Hardware mystery . . . USB 2.0 on a USB 1.1 chip???

[Mike T.]

Long story short, I'm rebuilding a system for a friend who is on an
extremely tight budget. He had a power supply failure recently that killed
his mainboard. He could afford a new (high quality) power supply and a
cheap new mainboard and nothing else. I saved as much data as I could
(using another system), then wiped the hard drive and reinstalled his copy
of Windows ME (yeah, I know, but it's not my system) from scratch, due to
the chipset change. System is up and running and rock-solid stable, with
only one problem . . .

If the onboard sound is enabled, system will not boot. Or sometimes it
boots, but then is extremely unstable. No conflicts are SHOWING in device
manager, but the sound card insists on sharing a IRQ (whatever IRQ) with a
USB Host Controller, "VIA 3038". I can change the IRQ in the BIOS for the
USB controllers. But if I do, the onboard sound IRQ changes to match, so
this accomplishes nothing. (IRQ conflict just moves to a DIFFERENT IRQ)
According to device manager, FIVE of these VIA 3038 are on the mainboard,
but one is not working correctly. (the other four are) It is the one that
is not working correctly that is conflicting with the onboard sound, causing
the system to not boot (or be extremely unstable). If I disable the
non-working VIA 3038 UHC and re-enable the onboard sound, this helps
somewhat. (system runs more stable that way, but not enough that I'd be
willing to depend on it). This system has the latest BIOS, runs the latest
4-in-1 drivers, and has the USB patch installed from viaarena. All updates
from Microsoft have been downloaded and installed. System scores well over
5000 on 3DMark03 (not bad, for a non-gaming system), runs Memtest and CPU
Stability (either one) for many hours with no errors. But this is only with
the onboard sound disabled in the BIOS! I threw in an older Turtle Beach
PCI sound card for testing, so I know that the system can run rock-solid
stable (and with SOUND, also), but not if the onboard sound is enabled.

Now obviously, I could just give him the turtle beach card (or another sound
card), leave the onboard sound disabled and be DONE with the matter. IF it
aint broke, don't fix it. BUT, in trying to research this issue, I
discovered something really odd:

According to the web site, this mainboard is supposed to support 8X USB2.0
ports. See:
http://www.biostar-usa.com/mbdetails.asp?model=p4m80-m4

But the working VIA 3038 USB Host Controllers (4 of them) and the
NON-WORKING one that is causing the conflict with onboard sound, are
apparently USB 1.1. See:
http://www.viaarena.com/default.aspx?PageID=420&OSID=5&CatID=1330&SubCatID=121

I've already e-mailed Biostar (maker of the mainboard) to ask about this,
and got no response.

Is it possible that Windows (ME) has mis-identified the USB Host Controller?
IF so, shouldn't the latest 4-in-1 drivers (already installed) solve that
issue? There is no separate USB driver, so updating that is not an option.

How is it that I've apparently got TEN USB 1.1 ports (including two that
would be running off a non-functional Host Controller) on a board that is
supposed to have EIGHT USB 2.0 ports????

Can anybody shed some light on this? -Dave

 

[JAD]

'reset configuration data' in the BIOS or reset the CMOS.
next boot to safe mode with the audio enabled, all is good? then there is a
good chance this is a driver issue rather than a hardware problem. remove
the sound card entry from device manager. reboot to normal.

 

[Mike T.]

'reset configuration data' in the BIOS or reset the CMOS.
next boot to safe mode with the audio enabled, all is good? then there is
a
good chance this is a driver issue rather than a hardware problem. remove
the sound card entry from device manager. reboot to normal.

I tried resetting CMOS a couple of times. This did not change things.
Although I only tried (safe mode with onboard sound enabled) once, I could
not boot to safe mode with the onboard sound enabled. I'd be inclined to
think this is a bad motherboard. Only, with onboard sound disabled,
everything else works perfectly. As I wrote before, rock-solid stable, even
with burn-in programs running for several hours. Also, if I disable the
non-functional (according to device manager) USB Host Controller and
re-enable the onboard sound, the system works OK (not good, but OK).
Meaning, it will boot fine, and usually run OK, but locks up frequently.
The latest sound card drivers are installed. The latest 4-in-1 drivers are
installed. There is no separate USB driver, so updating that is not an
option, unfortunately. I could be wrong, but I don't think it's a BIOS or
Driver issue. My best guess at this point is someone put the wrong chips on
the board at the factory?????? But that's a wild guess, based on the fact
that it's supposed to be USB 2.0, but apparently has USB 1.1 hardware, and
some of that USB 1.1 hardware is non-functional. -Dave

 

[JAD]

I tried resetting CMOS a couple of times. This did not change things.
Although I only tried (safe mode with onboard sound enabled) once, I could
not boot to safe mode with the onboard sound enabled. I'd be inclined to
think this is a bad motherboard.

