What else does power to mobo actually do?

  • Thread starter Thread starter Franklin
  • Start date Start date
F

Franklin

Why does the motherboard still get power supplied to it even after
the system is powered down?

I know that the front switch on the system box is just a signalling
switch going to the mobo in order to request the mobo calls for full
power when starting up (PS_ON signal to pin 14). And also to ask
for the mobo to kill power when powering down.

But I kind of get the feeling that there are other things which
remain energisied and/or partly functional while the power supply is
not fully disconnected. Maybe the Ethernet LAN card in home PC for
connecting to a cable modem is kept alive for some reason.

Can anyone tell me if this feeling of mine is true and what sorts of
other things might be kept alive by this near continuous power.

(I am thinking mainly of a home user connected to the net with no
other PCs attached.)
 
Why does the motherboard still get power supplied to it even after
the system is powered down?


All PCI card receive power when a PC is 'off' - this allows functions
like Wake on LAN, Wake on Keyboard to function.
 
Franklin said:
Can anyone tell me if this feeling of mine is true and what sorts of
other things might be kept alive by this near continuous power.

The 5v standby line, which should be the only thing active (Pin 9)
supplies power to the motherboard power-on logic, which includes things
like wake-on-lan.

Contrary to other people, I cannot find any mention of standby power in
the interface description for PCI slots. (Can anyone confirm this)

Cards that can wake up a system (by generating power management events
such as wake-on-lan, wake-on-ring etc.) presumably do so via some other
means (and all of the network cards I have with WOL are either built
into the motherboard, or come with a three-wire WOL cable, which
provides +5VSB, GND, and /PME ).

The system clock is also supplied with +5VSB on most board designs
(mainly to increase battery life) - You'll discover that when your
battery is dead, the system will keep time and CMOS settings even when
powered down, provided standby power is still available.

The larger chips (north and southbridge) are not likely to be (fully)
powered on, as they would draw too much current. Most +5VSB lines will
only supply one amp, and cheap PSU's probably quite a bit less.
 
Franklin said:
Why does the motherboard still get power supplied to it even after
the system is powered down?

Because you left it switched on at the wall! It is only doing the same
trick as your TV, VCR, Satellite box, Freeview box etc. Anyway, if
the m/b didn't get power how would you get it started again? (This
is occasionally a fault condition, BTW).
 
Jim Howes wrote:
8><......................................
Contrary to other people, I cannot find any mention of standby power in
the interface description for PCI slots. (Can anyone confirm this)

Cards that can wake up a system (by generating power management events
such as wake-on-lan, wake-on-ring etc.) presumably do so via some other
means (and all of the network cards I have with WOL are either built
into the motherboard, or come with a three-wire WOL cable, which
provides +5VSB, GND, and /PME ).
http://www.planetanalog.com/printableArticle.jhtml?articleID=12801381
 
Why does the motherboard still get power supplied to it even after
the system is powered down?

I know that the front switch on the system box is just a signalling
switch going to the mobo in order to request the mobo calls for full
power when starting up (PS_ON signal to pin 14). And also to ask
for the mobo to kill power when powering down.

But I kind of get the feeling that there are other things which
remain energisied and/or partly functional while the power supply is
not fully disconnected. Maybe the Ethernet LAN card in home PC for
connecting to a cable modem is kept alive for some reason.

Can anyone tell me if this feeling of mine is true and what sorts of
other things might be kept alive by this near continuous power.

(I am thinking mainly of a home user connected to the net with no
other PCs attached.)

5VSB keeps the logic available to turn on the system and
keep the clock from draining battery.

This includes any PCI devices triggering power on, a bios
setting for turning on at a user-specified time, a setting
to turn on (or not, or same-state) after a power failure,
and triggering system on by any other devices the
motherboard manufacturer implements. They may implement a
WOL (wake on lan) header for a network adapter, but more
commonly on modern systems the network adapter uses a PCI
feature to do this. 5VSB and sometimes 3V-derived (via
motherboard or rarely from proprietary systems) power to the
PCI slots enable this.

For example, it's now more common to be able to turn a
system on by PS2 or USB keyboard or mouse. TYpically there
is a motherboard jumper to determine if the main 5V rail or
the 5VSB rail is used to power the USB and/or PS2 ports,
with the 5VSB being necessary for the device availability to
turn system on.
 
Why does the motherboard still get power supplied to it even after
the system is powered down?
Because the ATX power management requires it.
I know that the front switch on the system box is just a signalling
switch going to the mobo in order to request the mobo calls for full
power when starting up (PS_ON signal to pin 14). And also to ask
for the mobo to kill power when powering down.

But I kind of get the feeling that there are other things which
remain energisied and/or partly functional while the power supply is
not fully disconnected. Maybe the Ethernet LAN card in home PC for
connecting to a cable modem is kept alive for some reason.
Correct. Wake On events.
Can anyone tell me if this feeling of mine is true and what sorts of
other things might be kept alive by this near continuous power.
It is true but why do you ask? Are you worried about someone hacking
your PC?
 
NBT said:

Aha, well spotted!.. (I went looking for +5VSB, and didn't find it.
with DC-DC converters it isn't particularly suprising that I didn't find
it with my meter..)

The PCI local bus specificiation says of pin A14:
3.3VAUX: An optional 3.3 volt auxiliary power source delivers power
to the PCI add-in card for generation of power management events when
the main power to the card has been turned off by software.
The use of this pin is specified in the PCI Bus Power Management
Interface Specification.

Thankyou.

I note that the 3com ethernet cards I have do not manage to assert PME#
without the +5VSB, although it should technically be possible. I expect
this is because they would otherwise pull too much current from 3.3VAUX.
 
if you want to be sure, turn off the master switch at the back of the
powersupply unit
 
Jim said:
The larger chips (north and southbridge) are not likely to be (fully)
powered on, as they would draw too much current. Most +5VSB lines will
only supply one amp, and cheap PSU's probably quite a bit less.

When ATX was first introduced, your above comment about the +5vsb
current *might* have been correct, but not in the last 5 years. The
ATX specified required current for the +5vsb has slowly increased
to the currently specified 2.5 amps.

http://www.formfactors.org/developer\specs\ATX12V_PSDG_2_2_public_br2.pdf

Section 3.3.3
 
Jim said:
The 5v standby line, which should be the only thing active (Pin 9)
supplies power to the motherboard power-on logic, which includes things
like wake-on-lan.

Contrary to other people, I cannot find any mention of standby power in
the interface description for PCI slots. (Can anyone confirm this)

PCI 2.2 has 5v standby power on the bus.
 
Back
Top