Got a bad USB port. How to fix?

[*Vanguard*]

This is for an old box using the AOpen AX6BC motherboard with a Pentium3
800MHz. There's one card slot left but I'll need that later for a SCSI
card. I have 2 printers so one is connected to the parallel port and the
other to one of the two USB ports on the backpanel. That left me with one
USB port. These are USB 1 ports but the peripherals I'm connecting probably
won't exceed their speed (the printer certainly won't but maybe the scanner
might approach or might exceed the USB 1 port speed but it seems to scan
just fine). When I tried using the 2nd USB port (which is part of the same
connector assembly on the motherboard at the backpanel), I found it was
dead. Anything connected to it doesn't work. If I swap out the printer and
scanner on the lower USB port then that device works but anything plugged
into the upper USB port does not work. Looks like the upper USB port is
physically dead. I found nothing in the BIOS to disable just one USB port
(or both of them); i.e., the only BIOS option regards USB support for the
keyboard (disabled since I'm using the PS/2 port).

I'm wondering about the hookup to the USB port connector assembly. Wouldn't
both of these ports go to the same USB controller? If so, then I would
think it is the connector assembly with the two USB ports that is damaged.
It's not like a RJ-11 or RJ-45 jack where the contact wires might be crushed
and you can pull them back out to make contact with the plug. The USB
connector is like a mini circuit board the connector slides onto. So there
doesn't seem to be anything within the connector that I could repair. If I
go the route of replacing the USB connector assembly (I'm good at soldering
but taking apart the box to extract the mobo will be a bitch), I'd like to
know that replacing it will actually fix it. No point in replacing it if
the problem is a foil on the mobo's PCB or a dead channel on the USB
controller.

For now, I've got a 4-port powered USB hub attached to the one working USB
port on the case backpanel. But I'd like to get rid of the hub and get the
2nd USB port working.

 

[tonj]

1) does your motherboard have pins waiting for usb expansion ports? many
boards do
2) if you can somehow haggle yourself another pci slot you could put a USB
card in.

 

[*Vanguard*]

"tonj" said in news:[email protected]:

1) does your motherboard have pins waiting for usb expansion ports?
many boards do

This is the 2-port USB connector at the backpanel. As such, it is an
assembly that looks to be soldered onto the motherboard. That's why I was
trying to figure out if the foils going to the connector were all shared for
both the port jacks or if maybe there are different channels and foils for
them going to each port jack (so one set or a unique foil for just the one
dead USB port maybe got damaged).

I didn't know if there was something obvious to look for to repair the
backpanel connector. If not, I'm not sure I want to remove the whole
connector assembly to just get one port working since the other port works
and I'm using a USB hub to overcome the restriction of just having the one
working USB port on the backpanel.

2) if you can somehow haggle yourself another pci slot you could put
a USB card in.

Unfortunately all card slots are filled except one and that's waiting for
the SCSI card to show up.

 

[kony]

This is for an old box using the AOpen AX6BC motherboard with a Pentium3
800MHz. There's one card slot left but I'll need that later for a SCSI
card. I have 2 printers so one is connected to the parallel port and the
other to one of the two USB ports on the backpanel. That left me with one
USB port. These are USB 1 ports but the peripherals I'm connecting probably
won't exceed their speed (the printer certainly won't but maybe the scanner
might approach or might exceed the USB 1 port speed but it seems to scan
just fine). When I tried using the 2nd USB port (which is part of the same
connector assembly on the motherboard at the backpanel), I found it was
dead. Anything connected to it doesn't work. If I swap out the printer and
scanner on the lower USB port then that device works but anything plugged
into the upper USB port does not work. Looks like the upper USB port is
physically dead. I found nothing in the BIOS to disable just one USB port
(or both of them); i.e., the only BIOS option regards USB support for the
keyboard (disabled since I'm using the PS/2 port).

I'm wondering about the hookup to the USB port connector assembly. Wouldn't
both of these ports go to the same USB controller? If so, then I would
think it is the connector assembly with the two USB ports that is damaged.
It's not like a RJ-11 or RJ-45 jack where the contact wires might be crushed
and you can pull them back out to make contact with the plug. The USB
connector is like a mini circuit board the connector slides onto. So there
doesn't seem to be anything within the connector that I could repair. If I
go the route of replacing the USB connector assembly (I'm good at soldering
but taking apart the box to extract the mobo will be a bitch), I'd like to
know that replacing it will actually fix it. No point in replacing it if
the problem is a foil on the mobo's PCB or a dead channel on the USB
controller.

