[geeks] weird power supply behaviour

[Sheldon T. Hall]

 Saith the Mouse ...

> I've just had some weird behaviour I don't understand.
> 
> The scenario: power supply, peecee motherboard (fairly old - 
> it's a Celeron 400), and two disk drives, A, and B.  The PS 
> is way over spec for the MB and disk; the (dead) PS that came 
> out of the machine is rated 250W and this one is rated 400W.
> 
> With the MB and disk A connected, disk B not involved, 
> everything works fine.
> 
> Swap out disk A for disk B and the PS apparently goes into 
> overload shutdown - it powers up long enough for fans to 
> spin, but then drops power almost immediately.  Standby power 
> remains up; I have to kill mains power and wait for standby 
> power to drop to reset it.  This occurs even if I unplug the 
> data cable between disk and MB.
> 
> Okay, so that disk is just failed in sokme way that looks too 
> much like a short to the power supply, I figure.
> 
> But no.  I put disk A on and get it running enough to display 
> stuff on the screen, then plug power into disk B live - and 
> disk B spins up fine.
> 
> So I reassemble the disk-B machine, only I don't plug powe 
> rinto disk B.  I let the MB spin up enough to start 
> displaying, then plug power into disk B.  Power supply shuts down.
> 
> The PS is willnig to power the MB and both disks, but not MB 
> and disk B alone.  Whether data cables are connected or not 
> appears to make no difference, so it's not something being 
> incorrectly grounded through the data cable.
> 
> Any theories?  I have no theory to explain this.

I have heard of, but never seen for myself, power supplies which, with too
little load, act just as if they had too _much_ load.

-She