Re: Computer power failure
From: w_tom (w_tom1_at_hotmail.com)
Date: Thu, 18 Nov 2004 11:08:25 -0500
If that procedure to verify power supply integrity is too
complex, then you are reading too much into it. You take a
meter. You put meter leads on black wire and yellow wire
(where power supply connects to motherboard). Number
displayed on meter is recorded on paper. Repeat this for
orange wire, red wire, etc. When done, compare those recorded
numbers with numbers in chart. If numbers are within (upper
3/4) limits, then power supply is just fine - move on. If
numbers are below limits (or within bottom 1/4 of those
limits), then power supply is probably slowly failing. Its
that simple - only two minutes.
Again, it takes many times more minutes even to swap a
supply. Do not start wildly replacing things since that also
can complicate your solution. Procedure to verify power supply
integrity is that quick, that comprehensive, and that simple.
Nothing in those procedures is complex (except if you are
trying to learn too soon why you are doing things). Don't get
caught up trying to understand why you are doing that.
Learning the whys comes later with your next questions. Do
the procedure and don't even look back.
Also ignore nonsense about removing dust. Computers are
designed with significant dust accumulation already expected.
Furthermore, those who install too many fans to fix defective
hardware have excessive dust accumulation. If heat was a
problem, then when hair dryer (on high) was directed to a
defective component, component would fail immediately. Why?
Everything in that computer must work just fine even when room
temperature is above 100 degree F. OS also will not cause
this. If OS did, then fact would probably be posted in event
You have a perfect situation to learn which component caused
power failure. A power supply system is three components: the
PSU, a controller on motherboard, and power switch. Again,
get the meter. You cannot see electrons. Record voltage
readings of purple, green, and gray (etc) wires when power is
off AND when power switch is pressed AND when computer is on.
If you don't understand what those readings mean, then post
them here. You problem is that you are starving everyone
(including yourself) of important facts. I can take you far
beyond what other posters can accomplish BUT you must not
starve me of facts. Get the 3.5 digit multimeter.
Blindly follow the procedure in those previous posts. If
you don't understand, then post numbers here. Collecting
those numbers should take but a few minutes. You will learn
how to fish rather than just be feed fish. IOW you will
accomplish far more than just fix a computer.
In the meantime, start the process of getting additional
free tools. For example, every responsible computer
manufacturer provides comprehensive diagnostics on their web
site - for free. If manufacturer is not so responsible, then
start collecting freeware diagnostics such as memory (Memtst86
or Docmem), components (ie go to hard drive manufacturer for
his diagnostics), etc. Once we establish power supply
integrity, then we can move on to other 'usual' suspects. We
must first verify power supply integrity with the 3.5 digit
multimeter long before we can move on to other 'usual'
Some obvious questions - your computer is plugged into a
properly safety grounded three prong outlet. All connected
peripherals share a common safety ground. You don't have
static electricity problems - the building has a humidifier.
No reason to assume anything. Computer repair is really quite
straight forward. You replace a part because part is
obviously defective. In but two minutes with a 3.5 digit
multimeter, you have all the facts necessary to say whether
power supply is good or bad.
If you do start looking at power supplies, then you are
asking for more and future problems if the supply is a
discounted (less than $60 retail) type. How do they lower
price? Forget to include essential functions that were even
defacto standard 30 years ago. Its called dumping. And so
many North Americans have so little computer knowledge as to
buy these 'defective by design' supplies.
> Hi Tom,
> Thanks for your reply. I went and read over the postings from the
> links suggested. I'm sorry but I did not understand any of it.
> I'm not too familiar with power supplies and hardware of that
> There are no errors in event viewer. I've gone as far as dusting
> the fans to no avail. I unhooked the power supply from the
> motherboard and replugged it in. Still nothing. The computer
> has cut off twice in a 45 min. time frame. The first 20 min and
> the second 10 min.
> I'm beginning to rule out an overheating problem because the room
> the computer is in is very cold and I took the side casing off to
> see if that would help the inside breath a little. I noticed that
> the green power light on the front of the computer remained on, but
> the power supply light on the back wasn't. I let the computer sit
> for 15 min. in case overheating was a factor and then tried to turn
> it back on but it did not power up. I again had to unplug the cord
> from the back of the power supply, reconnect and then power on
> the computer.
> I don't know if the O/S could be causing this. Could it? Should
> I now assume I have a faulty power supply? Any other suggestions?