Prev: Power light blinking (amber color)
Next: Dimension 4700 Boot Failure at Checkpoint Ithr
From: S.Lewis on 9 Feb 2006 18:31

"Bradley Walker" <bawalkerREMOVE(a)THISmodemnet.net> wrote in message
news:SfPGf.13907$qg.3514(a)news01.roc.ny...
> The machine is only 2-3 years old at tops. It's got a Pentium 4, SDRAM,
> came factory preinstalled with Windows XP. The system has been running
> unformatted, unmodified since the day it was purchased. These power dumps
> just started happening yesterday afternoon for no reason (so the client
> says).
>
>

That machine is likely 4-5 years old. It's a P4/478/400mhz FSB system that
supports only PC2100/266, if that give you an idea of the timeline.

The good thing is that (if it is a PS) parts are common, even on ebay.


Stew


From: w_tom on 9 Feb 2006 23:16
The suspect 'system' includes power supply controller, power supply,
and power switch. Stuck contacts not quite disconnecting on switch.
Loose wire strand. Failing capacitors inside power supply contoller or
inside power supply (one of the reasons Dell was taking some major
charges recently), or just something inside the power supply. Get a
meter most anywhere - Lowes, Radio Shack, Sears. They are so
ubiquitous and so inexpensive that to have a few - just like a few
screwdrivers - is normal. A tech without a voltmeter? Do you also not
have screwdrivers? It's really not a valid excuse.

Without those voltmeter readings, then just wildly start replacing
everything. Without the facts, one can only shotgun. One cannot even
move on to other smarter diagnostic techniques that actually would
identify the problem.

Bradley Walker wrote:
> A client called me earlier today regarding their office computer
> sporatically shutting itself down. The lady at the office said the screen
> would go blank, all the fans and everything would stop functioning inside
> the tower. She could *not* power the system back on unless she unplugged
> the powercable from the back of the system.
>
> The problem started midafternoon and after troubleshoot all of the basic
> items such as surge protector, snug cables, etc. I decided to get the tower
> and bring it back to my office for a look over. Before I picked the tower
> up, the system was shutting down on her every 45 seconds. She barely had
> enough time to log into windows before the system would once again shut
> down.
>
> I brought the tower back to my office only to do windows updates, all the
> anti-virus/spyware scans I normally do on systems as well as a stress tester
> which ran for nearly for nearly 5 hours. I called the client back and
> suggested they could have a faulty surge protector. However when I went to
> my lab a few minutes ago, the tower was indeed turned off. I noticed the
> power buttom light was a light yellowish-amber color and pushing or holding
> that button did not do anything to turn the power off. Unfortunately I do
> not have any voltmeters on hand so I can not test the output of the
> powersupply. But before I run out and jump the gun on what to replace with
> this... could the powersupply be flakey enough to work for long stretches
> then die out like the symptoms above indicate?

From: User N on 10 Feb 2006 01:39

"Bradley Walker" <bawalkerREMOVE(a)THISmodemnet.net> wrote in message news:dCzGf.13790$qg.10236(a)news01.roc.ny...

> [] could the powersupply be flakey enough to work for long stretches
> then die out like the symptoms above indicate?

I've seen bad power supplies spontaneously shutdown, but IIRC they
did so within 5 or 10 minutes of power on. However, when working at
the board level I've seen bad capacitors run for a couple of hours before
acting up enough to cause a problem. Not fun.

I threw a multimeter on one of those flaky PS's in the hopes of seeing a
rail act up just before shutdown... didn't. I was too lazy to check PSON.
I just grabbed a spare PS and tried it, and when the problem went away
and stayed away, that was good enough for me.
From: S.Lewis on 10 Feb 2006 04:44

"User N" <usern(a)invalid.invalid> wrote in message
news:-_adnd7UHIeIqHHeRVn-vw(a)comcast.com...
>
> "Bradley Walker" <bawalkerREMOVE(a)THISmodemnet.net> wrote in message
> news:dCzGf.13790$qg.10236(a)news01.roc.ny...
>
>> [] could the powersupply be flakey enough to work for long stretches then
>> die out like the symptoms above indicate?
>
> I've seen bad power supplies spontaneously shutdown, but IIRC they
> did so within 5 or 10 minutes of power on. However, when working at
> the board level I've seen bad capacitors run for a couple of hours before
> acting up enough to cause a problem. Not fun.
>
> I threw a multimeter on one of those flaky PS's in the hopes of seeing a
> rail act up just before shutdown... didn't. I was too lazy to check PSON.
> I just grabbed a spare PS and tried it, and when the problem went away
> and stayed away, that was good enough for me.


But did you do so "wildly"? heh


From: PeterD on 10 Feb 2006 09:24
On Fri, 10 Feb 2006 01:39:26 -0500, "User N" <usern(a)invalid.invalid>
wrote:


>I threw a multimeter on one of those flaky PS's in the hopes of seeing a
>rail act up just before shutdown... didn't. I was too lazy to check PSON.
>I just grabbed a spare PS and tried it, and when the problem went away
>and stayed away, that was good enough for me.

You're unlikely to see anything on a digital meter (response times are
way too slow), and even an analog meter probably won't show anything
useful. For this you'd need a scope with storage, with a trigger set
to fire if the voltages go oout of spec. Not trivial to setup but
possible. Why bother? Replace the power supply, if that doesn't fix
it,then the MB is probably toast.

Sure you could have an intermittant short on the power switch lines,
but that is very unlikely...
First  |  Prev  |  Next  |  Last
Pages: 1 2 3
Prev: Power light blinking (amber color)
Next: Dimension 4700 Boot Failure at Checkpoint Ithr