USB port problem

Here's what I'm running:

Motherboard - Intel DP55KG
CPU - Intel Core i7 K 875 @ 2.93Ghz
Memory - 8 Gb (3.49 Gb in XP)
Nvidia GeForce GTS 250
Dual-boot Windows 7 Ultimate 64 bit, XP Pro SP3

I built this computer 6 months ago and have been generally satisfied
although it didn't quite set the world on fire as I expected.

In Windows XP (my preferred OS) I have been experiencing gradually
increasing problems with the two front USB ports, the ones on the case. I
examined the connection between the case and the motherboard and
everything seemed to be OK.

I made the mistake of trying to uninstall all the USB hubs, controllers,
etc. until my mouse and keyboard quit working. My thoughts were that when
the devices reinstalled themselves the problem would be corrected.
According the Device Manager, they did, but the mouse never did come
back.

I restored my XP partition from a disk image from about 10 days ago. So
I'm back to square one. When I plug a flash memory stick in one of the
front ports, the light comes on but no "Device Connect" sound. If I go
into Disk Management it shows but with no drive letter. I can assign it a
drive letter, but it still isn't accessible in Windows Explorer. The
drive letter I assigned it is there as removable media but clicking on it
yields "please insert media".

I have the installation CD that came with the motherboard. I ran the
chipset drivers installation but it changed nothing. I'm sure that if I
restored from a month old disk image, it would be OK again. But I don't
want to have to reinstall all the stuff I've installed in the last month.

BUT... in Windows 7 there is absolutely no problem with those front ports
whatsoever. It has to be something with the drivers in XP for those front
ports. Doesn't it? :-)

I'm getting by with an external 4 port hub that I can reach, but I'd
prefer to have everything working as it should.

Crossposted to alt.comp.hardware although I do think it's an OS problem.

I repeat... the problem does NOT exist in Windows 7 (which sucks).

As usual, any help would be greatly appreciated!

--
-- I'm out of white ink --

Menno Hershberger wrote in
news:Xns9F09E84B161D4butter@wefb973cbe498:

Here's what I'm running:

Motherboard - Intel DP55KG
CPU - Intel Core i7 K 875 @ 2.93Ghz
Memory - 8 Gb (3.49 Gb in XP)
Nvidia GeForce GTS 250
Dual-boot Windows 7 Ultimate 64 bit, XP Pro SP3

I built this computer 6 months ago and have been generally satisfied
although it didn't quite set the world on fire as I expected.

In Windows XP (my preferred OS) I have been experiencing gradually
increasing problems with the two front USB ports, the ones on the
case. I examined the connection between the case and the motherboard
and everything seemed to be OK.

I made the mistake of trying to uninstall all the USB hubs,
controllers, etc. until my mouse and keyboard quit working. My
thoughts were that when the devices reinstalled themselves the problem
would be corrected. According the Device Manager, they did, but the
mouse never did come back.

I restored my XP partition from a disk image from about 10 days ago.
So I'm back to square one. When I plug a flash memory stick in one of
the front ports, the light comes on but no "Device Connect" sound. If
I go into Disk Management it shows but with no drive letter. I can
assign it a drive letter, but it still isn't accessible in Windows
Explorer. The drive letter I assigned it is there as removable media
but clicking on it yields "please insert media".

I have the installation CD that came with the motherboard. I ran the
chipset drivers installation but it changed nothing.

Hold everything! When I ran that installation, it ran a bit and the
window went away. No closing remarks. The ports still didn't work so I
rebooted into Windows 7 and in my frustration made this post.

Now, back in XP again, the front ports seem to be working flawlessly. So
maybe the reinstall of the drivers DID fix it, but required a reboot to
take effect.

I wanted to post this right away so no one would waste a bunch of time
trying to figure out a solution for me. I've received plenty of
assistance in these groups and appreciate it all. Just don't want anyone
to waste any time on this one.

Crossing my fingers.... :-)


--
-- I'm out of white ink --

Menno Hershberger wrote in
news:Xns9F0A76661379butter@wefb973cbe498:

Menno Hershberger wrote in
news:Xns9F09E84B161D4butter@wefb973cbe498:

Here's what I'm running:

Motherboard - Intel DP55KG
CPU - Intel Core i7 K 875 @ 2.93Ghz
Memory - 8 Gb (3.49 Gb in XP)
Nvidia GeForce GTS 250
Dual-boot Windows 7 Ultimate 64 bit, XP Pro SP3

I built this computer 6 months ago and have been generally satisfied
although it didn't quite set the world on fire as I expected.