For now, I've got a 4-port powered USB hub attached to the one working USB
port on the case backpanel. But I'd like to get rid of the hub and get the
2nd USB port working.

Both ports are to the same controller.

You might first check continuity on the fuses behind it. From a picture I
saw they appear to be just below and to the right of the top left board
mounting hole.

The port itself isn't all that complicated, a visual inspection (to verify
the contacts are clean) of the contacts as they appear from the back of
the case, should be enough. Not only should the contacts be clean, but
they should have a slight bend inwards towards the end to provide a
spring-like friction contact with the inserted connector. If there's a
problem with the physical port connector it should be visually obvious,
and if the contact isn't springy enough, it may be possible to use a
couple probes to hold down the contact strip while prying up from
underneath with a needle, to recreate the slight bend that the other
contacts have.

The traces could be lifted, damaged in some other way, but if the board
worked while installed in it's current case I wouldn't expect this to be
the problem. Then again, "something" is the problem, however unlikely.
It's also unlikely that the USB controller channel is dead unless some
sort of incorrect connection were made to it.

The larger question is how much time this old board is worth. The fuse
can be checked while it's in the case still, as well as the visual
inspection of the contacts, but beyond that it'll be pretty time consuming
to pull the board and check continuity on the traces, fix (whatever),
then reinstall in case. If it did turn out to be the fuse, I wouldn't go
to the bother of finding and installing another fuse for such an old
board, just carefully put some conductive paint or solder a wire overtop
the fuse (or desolder it before doing same). Ideally it should have a
working fuse, but then again some junk boards like PCChips have done
without these fuses in their design.

 

[*Vanguard*]

"kony" said in news:[email protected]:

Both ports are to the same controller.

You might first check continuity on the fuses behind it. From a
picture I saw they appear to be just below and to the right of the
top left board mounting hole.

The port itself isn't all that complicated, a visual inspection (to
verify the contacts are clean) of the contacts as they appear from
the back of the case, should be enough. Not only should the contacts
be clean, but they should have a slight bend inwards towards the end
to provide a spring-like friction contact with the inserted
connector. If there's a problem with the physical port connector it
should be visually obvious, and if the contact isn't springy enough,
it may be possible to use a couple probes to hold down the contact
strip while prying up from underneath with a needle, to recreate the
slight bend that the other contacts have.

The traces could be lifted, damaged in some other way, but if the
board worked while installed in it's current case I wouldn't expect
this to be the problem. Then again, "something" is the problem,
however unlikely. It's also unlikely that the USB controller channel
is dead unless some sort of incorrect connection were made to it.

The larger question is how much time this old board is worth. The
fuse can be checked while it's in the case still, as well as the
visual inspection of the contacts, but beyond that it'll be pretty
time consuming to pull the board and check continuity on the traces,
fix (whatever), then reinstall in case. If it did turn out to be the
fuse, I wouldn't go to the bother of finding and installing another
fuse for such an old board, just carefully put some conductive paint
or solder a wire overtop the fuse (or desolder it before doing same).
Ideally it should have a working fuse, but then again some junk
boards like PCChips have done without these fuses in their design.

Fuses. I never thought of that but it makes sense since these ports have to
provide power to the USB device and why I ended up getting a powered USB hub
at one time because the mobo USB ports might not be capable of supplying
enough juice for high-powered USB devices, like USB headphones. Maybe
that's what caused it, using the USB headphones sucked too much current and
the fuse blew.

If the fuse is a tiny oblong rectangular component soldered onto the
motherboard then maybe I can replace it. I should be able to tell which one
is open with a multimeter. But as you say, trying to find a correct
replacement might be very hard and consume more time than it's worth. The
manual has an arrow to a couple of "resettable fuse" chips on the mobo but
never explains what they are used for. While the manual didn't explain
them, a Google search found a link back to AOpen at
http://english.aopen.com.tw/tech/techinside/r_fuse.htm. Again, I can
probably ohm out each side of the fuse to see which one(s) go to the USB
ports. If they are used for the USB ports then it looks like one of them
didn't reset and instead went dead (i.e., is always open).