When you say boot, I assume your speaking booting into windows. This would
point in the direction of driver issue, as it does not prevent you from
entering bios. What MB is this?
Have you the bios switches for ' PnP OS installed ' (yes /no) and 'Reset
configuration data' (yes / no)?
try 'PnP os' to no and 'reset config' to yes.

Only, with onboard sound disabled,

everything else works perfectly. As I wrote before, rock-solid stable, even
with burn-in programs running for several hours. Also, if I disable the
non-functional (according to device manager) USB Host Controller and
re-enable the onboard sound, the system works OK (not good, but OK).
Meaning, it will boot fine, and usually run OK, but locks up frequently.

Do you have an 'other devices' heading in the device manager?
Does the non functional USB heading give an error or reason?

The latest sound card drivers are installed. The latest 4-in-1 drivers are
installed.

There is no separate USB driver, so updating that is not an

option, unfortunately.

Most USB (like sata drivers) are a separate (download)File / install. These
are usually found on a MB install disk.

I could be wrong, but I don't think it's a BIOS or

Driver issue. My best guess at this point is someone put the wrong chips on
the board at the factory?????? But that's a wild guess, based on the fact
that it's supposed to be USB 2.0, but apparently has USB 1.1 hardware, and
some of that USB 1.1 hardware is non-functional. -Dave

More of an example of a driver issue, especially if the OS tried to install
native drivers for a proprietary USB host controller.

 

[Mike T.]

When you say boot, I assume your speaking booting into windows. This

would
point in the direction of driver issue, as it does not prevent you from
entering bios. What MB is this?
Have you the bios switches for ' PnP OS installed ' (yes /no) and 'Reset
configuration data' (yes / no)?
try 'PnP os' to no and 'reset config' to yes.

Do you have an 'other devices' heading in the device manager?
Does the non functional USB heading give an error or reason?

There is no separate USB driver, so updating that is not an

Most USB (like sata drivers) are a separate (download)File / install.
These
are usually found on a MB install disk.

More of an example of a driver issue, especially if the OS tried to
install
native drivers for a proprietary USB host controller.

Biostar P4M80-M4
http://www.biostar-usa.com/mbdetails.asp?model=p4m80-m4
I've tried PNP OS setting both yes and no, and reset configuration both yes
and no. Nothing seems to change anything, as far as what devices show up in
device manager. I don't have "other devices" showing in device manager.
All hardware appears to be properly recognized, except for one VIA 3038 USB
Host Controller (#5 of 5). The USB heading says device not present or not
working properly. What I don't understand is, if it's not there, why is it
being assigned an IRQ? There is no USB driver available on the motherboard
install disk. There is no USB driver available from the Biostar web site.
The viaarena web site specifically states that there is no driver for the
VIA 3038, as Windows has native support for it. So it's not a proprietary
host controller. As 4 out of 5 of them are working correctly, it's unlikely
that it's a driver issue. -Dave

 

[JAD]

have you tried 'removing' all the offending hardware from device manager in
safe mode? then rebooting into normal?
Have you 'disabled' the dead USB from device manager in normal mode?

The dead USB listing in USB is because it most likely the 2.0 port and since
the host controller didn't load you get lamed hardware.

http://www.biostar-usa.com/downloaddetails.asp?fileid=106

however if you are using XP you need XPsp2 so it says.......this would be a
'successful' sp2 install

 

[Mike T.]

have you tried 'removing' all the offending hardware from device manager
in
safe mode? then rebooting into normal?

Yes. Didn't work.

Have you 'disabled' the dead USB from device manager in normal mode?

Yes, with limited success. At least the board will boot that way, with
onboard sound enabled.

The dead USB listing in USB is because it most likely the 2.0 port and
since
the host controller didn't load you get lamed hardware.

http://www.biostar-usa.com/downloaddetails.asp?fileid=106

This makes the most sense, if I assume that Windows mis-identified the USB
Host Controller and loaded a driver for an earlier version of it. If this
driver is needed though, it is really odd that it is not on the driver page
for this specific motherboard. But I'll try it, and I'd be somewhat
surprised if that doesn't work. Thank You! -Dave

 

[David Maynard]

Long story short, I'm rebuilding a system for a friend who is on an
extremely tight budget. He had a power supply failure recently that killed
his mainboard. He could afford a new (high quality) power supply and a
cheap new mainboard and nothing else. I saved as much data as I could
(using another system), then wiped the hard drive and reinstalled his copy
of Windows ME (yeah, I know, but it's not my system) from scratch, due to
the chipset change. System is up and running and rock-solid stable, with
only one problem . . .

