AdeW said:
If its dependant on the motherboard, I have a Dell computer which has
the USB sockets at the back Mar 2000 (ship date) Optiplex GX1 Pentium
III (I've looked inside and can't see a model # on the motherboard)
and...
...two others I get to use which have the USB sockets at the front...
Dell Optiplex 740 AMD Athlon 64, WinXP 2002 SP3 and
Dell Dimension 3100 Pentium 4 WinXP 2002 SP3.
Would their motherboards be likely to be modern enough?
Does having the USB socket at the back of the Mar2000 Pentium III one
give a clue as to how robust it is?
The point of my statement was, that instances of designs that
have problems with USB port reliability are relatively rare.
Intel has not admitted there is a problem with ICH5/ICH5R.
The only site admitting there is a problem with chips like
that, is the Gigabyte site. They are mainly concerned
with static discharge. I'm not convinced their analysis
is totally correct. There was one user who had a USB failure
that happened when his system was rebooted. In that case,
the initial damage may have been by static, but the only
trigger was the reboot. I think there is more to the issue
than just static electricity. But if Intel chooses to not
explain the problem, we'll never know. I'm sure Intel has
done post-mortem analysis and knows exactly what the mistake
was.
http://tw2005.giga-byte.com/Motherboard/Support/FAQ/FAQ_456.htm
Your Dimension 3100 comes just after that generation of
chip, so should not have a USB issue. The 915GV likely has
something like ICH6 on it. Using a utility like Everest,
you may be able to list the hardware inventory in the computer,
and figure it out from that.
http://support.dell.com/support/edocs/systems/dim3100/en/sm/specs0.htm#wp1052310
Other brands and model numbers of chips should not have a problem.
The only other instance I know of, is PCI USB2 cards with
NEC chips, tend to be static sensitive. I've seen reports
from a number of people, where one or more ports on their
PCI USB2 cards end up blown. While the other ports
continue to work. The ports seem to fail independently
on the NEC chip.
The above Gigabyte article makes a reference to ICH4
also having the problem, but I haven't seen evidence of
that in the newsgroups. Whenever the USB problem shows
up, it seems to be the ICH5/ICH5R at fault. If it is
happening to ICH4, nobody has complained about it.
Paul