Robin said:
That's all true, and no doubt the m/b might do something with the
signals, but on my ASUS P5E3 neither the BIOS nor ASUS's Probe
software monitors the pwr_fan socket. I connected the Antec lead to a
cha-fan socket.
What you can do, is download the datasheet for the hardware
monitor chip, and see how many measurement channels it has
for fan RPMs. Sometimes, it is possible to predict in
advance, that one of the headers will be unable to measure
anything (more fan headers than channels).
(Page 28 - five "fanin" signals)
http://www.winbond.com/NR/rdonlyres/7885623D-A487-4CF9-A47F-30C5F73D6FE6/0/W83627DHG.pdf
That chip has five channels for RPM measurement, of which
two have dual definition pins. Three channels are definitely
available for measurements. A fourth can be used without
too much trouble (the alternate definition is a GPIO pin).
The fifth channel pin, is also used for a serial flash
interface, so if there is a serial flash chip on the
board, that channel may be "hijacked" for interfacing
to flash.
In any case, the user manual for P5E3 shows six fan
headers, and there are a maximum of five measurement
channels on the Winbond chip, so the Asus manual
is wrong. One of the headers should have the "Rotation"
label removed, as the pin obviously cannot be connected
to the hardware monitor. No channels left.
Paul