Archived from groups: alt.comp.periphs.mainboard.asus (
More info?)
In article <0DEpc.56518$536.9773871@attbi_s03>, "Marco415"
<marco415@comcast.net> wrote:
> I just installed this board, connected the ATX power connector to the board
> and power switch from case and CPU fan. When I turn on the toggle switch on
> the back of the power supply, everything will turn on and power up, but you
> can't turn off from the case switch, it will only turn off by switching off
> power supply switch. Tried to reverse the case power switch connector, still
> won't power down from switch. shouldn't the board turn on only when I use
> the power switch from the case?.
>
> - MM
The case power switch doesn't directly control the power.
The case power switch is momentary contact. You press the
button, and a pulse is fed to the logic on the motherboard
for as long as you depress the button. A short pulse is needed
to start things, and pressing for up to four seconds is required
as part of an unscheduled shutdown sequence.
The case power switch feeds to several chips on the motherboard.
For example, if the motherboard designer has designed the
mobo to shutdown on overheat or a CPU fan is missing, those
events have to feed into the logic. Same goes for the AGP_warn
circuit, which stops the ATX power supply from starting, if
an illegally keyed video card is being used.
The case power switch is latched by the motherboard logic.
The motherboard logic responsible for controlling things
runs from +5VSB. That is the voltage that starts to flow
once the switch on the back of the computer is in the ON
position.
The output of the motherboard logic, drives the PS_ON# signal
on the ATX PSU. The PS_ON# signal is active low and the signal
must be constantly present for the ATX supply to stay on.
On the power supply end, PS_ON# is pulled up to +5VSB, and
it is the motherboard's job to ground it, when all ATX power
outputs are desired.
So, anything which is causing the PS_ON# signal to be grounded,
would cause the behavior you are seeing. You could, for example,
disconnect the PSU from the motherboard, and then flip the
switch on the back to the ON position - the PSU should not
start when you do that, as there is nothing to ground the
PS_ON# signal. If the PSU fan starts to spin, when the switch
on the back is flipped, and all of the output cables are
disconnected, then the PSU is defective.
I've had one strange occurrence on an old motherboard here,
where one of my IDE cables was slightly loose, so not all
the logic signals were making contact. My computer would
start immediately when the switch on the back was set to
ON. Once the IDE cable was reseated, every thing went back
to normal.
Any time your build exhibits mysterious behavior, you
have to simplify the configuration, and then add one component
at a time, until the malfunction occurs. On motherboards
which have Voice POST, you can start with a bare board, and
listen to the Voice POST messages advance through the list
of messages, as you add CPU,memory,video card. Without Voice
POST, the simplest config you can do is mobo, CPU, memory,
video card, keyboard, mouse, to be able to test for a working
mobo.
You might also consider the "cardboard test". Assemble the
components, one at a time, while the motherboard sits on
a piece of cardboard. This eliminates a short caused by a
misplaced brass standoff underneath the motherboard, or
mechanical stress (bending of the board), as possible reasons
for the board to malfunction.
HTH,
Paul
Facebook
Twitter
Instagram