OT: Kludging a PC's power-supply control

Vaughn Treude vltreude at deru.com
Fri Apr 7 18:31:44 MST 2006


P.S. If anyone has a used socket-462 motherboard they'd like to get rid
of, I'll buy it from you.  (I can probably find one on-line somewhere,
but I thought I'd ask anyway.)

Vaughn


On Fri, 2006-04-07 at 18:25, Vaughn Treude wrote:
> Hello,
> Here's one for you hardware gurus.
> 
> I've got a PC that has behaved badly on and off; shuts off sporadically.
> It's a problem that was occasional at first and has gotten worse as
> time's gone by. After a lot of messing around I've convince myself that
> it's a hardware problem. (There are absolutely no messages in software;
> the logs just end suddenly. And I've swapped out everything except the
> video card, a Radeon 7500 AGP, seems doubtful that's doing it, but I
> suppose it's possible.) Troubles are, (a) the motherboards's out of
> warranty, and (b) I really like the board otherwise, it's got the best
> on-board sound hardware I've ever heard. (It's an Abit, with a
> socket-462 AMD processor, if you wonder.) Yeah, I should've returned it,
> but it's so much hassle when it's mail-order and home built - take
> everything apart, take out the CPU and RAM, package it up, and lose the
> use of the system for a month until they send you a new one.  So if the
> problem is at all tolerable I live with it - foolish, but that's water
> under the bridge.
> 
> The deal is that I've always assumed this was the ACPI malfunctioning;
> when it shuts off it's hosed and you can't reboot it until you manually
> shut off the power-supply switch for 10 seconds or so - I'd guess that
> has to turn off the 5V standby power so something can be reset.  So I'd
> like to just shut ACPI off, but it appears that the BIOS just won't let
> you shut off this feature. (How obnoxious is that?  Is that some kind of
> "Energy Star" certification requirement?) So I Googled for a while and
> found some postings somewhere from someone who got the bright idea to
> bypass the PS-ON signal in the ATX power supply connector, and hold it
> low so the mobo can't turn the PS off.  As an experiment, I rigged a
> power supply cable extender to a SPDT switch which would allow me to
> either route the PS-ON signal normally or tie it to ground.  It's weird;
> I can turn it on via that switch but the system won't boot, probably
> because nothing tells the mobo that's what it's supposed to be doing. 
> So I figured that if I did the bypass switch after the system was on, it
> would bypass PS-ON, thereby preventing the sporadic shutoff.  That
> doesn't appear to work, as it shut itself off while I had the signal
> bypassed.  Now I wonder if they actually use the PWROK (AKA "Power
> Good") signal for this ACPI garbage.  Or is it possible that PWROK
> signal can also indicate some sort of sporadic problem with the power on
> the board?  (Kind of like the "Check Engine" light on your car, which
> can mean anything from "your gas cap's loose" to "your engine's about to
> explode.")
> 
> Oh well.  I think this board's ready for the garbage, but if anyone has
> any other suggestions, it would make for an interesting experiment.
> 
> Vaughn
> 
> 
> 
> 
> ---------------------------------------------------
> PLUG-discuss mailing list - PLUG-discuss at lists.plug.phoenix.az.us
> To subscribe, unsubscribe, or to change  you mail settings:
> http://lists.PLUG.phoenix.az.us/mailman/listinfo/plug-discuss
> 
> 



More information about the PLUG-discuss mailing list