Bob Epinette wrote:
> Yup, got BIOS update to v1019 for my P4P800 mobo from the Asus web site. Yup,
> machine is *home* built as in from a local, reputable builder, with top-notch
> parts. Chipset drivers obtained direct from Intel web site as well.
>
> I will try tonight the boot into Linux idea, and add the USB sniffer or view
> the Event log.
>
> In the meantime, there is some addtional info from Microsoft about quikly
> removing and re-inserting flash storage that can cause ports to stop
> functionning completely:
>
> "
> SYMPTOMS
> If you quickly and repeatedly insert and remove a USB device, the USB port
> may stop responding. When the port is in this state, it no longer recognizes
> any USB device.
>
> CAUSE
> This problem may occur because of a rare condition in the Selective Suspend
> feature of USB. Microsoft and the hardware vendors that manufacture the chips
> that are used for USB have investigated the problem. However, the results
> have been inconclusive because of the intermittent nature of the problem.
> "
>
> Full article at: http://support.microsoft.com/kb/817900/en-us
>
> Talk to you all later....
>
> Bob
>
I'd be careful with that motherboard. There are issues with the ICH5 Southbridge
and in particular, the USB ports. There is a problem called "latchup", where the
ICH5 fails completely. There is a picture here, of what a failed Southbridge
looks like. The appearance is of a problem called "latchup" and the location
of the burn marks here, correlates with the position of the power pads that
feed the USB interface.
http://www.abxzone.com/forums/showth...light=usb+port
It could be, that this is just another instance of latchup. There are different
levels of latchup, from the spectacular "chip burning" flavor, to "micro latchup",
where individual CMOS gates stop functioning. In "micro latchup", if you
pull the AC plug, and remove all power from the computer, the chip can recover and
function again. But if the latchup occurs in a part of the circuit connected to
a strong source of power, it can burn instead. Even the CMOS battery could provide
enough power to keep a latched up circuit, in the latched up state. It all
depends on whether the thing latched up, is near enough to the CMOS well to
get power to it.
On my own P4C800-E Deluxe, I've stopped using the motherboard USB ports.
And if I wanted some USB ports on my ICH5/ICH5R based motherboard, I'd
use a separate PCI USB card for that. Something with a NEC chip on it
would be nice. If you ruin a PCI USB card, it is a simple matter to unplug
it, and replace it with another $7 USB card. Once the warranty on the
motherboard has expired, a repair is more expensive.
This is all that remains of the only motherboard company to issue a warning.
View this in IE, as Firefox cannot handle the code. This is an archived
copy, as the original has been removed.
http://web.archive.org/web/200512221...AQ/FAQ_456.htm
There was a rash of failures at one time, but there have been few reports
lately.
Paul