He's referring to the ICH5 Southbridge of his motherboard, and the
logic block inside there associated with USB2. That is the
Intel(R) 82801EB USB2 Enhanced Host Controller 24DD
that controls USB2 operation on up to eight USB connectors on
the motherboard.
And I do not have a clue why it would be doing that, unless
there are USB bus resets, followed by an incorrect negotiation.
It still implies a hardware malfunction, as the software and
hardware would be attempting to run at USB2 rates, but for whatever
reason, the attempt fails and the hardware connected to the port ends
up at USB 1.1 rates. So you would have a bus reset (which is a normal
part of the protocol, but needs an event to kick it off), followed
by a failed negotiation.
Even if you had a copy of UVCVIEW or equivalent, I doubt it would
tell you anything.
There was at least one motherboard, that had a "drive strength" setting
in the BIOS for the USB ports. Changing that, affects the current flow in
USB2 mode. But such a setting has not been made available in years, implying
tuning it is not necessary. On the motherboard I spotted that setting on,
the hardware never needed any adjustment of that value, and worked as well
as any other motherboard with USB2 I have got.
I'd be more concerned, if a USB2 PCI card was added to the PC, and the
symptoms continued. Using a USB2 PCI card, allows testing with another
USB chip. If it cures the problem, then you would know there was something
about the motherboard and its ICH5, which was at fault.
Since the ICH5 has the "latchup failure" problem associated with the
USB ports, I'd insert and use a USB2 PCI card anyway, just to protect
the motherboard. It would be hard to find a nice replacement motherboard,
if the ICH5 burns up.
(ICH5 after failure related to USB ports... That's a burn mark.)
http://onfinite.com/libraries/179057/2ea.jpg
Paul