 volt current spec. If not, well, at
least the new supply should have same or greater current numbers
compared to the original supply. Ignore the watt number. Instead
review current numbers for each voltage.
As noted earlier, when you provided numbers, then others with better
knowledge can add additional facts.
When the supply is replaced (do it before replacing motherboard),
then take voltage readings to confirm that new supply is functional
and sufficient - the two minute procedure - and does not spike the 12
volts.
A final observation. 12 volts powers another power supply system
installed for and located adjacent to the CPU. Those upright
cylinders are called electrolytic capacitors. One type of failure
means the capacitors bulge - top pushes out. A visual examination
may provide some useful information. Pictures of a failing capacitor:
http://www.badcaps.net/images/caps/kt7/image005.png
Provided from numbers identify two suspects, better information for
selecting a new power supply unique to your computer's load, lessons
learned from the failure, and a shorter list of things to examine. As
I said early on, your posts will only be as helpful as informatiion
provided. Two minutes to obtain numbers results in much new
information.
please bear in mind that i do not have