"Jody" said:
What do you make of this:
Two different power supplies (350w, 20pin+4pin)
P5P800
3.0ghz p4 530(540?)
ATI Radeon 9000 128mb
Install windows xp&drivers.
So far so good.
Run standby mode
CPU Fan turns off (4pin)
-Celeron Fan (The orig fan only went 1000rpm and rec'd fan warnings on this
board; prev discussion)
Power Supply Fan turns off
Case fan(4 pin, ala hard drive connector) turns off
The Celeron Fan keeps up no problem.
CPU temp after 3dmarking like crazy is around 45C
Bios:
Internal CPU Thermal Control (On and Off both tried)
Q-Fan(On and Off both tried)
What is especially odd is the 4 pin case fan wouldn't have a sensor. So that
probably is a bad sign.
Wake it up, everything turns back on.
Have had some problems with the lan controller on wake up(intermitent(sp?))
Thoughts?
jody
Call me dense, but I'm not sure what the problem is here.
AFAIK, when a system goes into S3 standby, the power supply is switched
off via PS_ON# signal. The switch on the back of the computer, of
course, is still in the "on" position. You should find the LED on
the motherboard lit, indicating that +5VSB is flowing. The +5VSB
is used to power "wake-on" components, and also powers the all
important memory refresh, so that the system can quickly return
from suspend-to-ram state, to the fully running state. The CPU
and the fans are not powered, because the +12V is off. (In other
words, in the suspend-to-ram state, the CPU is no longer computing,
but the Northbridge is still sending signals to the DIMMs, and the
DIMM contents are valid, because regular refresh is happening. In
standby, you should not attempt to pull a DIMM out of its socket,
as it is still powered.)
All of the fans are powered by +12V. When in suspend-to-ram, there
is no voltage available to run any fan. When the computer returns from
suspend-to-ram state, the +12V will again become active. The
fans should spin. The +12V also powers the processor.
Fan speed is modulated by means of voltage. Some number of the
fans on the motherboard, may be connected to "Q-fan", a feature
that can reduce the voltage to the fan. Depending on whether a
motherboard features Q-fan or Q-fan2, determines whether just
one fan (the CPU fan) or a couple of fans, have the Q-fan
adjustment available.
Some fans are temperature sensing. The Intel retail boxed fan,
senses air temperature. A thermistor in the hub of the fan,
sniffs the passing air. As the air temperature rises, the fan
speed increases.
If two fan speed control mechanisms are used at the same time,
both of them reducing the voltage to the fan motor, the voltage
may drop to the point that the fan can no longer maintain
rotation. The rotor stops turning, or perhaps the rotation rate
is low enough, that the fan monitor in the BIOS reads zero.
Probing the voltage present on the red and black wires of the
fan, may indicate what is going on.
Now, out of all that description, exactly what are you
seeing again ? Is it that the fans will not return to
proper operation when you return from standby ? Any fan,
such as a case fan, that is connected to a non-Qfan equipped
header, should return to full speed. That is because the
+12V on that header, comes straight from the +12V wire on
the 20 pin ATX power connector. If you burn the copper track
on the motherboard, that feeds the fan header, then the
fan will stop.
According to this table, the P5P800 is a Q-fan board, so only
the CPU fan header has a MOSFET to change the fan voltage. If
the MOSFET fails, the copper track is burned, the BIOS control
algorithm has a bug, then the CPU fan speed could be reduced,
as the voltage at the fan motor is no longer +12V.
http://usa.asus.com/products/mb/feature.htm
If there is some problem with the software install, and you
have actually entered S1 standby state, all that happens there,
is the video card output is turned off, so the monitor goes
to sleep. In that case, PS_ON# is unaffected, and all voltages
are present. All fans should continue to run. S1 won't be
saving too much power, but S3 (suspend-to-ram) only uses +5VSB,
and that can save considerable power. S4 Hibernate saves the
most of all, because while +5VSB is still present, I don't think
the system expects any component to be holding data - recovery is
via rebooting from the hard drive when power returns. In S4
Hibernate, there is no need to refresh the DRAM, since the system
doesn't rely on RAM contents to return from S4, and instead
relies on the hard drive. That is why, when you are in S4
Hibernate, and you accidently flip the switch on the back of
the computer, the computer doesn't care - all the info needed to
recover is on the hard drive.
In summary:
S1 - "poor mans standby" - video card output off, monitor will sleep,
fans still spin, CPU is running
S3 - suspend-to-ram - DRAM holds system image, only +5VSB available,
fans off, CPU is off
S4 - hibernate - hard drive holds system state, +5VSB available but
unnecessary for the DRAM. Wake functions still work ? Fans off.
CPU off.
If you do an ACPI system install, and the BIOS was not set to S3 STR,
the OS will use S1 for standby. You can use the Microsoft "dumppo"
utility, to change from S1 to S3, once the BIOS setting is corrected.
Selecting "standby" in Windows, will then cause the fans to go off,
and the system will enter a lower power state than the S1.
If you did not do an ACPI-ready install, power functions will be
dumb. Like the computer next to me, you will see "it is safe to turn
off your computer", if you actually installed as "standard pc".
A "standard pc" won't have a standby option in the BIOS (so my
comp won't save any power until I shutdown). I didn't do this
on purpose, but inherited this characteristic because the hard
drive is a transplant from a previous system (and I hate
re-installing everything). When a system is installed without ACPI,
AFAIK, it cannot be successfully be added after the fact.
HTH,
Paul
