[Gllug] Power supply problems - question...

[Darren]

First post, so hello folks! :-)


Regarding power supplies...I'm wanting to build a low-power-consumption 
mini-itx box, just running Debian base.

Standard power supplies are a lot cheaper than the specialised 65 or 100 watt 
ones and I just want to confirm what I think I remember from GCSE Physics ;-) 
"Conservation of Energy" - If I use a 350w power supply to power this setup, 
it'll only consume whatever power the connected equipment draws (excluding 
losses through heat generated etc).

Does this sound reasonable?


Thanks for your time,
Darren


On Thursday 09 Oct 2003 6:45 pm, Chris Bell wrote:
> On Wed 08 Oct, Roger Whittaker wrote:
> > We have a machine which has two hot-swappable power supplies which slide
> > in.  It has started crashing - the BIOS contains a "Power management
> > setup" section, and at boot time there's a warning from the BIOS that
> > there's a power error.  Looking at the power management stuff in the BIOS
> > shows that the +5V line is at 4.5V or lower - it reports an error if it
> > goes below 4.6V.
> >
> > My questions are
> >
> >  - what might have caused this to start happening?  Is there anything
> > that might be casuing a drain on the +5V in particular?
> >
> >  - is it a good idea to try new power supplies?
> >
> >  - any other clues ...
>
>    Sorry, I have only just received your email, and have just walked from
> Olympia past the SUSE office on my way home.
>
>    Have you been able to measure the real voltages? It could be a
> monitoring system error.
>
>    A pair of power supplies sharing the load should have their outputs
> matched, so they may be adjustable. An adjustable control may have problems
> with corrosion or contamination.
>
>    Are both power supplies driving the load all the time? This may use a
> pair of power diodes to isolate the supplies in case either PSU fails, and
> I have had both diodes go open circuit together in a similar situation so
> that neither PSU was connected to the load. (Both power supplies and both
> diodes should have been capable of driving the entire maximum load).
>
>    If either diode goes short circuit it might pull that
> power supply down enough to trigger the fault monitor.
>
>    It could be a linking problem or dry joint outside of the power supply.
> Check both the positive and negative supply connections.
>
>    I have had to repair several burnt-out high power mains connections, all
> of which were in the neutral return links, not the live feeds. Not sure
> why, perhaps people take more care over the live connections.


-- 
Gllug mailing list  -  Gllug at linux.co.uk
http://list.ftech.net/mailman/listinfo/gllug