There must be a separate PTC fuse for each USB port; otherwise, the fuse
going bad and remaining at high resistance would kill both ports, but one of
the ports does work. If they are thick enough, I can just use an alligator
clip to test if shorting a fuse brings back the dead USB port as functional.
But where to find a replacement fuse and how to tell what is the rating for
the current one? I haven't checked yet but the link above to the help
article at AOpen shows markings on the mobo for the current and voltage of
the example shown, so hopefully there'll be markings to tell me what I could
use. Even if I can't get a surface mount fuse chip replacement, I suppose I
could solder on a same-rate wire fuse using piggy tails. I just started
hunting around digikey.com which always has lots of components. It looks
like I need a PTC resettable fuse and maybe http://snipurl.com/4m74 fits the
bill. It's the same amperage and voltage as shown in the AOpen article but
I'll have to check if it's the same size.

I really don't like the idea of just shorting across it. It blew once and
so badly that the resettable fuse never reset to a usable state or burn up,
so it could happen again. Thanks for the hint regarding the fuses. Makes
sense now that I think of it. Sometimes they cram too much around those
mounting holes. Could be a slip of the screwdriver damaged a fuse. I've
seen it happen that a slip of the screwdriver disabled the keyboard port
(which also has a fuse).

 

[kony]

Fuses. I never thought of that but it makes sense since these ports have to
provide power to the USB device and why I ended up getting a powered USB hub
at one time because the mobo USB ports might not be capable of supplying
enough juice for high-powered USB devices, like USB headphones. Maybe
that's what caused it, using the USB headphones sucked too much current and
the fuse blew.

If so I'd be very cautious using those headphones with any board, though
it's odd, they ought to draw a max 500mA and that "shouldn't" be enough to
trip the fuse.

If the fuse is a tiny oblong rectangular component soldered onto the
motherboard then maybe I can replace it. I should be able to tell which one
is open with a multimeter.

Or the other option is to check for 5V on the port itself.

But as you say, trying to find a correct
replacement might be very hard and consume more time than it's worth. The
manual has an arrow to a couple of "resettable fuse" chips on the mobo but
never explains what they are used for. While the manual didn't explain
them, a Google search found a link back to AOpen at
http://english.aopen.com.tw/tech/techinside/r_fuse.htm. Again, I can
probably ohm out each side of the fuse to see which one(s) go to the USB
ports. If they are used for the USB ports then it looks like one of them
didn't reset and instead went dead (i.e., is always open).

I no longer have an AX6BC but still have a high-res pic of a revision 1.1
board. The silkscreening appears to spec the fuses as 125V, 3A, which
seems WAY too high to me but who knows?

http://69.36.189.159/usr_1034/ax6bc_tl.jpg

There must be a separate PTC fuse for each USB port; otherwise, the fuse
going bad and remaining at high resistance would kill both ports, but one of
the ports does work. If they are thick enough, I can just use an alligator
clip to test if shorting a fuse brings back the dead USB port as functional.
But where to find a replacement fuse and how to tell what is the rating for
the current one? I haven't checked yet but the link above to the help
article at AOpen shows markings on the mobo for the current and voltage of
the example shown, so hopefully there'll be markings to tell me what I could
use. Even if I can't get a surface mount fuse chip replacement, I suppose I
could solder on a same-rate wire fuse using piggy tails. I just started
hunting around digikey.com which always has lots of components. It looks
like I need a PTC resettable fuse and maybe http://snipurl.com/4m74 fits the
bill. It's the same amperage and voltage as shown in the AOpen article but
I'll have to check if it's the same size.

Considering the location you could probably make any fuse work, but at
this point I'd just do without the fuse, if that is the problem.
Considering the fuse, shipping, time, etc, it might be just as well to
replace the board with something using Apollo Pro 133A/694X chipset,
giving you a bit newer features, higher ATA speed and modern PC133 high
density memory support.

I really don't like the idea of just shorting across it. It blew once and
so badly that the resettable fuse never reset to a usable state or burn up,
so it could happen again. Thanks for the hint regarding the fuses. Makes
sense now that I think of it. Sometimes they cram too much around those
mounting holes. Could be a slip of the screwdriver damaged a fuse. I've
seen it happen that a slip of the screwdriver disabled the keyboard port
(which also has a fuse).

That's the part that's nagging at me, that the keyboard port usually has a
fuse. I only saw two fuses leaving only one for both USB ports... so now
I wonder if it isn't the fuse that's the problem.