usb port on fire

  • Thread starter Thread starter Bob
  • Start date Start date
B

Bob

A couple of nights ago I was reinstalling a driver with cd rom for a hp
printer. Smoke started coming out the bottom of the computer, the front usb
ports were actually so hot they melted. I disconnected everything, had to
pull the wires from case. Bottom of case had melted along with cover door
for usb. Printer was connected to back usb, fronts were not even being used.
I built the computer about a year ago, have had no problems. Front and back
usb's always worked. Computer actually still works, just no front usb now.
What could have caused this to overheat like this. Hate to think what would
have happened had I not been sitting at desk. Thanks for any info.
 
A couple of nights ago I was reinstalling a driver with cd rom for a hp
printer. Smoke started coming out the bottom of the computer, the front usb
ports were actually so hot they melted. I disconnected everything, had to
pull the wires from case. Bottom of case had melted along with cover door
for usb. Printer was connected to back usb, fronts were not even being used.
I built the computer about a year ago, have had no problems. Front and back
usb's always worked. Computer actually still works, just no front usb now.
What could have caused this to overheat like this. Hate to think what would
have happened had I not been sitting at desk. Thanks for any info.

Had to have been a short circuit. No fried mice inside the case?

Take the motherboard out, LOOK for any loose screws, metal bits, that
sort of thing shorting inside. Do a *thorough* inspection before using
it again.
 
A couple of nights ago I was reinstalling a driver with cd rom for a hp
printer. Smoke started coming out the bottom of the computer, the front usb
ports were actually so hot they melted. I disconnected everything, had to
pull the wires from case. Bottom of case had melted along with cover door
for usb. Printer was connected to back usb, fronts were not even being used.
I built the computer about a year ago, have had no problems. Front and back
usb's always worked. Computer actually still works, just no front usb now.
What could have caused this to overheat like this. Hate to think what would
have happened had I not been sitting at desk. Thanks for any info.

Is it possible your front panel USB was on a small circuit
board with tantalum capacitor(s) on it, and that one went in
flames? Did it use a ribbon cable that was subject to
fraying against a sharp corner of the metalwork on the case,
or if damaged enough even a round cable with visually
degraded insulation?

A forensic analysis of the front port assembly should find
the cause.
 
Bob said:
A couple of nights ago I was reinstalling a driver with cd rom for a hp
printer. Smoke started coming out the bottom of the computer, the front usb
ports were actually so hot they melted. I disconnected everything, had to
pull the wires from case. Bottom of case had melted along with cover door
for usb. Printer was connected to back usb, fronts were not even being used.
I built the computer about a year ago, have had no problems. Front and back
usb's always worked. Computer actually still works, just no front usb now.
What could have caused this to overheat like this. Hate to think what would
have happened had I not been sitting at desk. Thanks for any info.

Overcurrent protection exists in two forms. Some motherboards use small green
Polyfuses, which recover when they cool off. They are a fuse which doesn't
need to be replaced, and if present, they open at about 1.1 amps or so. I think
as a rule, a two connector stack runs off the same fuse.

A second method is USBOC#. That is a signal passed from any adapter assembly,
to the motherboard. The idea is, the adapter assembly (like on the front of
the computer case), measures the current. If the current is too high for the
USB standard, the USBOC# signal is sent to the motherboard. A logic low
level means "switch off this port". Then, it is up to the motherboard to
remove the power. To cause a problem, there would have to be a double
fault, where the USB assembly or a plugged in device, shorted +5V, and there
was also a failure of the switching device. And frying the switching device,
should stop the flow of current, eventually.

The user can complicate life, if mixing a non-USBOC# front panel assembly,
with a USBOC# motherboard. Failure to use the right kind of assembly, means
there is no protection via the USBOC# signal. (The "#" in the signal name,
means the current is shut off when the signal is grounded.) This is why, if
I'm shopping for a motherboard, I look at the pictures to see if it has
Polyfuses or not. A Polyfuse is a better solution, simply because computer
case assemblies are too cheap to include USBOC#.