In Windows XP (my preferred OS) I have been experiencing gradually
increasing problems with the two front USB ports, the ones on the
case. I examined the connection between the case and the motherboard
and everything seemed to be OK.

I made the mistake of trying to uninstall all the USB hubs,
controllers, etc. until my mouse and keyboard quit working. My
thoughts were that when the devices reinstalled themselves the
problem would be corrected. According the Device Manager, they did,
but the mouse never did come back.

I restored my XP partition from a disk image from about 10 days ago.
So I'm back to square one. When I plug a flash memory stick in one of
the front ports, the light comes on but no "Device Connect" sound. If
I go into Disk Management it shows but with no drive letter. I can
assign it a drive letter, but it still isn't accessible in Windows
Explorer. The drive letter I assigned it is there as removable media
but clicking on it yields "please insert media".

I have the installation CD that came with the motherboard. I ran the
chipset drivers installation but it changed nothing.

Hold everything! When I ran that installation, it ran a bit and the
window went away. No closing remarks. The ports still didn't work so I
rebooted into Windows 7 and in my frustration made this post.

Now, back in XP again, the front ports seem to be working flawlessly.
So maybe the reinstall of the drivers DID fix it, but required a
reboot to take effect.

I wanted to post this right away so no one would waste a bunch of time
trying to figure out a solution for me. I've received plenty of
assistance in these groups and appreciate it all. Just don't want
anyone to waste any time on this one.

Crossing my fingers.... :-)

I spoke too soon... :-(
Rebooted for an update and it's back to the same thing again.

--
-- I'm out of white ink --

Menno Hershberger wrote:


I rebooted again today, and the ports started working again. So it's on
again off again behavior. I think I'll peek inside and see if there's an
extra USB header that I can plug those front two ports in to. After all
this playing around, I can't say positively that it *always* works in
Windows 7 either. I don't use Windows 7 that much and don't hardly ever use
those USB ports in Windows 7. So maybe it *is* a hardware problem.

Thanks for further educating me!


The USB section of Device Manager, is going to list USB logic blocks
in the Southbridge. And then, any USB devices connected to the ports,
are going to get enumerated, and if there are standard protocol stacks
for them, get installed too. That fills your setupapi.log with all sorts
of entries, and only some of those are critical to getting the USB port working.

You have to decode the VEN/DEV or VID/PID Plug and Play information,
to understand what is installing. You use things like this, to figure out
what you're looking at in setupapi.log . You can also look at the
INF files, and when Windows installs those, they're given names like
"OEM23.INF" instead of their original file names.

http://www.linux-usb.org/usb.ids
http://pciids.sourceforge.net/pci.ids

To start, I'd be looking through setupapi.log for reports of failures. In the
hope that if there is a repetitive condition, you can see the pattern in
the file and focus on that.

Copying the whole file isn't the right answer, as it could be megabytes.
Even selecting sections related to USB and copying those sections out,
would be a whole lot of work.

*******

As you noted in a previous post, USB2 drivers are not provided by
the hardware manufacturer. They're provided by Microsoft. If a
hardware manufacturer includes a USB driver, it is a stub and
calls something like USBPORT.INF. In other words, the installer
calls the standard Microsoft installer.

You can clean out the USB section of device manager, reboot, and
let Windows redetect all the hardware. This would be all fine and
good, if I could guarantee you that all related registry information
would get flushed. But I don't know that for a fact. Windows has
a bad habit of remembering things you would rather it forget.

http://www.usbman.com/Guides/Cleanup...afe%20Mode.htm

Note that the procedure there, doesn't address Vista and Windows 7.
It also warns that if you have a USB keyboard and mouse, and
flush everything like they tell you to, you could lose control of the
machine (i.e. cause keyboard, mouse, or both, to stop working).

There is also a script here, that uses "devcon" program, to flush
that section of Device Manager. The script programmatically looks
for things related to USB and tries to kill all of them.

http://www.robvanderwoude.com/devcon.php

"RenewUSB.bat"

You can read the script, and follow the logic it is using.

