Another dead power supply

[Lennart Sorensen]

On Tue, Dec 06, 2005 at 08:19:34PM +0200, Peter wrote:
> You could introduce some metrics using f.ex. (power * known_life)/price, 
> otherwise it's comparing apples with prunes. A brand name PSU would have 
> to run 4 times longer than a no name cheap one on account of the price 
> difference alone. That is equivalent to roughly being twice as large or 
> running 20 degrees (Celsius) cooler (imagine the fans this requires, 
> given standard PSU size).

How many $s is a stable system worth to you?  I have seen systems that
would crash randomly and corrupt the filesystem once in a while, which
in the end were fixed by replacing the power supply.  So while the power
supply was "working" (it was delivering power) it wasn't working very
well and was wasting a lot of time on reinstalls to fix the damage it
was causing.  Probably wasn't good for the components in the system
either.  Not delivering power isn't the only failure mode of a power
supply.  Delivering the wrong power may be even worse.  Had it simply
not worked at all, it would have been much quicker since then the power
supply would have simply been changed rather than blaming windows (the
machine had win98 at the time) for being unstable.

> This is very roughly estimated, based on the fact that electronic 
> component life (esp. capacitors) halves with each 10 degree C 
> temperature rise. To put this into perspective, 4khours 105degC caps 
> (this is very high quality for todays standards) run at 55 degC will 
> last about 14 years. Put in 2k hours caps (run of the mill) and run them 
> at 65 degC and they will last at most 3.5 years. Put in standard 85degC 
> caps with 2k hours and it will last just under two years at 55 degC. The 
> actual life of the unit is influenced by all of its parts, so it will 
> last less than calculated like this.
> 
> So imho barring any drastic departures from the laws of physics or 
> severely flawed design, temperature is the most important factor in PSU 
> life.
> 
> Of course a surge filter is a very good investment if the PSUs are to 
> die of old age ONLY.

Lennart Sorensen
--
The Toronto Linux Users Group.      Meetings: http://tlug.ss.org
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://tlug.ss.org/subscribe.shtml