On Tuesday, November 08, 2011 11:40:51 AM, Allen Weiner wrote:
> On Mon, 2011-11-07 at 22:22 -0500, Chris Knadle wrote:
> > On Monday, November 07, 2011 07:03:50 PM, Adam wrote:
> > ...
> > 
> > > The first time this happened was two weeks ago, and somehow after a
> > > few hours the problem, whatever it was, "fixed" itself.  This past
> > > weekend I powered it down for the first time since then, and it did
> > > not power up at all.  During that time, I used my DMM to measure one
> > > of the peripheral power connectors (the kind that would power a PATA
> > > drive), and discovered that when AC is applied, the +12V line rises
> > > within about half a second to about 0.67V and stays there.  When AC
> > > is removed, that voltage gradually drops to zero as the LED flashing
> > > slows down.  When the +12V line is steady at about 0.67V, the +5V
> > > line is steady at about 0.27V.
> > 
> > ...
> > 
> > > Obviously something is very wrong here.  My question here is, is it
> > > more likely that the problem is the power supply or the
> > > motherboard?
> > 
> > I'd say this is got to be a power supply problem.  I've recently found
> > out some very interesting things about power supplies, which I'll share.
> >  [This came up on the LIMARC "Tech Net" a few weeks ago, which had a
> > power supply expert on the radio that explained the follow details.]
> > 
> > First, non-air capacitors have a limited lifetime.  Here's where it gets
> > really interesting, though: a typical lifetime rating is around 2,000
> > hours. Yes, you read that right: in 24/7 use terms that's about three
> > months.  The reason is that over time, non-air capacitors slowly break
> > down such that the voltage rating slowly lowers.  When the voltage
> > rating finally lowers to the voltage that is in use, it arcs over and
> > the capacitor becomes a dead short.
> 
> The problem with Adam's PC may indeed be the PSU. But regarding the
> lifetime of capacitors, here is a contrary view on PSU life from some
> non-experts on the Anandtech Power-supplies forum:
> 
> It is from the recent thread "How long does a PS actually last?"
> http://forums.anandtech.com/showthread.php?t=2201228
> 
> Here is one comment from that thread: "I've seen computers from 1990
> still running fine. That would be a 20 year old power supply." There are
> other similar comments on that thread.
> 
> I own two year 2000 Dell PC's each of which has seen over five years of
> 8-10 hours per day, 6-7 days per week of use with no problems.

Hold on -- I didn't say /anything/ about how long a /power supply/ will last.

What I /said/ was concerning /stated rating/ of the longevity of /capacitors/.  
I'm very well aware that power supplies can operate for decades.  What I'm 
pointing out here is that there's something /confusing/ and contradictory 
going on concerning the MTBF rating of [ceramic|electrolytic|tantalum|oil-
filled|etc] capacitors verses how long the items those capacitors are /in/ are 
in use for, and something new I've discovered concerning the failure [mode|
mechanism] that those capacitors have.

  -- Chris

--
Chris Knadle
[email protected]
_______________________________________________
Mid-Hudson Valley Linux Users Group                  http://mhvlug.org
http://mhvlug.org/cgi-bin/mailman/listinfo/mhvlug

Upcoming Meetings (6pm - 8pm)                         MHVLS Auditorium
  Dec 7 - An Intro to Chef
  Jan 4 - Recovering the Brownfield: Revitalizing Open Source Projects
  Feb 1 - Home Networking Made Simple with Amahi Home Server

Reply via email to