If you could map the failing item to a particular entry in Device
Manager, you could cause that alone to reinstall, and try to fix things
that way. But that might not clean out things like any USB
storage devices you were using.

Paul

Rob wrote:

This kind of protection is sometimes called a 'silicon fuse'.
When the bad device is unplugged AND the system is rebooted,
the port will work again (until the bad device is re-inserted.)
HTH,

This can be done with a Polyfuse, and the OC# signal.

+5 supply -------- Polyfuse -------+---------- To USB connector +5V pin
|
|
v
To OC# logic input
on Southbridge.
[OC = OverCurrent]

When the Polyfuse opens on an overload, the rail voltage on
the +5V pin on the right, drops to zero, and the hardware can be
alerted to the fact, by seeing a state change on OC#.

OC# is also used in older designs, by a "smart front panel".
The front panel can monitor current flow, and send OC# back
to the motherboards. Some of the cheaper motherboard
companies do it that way, to save component cost, and
make a nuisance of themselves. (One of the pins on their
2x5 connector, is OC#.) I don't think motherboards
are built like that any more, but some older ones were.
Builders would ground the OC# by accident, and not be
able to get the USB ports working :-)

I don't know what happens, when OC# is asserted, and what
a "normal" response would be. Maybe just a dialog box,
stating there is an overload ? I don't know if there
are any additional steps carried out.

Laptops are sometimes equipped with a silicon power bug
(like an 8 pin DIP chip), where the bug opens the power signal
with a MOSFET, and also has a logic output flag signal
which can be connected to OC# for monitoring.

The Southbridge might have half a dozen OC# signals on it,
to allow individual ports or pairs or the like, to report
their status. So there is more than one signal available
for monitoring.

It takes time for the Polyfuse above to cool off and
recrystallize. And then it's ready to conduct current
again.

Simply monitoring the voltage on the +5V pin on the
USB connector, can tell you what the motherboard is
thinking. And what level might be present on OC#.

Paul

Menno Hershberger wrote:


My back USB ports are full. Two printers, mouse, keyboard, a scanner, and
a 4 port hub. All those are functional. When the front ones don't work, I
use the 4 port hub. I have several memory sticks, 2 external USB hard
drives, a USB wireless adapter, and a cord that connects to my digital
camera. These are the things that go in front, but usually just one at a
time. When one works, they all work. And vice versa. When they *don't*
work, the LED's on the items that have them always light up, so there's
power to the ports all the time.

I did change to a different header but the behavior remains the same.
Changing the ports on the case would be major surgery. The case looked
good on paper but when I got it, it turned out to be a huge, heavy and
complicated. It looks pretty tame at http://tinyurl.com/69ynsff but looks
can be deceiving. The power and reset buttons some audio jacks, an
external eSata socker, and the two USB ports are on a slanted
"dashboard" on top. That part can't even be seen from the inside. The
damn thing weighs a ton and is so tall that I had to leave it stick out
from under my hutch about 7 inches in order to access the power switch
and USB ports.


Have you tried a low speed USB device on the front ? Does that work
consistently ? Perhaps you could move the optical mouse up front for
a test ? A low speed device, is to test whether your problem is only
"high speed related".

How many other computers have you built ? Would you happen to
have a 2x5 USB to slot adapter assembly from another motherboard ?
What happens if you take the side off the PC, connect a slot adapter
assembly instead of your top front panel, and use the slot adapter
USB ports ? The purpose of this test, is to compare the
electrical performance of the top front panel, versus a
known good USB assembly.

This is an example of such an adapter. It connects two dual headers to
make four rear connectors. You'd use a dual or a quad like this, as a means
to test the USB 2x5 headers on your motherboard. Assemblies like this
usually work. The assembly in the computer case, may be built on a
PCB, and sometimes the engineers gets a little too clever for
their own good, and they add filter components, series resistors
and the like. As a general rule, the companies that make computer
cases are "metal bangers" and they tend to sub-contract the making
of the front panel PCB thing. And in some cases, they don't even
bother to verify what they sub-contracted (because all their staff
are metal shop workers).

http://us.estore.asus.com/index.php?..._images&p=2091

Paul