If the onboard sound is enabled, system will not boot. Or sometimes it
boots, but then is extremely unstable. No conflicts are SHOWING in device
manager, but the sound card insists on sharing a IRQ (whatever IRQ) with a
USB Host Controller, "VIA 3038". I can change the IRQ in the BIOS for the
USB controllers. But if I do, the onboard sound IRQ changes to match, so
this accomplishes nothing. (IRQ conflict just moves to a DIFFERENT IRQ)

Your description suggests they're using the same physical IRQ on the
motherboard.

According to device manager, FIVE of these VIA 3038 are on the mainboard,
but one is not working correctly. (the other four are) It is the one that
is not working correctly that is conflicting with the onboard sound, causing
the system to not boot (or be extremely unstable). If I disable the
non-working VIA 3038 UHC and re-enable the onboard sound, this helps
somewhat. (system runs more stable that way, but not enough that I'd be
willing to depend on it). This system has the latest BIOS, runs the latest
4-in-1 drivers, and has the USB patch installed from viaarena. All updates
from Microsoft have been downloaded and installed. System scores well over
5000 on 3DMark03 (not bad, for a non-gaming system), runs Memtest and CPU
Stability (either one) for many hours with no errors. But this is only with
the onboard sound disabled in the BIOS! I threw in an older Turtle Beach
PCI sound card for testing, so I know that the system can run rock-solid
stable (and with SOUND, also), but not if the onboard sound is enabled.

Now obviously, I could just give him the turtle beach card (or another sound
card), leave the onboard sound disabled and be DONE with the matter. IF it
aint broke, don't fix it. BUT, in trying to research this issue, I
discovered something really odd:

According to the web site, this mainboard is supposed to support 8X USB2.0
ports. See:
http://www.biostar-usa.com/mbdetails.asp?model=p4m80-m4

But the working VIA 3038 USB Host Controllers (4 of them) and the
NON-WORKING one that is causing the conflict with onboard sound, are
apparently USB 1.1. See:
http://www.viaarena.com/default.aspx?PageID=420&OSID=5&CatID=1330&SubCatID=121

I've already e-mailed Biostar (maker of the mainboard) to ask about this,
and got no response.

Is it possible that Windows (ME) has mis-identified the USB Host Controller?
IF so, shouldn't the latest 4-in-1 drivers (already installed) solve that
issue? There is no separate USB driver, so updating that is not an option.

How is it that I've apparently got TEN USB 1.1 ports (including two that
would be running off a non-functional Host Controller) on a board that is
supposed to have EIGHT USB 2.0 ports????

Just off hand I'd say you don't have 10 when there are 8 and that would
seem to be confirmed by the 'non-functional' two, because they aren't there.

Yes. Windows can misidentify hardware and, worse, it often 'saves' that
misinformation even when the driver is removed (the registry entry remains).

Just off hand I'd guess it mistakenly identified a piece of the on-board
sound card as a USB Host controller.

Not sure this will work but I'd try deleting each of the working USB ports
one at a time and then telling it to not install anything when they're
redetected, to get them out of the way (the point is to avoid being
confused as to which USB host controller it's trying to install on the
reboots: the 'good ones' or the 'bad one'). Then remove both the sound card
drivers and the 'non-functional' USB port and let it redetect on the next
reboot. If you're lucky, since all of that should be 'sound card' bits
maybe it'll figure it out. If not I'd then try manually reinstalling the
'non-functional' USB Host controller pointing to the sound card drivers and
hope it figures out which section it is. Then reinstall the working USB
host controllers.

 

[fj]

Your description suggests they're using the same physical IRQ on the
motherboard.


Just off hand I'd say you don't have 10 when there are 8 and that would
seem to be confirmed by the 'non-functional' two, because they aren't
there.

Yes. Windows can misidentify hardware and, worse, it often 'saves' that
misinformation even when the driver is removed (the registry entry
remains).

Is this true for Win 2K and Win XP as well? I've had lots of experience
with this type of problem in Win 95 and Win 98, but don't seem to have run
into it in Win XP or Win 2K.
Thanks

 

[David Maynard]

Is this true for Win 2K and Win XP as well? I've had lots of experience
with this type of problem in Win 95 and Win 98, but don't seem to have run
into it in Win XP or Win 2K.

Well, in theory it's not supposed to happen at all since the hardware's
"unique ID" is supposed to link up with the appropriate .inf file but there
are a ton of ways that can get messed up, third party drivers replacing
..inf files, for one.

IMO most of them are caused by 'not quite compliant' hardware
implementations that interfere with each other rather than 'Windows'.