In cases where a motherboard has USBOC#, and the wiring assembly assumes the
10th pin is a shield GND, the front panel USB may fail to work. This happens,
because the shield wire is being grounded at the computer case end, and is
being interpreted as a logic 0 by the motherboard. Which means the motherboard
cannot turn the port on. I've only run into that maybe once on these newsgroups.

In any case, I'm still curious as to how the situation turned into such a mess.
It is possible, for a Polyfuse to be present, and a solder bridge on the motherboard
to join the two sides of the fuse together. So I'd be taking a real good look at
the whole setup, to see what went wrong. Either a mismatch of technologies or
a fault on the motherboard. (And the motherboard companies using USBOC#, don't
explain anything about it. So it's not like the user is warned or anything.)

Paul
 
Paul said:
Overcurrent protection exists in two forms. Some motherboards use small
green
Polyfuses, which recover when they cool off. They are a fuse which doesn't
need to be replaced, and if present, they open at about 1.1 amps or so.
I think
as a rule, a two connector stack runs off the same fuse.

A second method is USBOC#. That is a signal passed from any adapter
assembly,
to the motherboard. The idea is, the adapter assembly (like on the front of
the computer case), measures the current. If the current is too high for
the
USB standard, the USBOC# signal is sent to the motherboard. A logic low
level means "switch off this port". Then, it is up to the motherboard to
remove the power. To cause a problem, there would have to be a double
fault, where the USB assembly or a plugged in device, shorted +5V, and
there
was also a failure of the switching device. And frying the switching
device,
should stop the flow of current, eventually.

The user can complicate life, if mixing a non-USBOC# front panel assembly,
with a USBOC# motherboard. Failure to use the right kind of assembly, means
there is no protection via the USBOC# signal. (The "#" in the signal name,
means the current is shut off when the signal is grounded.) This is why, if
I'm shopping for a motherboard, I look at the pictures to see if it has
Polyfuses or not. A Polyfuse is a better solution, simply because computer
case assemblies are too cheap to include USBOC#.

In cases where a motherboard has USBOC#, and the wiring assembly assumes
the
10th pin is a shield GND, the front panel USB may fail to work. This
happens,
because the shield wire is being grounded at the computer case end, and is
being interpreted as a logic 0 by the motherboard. Which means the
motherboard
cannot turn the port on. I've only run into that maybe once on these
newsgroups.

In any case, I'm still curious as to how the situation turned into such
a mess.
It is possible, for a Polyfuse to be present, and a solder bridge on the
motherboard
to join the two sides of the fuse together. So I'd be taking a real good
look at
the whole setup, to see what went wrong. Either a mismatch of
technologies or
a fault on the motherboard. (And the motherboard companies using USBOC#,
don't
explain anything about it. So it's not like the user is warned or
anything.)

Paul

I couldn't find this document, when I wrote the above. I missed a key point
in the doc. The USBOC# can only be used, if there is a fuse on the computer
case USB assembly. I had assumed the protection was via the power switching
device on the motherboard. In an Intel reference schematic, I've seen them
use a power switch, and assumed it was for protection. In any case, this
doc provides the info about USB and powering.

http://www.usb.org/developers/whitepapers/power_delivery_motherboards.pdf

Paul
 
Bob said:
A couple of nights ago I was reinstalling a driver with cd rom for a hp
printer. Smoke started coming out the bottom of the computer, the front usb
ports were actually so hot they melted. I disconnected everything, had to
pull the wires from case. Bottom of case had melted along with cover door
for usb. Printer was connected to back usb, fronts were not even being used.
I built the computer about a year ago, have had no problems. Front and back
usb's always worked. Computer actually still works, just no front usb now.
What could have caused this to overheat like this. Hate to think what would
have happened had I not been sitting at desk. Thanks for any info.


The old USB to "firewire" conversion it looks like.

Something must have shorted -out your USB connection
 
Back
Top