USB fuse ?

[Eeyore]

I have a (several actually) KT7A mobos including RAIDS in a few PCs.
Including one running a 2400XP CPU ! That FLIES and I picked up the
2600XP-M CPU too which has much the same settings but a larger cache
that gives it the extra speed.

Anyway, a while back, my scanner fell of the table it was on and landed
on its backside. It would no longer work, so I though I must have killed
it, even though there were volts present on all the USB connector pins.

Later, trying another USB device, the power LED would 'blip' on for a
second and then go out.

I twigged I'd probably blown a protective fuse/fusible resistor in the
USB 5V supply most likely. Probably gone high resistance and could
charge a capacitor on the mobo enough just to get that 'blip' on the
power LED.

I'm good with soldering and electronics generally but I don't know how
to locate the part. Does anyone have any clues ?

Graham

[Paul]

Post by Eeyore
I have a (several actually) KT7A mobos including RAIDS in a few PCs.
Including one running a 2400XP CPU ! That FLIES and I picked up the
2600XP-M CPU too which has much the same settings but a larger cache
that gives it the extra speed.
Anyway, a while back, my scanner fell of the table it was on and landed
on its backside. It would no longer work, so I though I must have killed
it, even though there were volts present on all the USB connector pins.
Later, trying another USB device, the power LED would 'blip' on for a
second and then go out.
I twigged I'd probably blown a protective fuse/fusible resistor in the
USB 5V supply most likely. Probably gone high resistance and could
charge a capacitor on the mobo enough just to get that 'blip' on the
power LED.
I'm good with soldering and electronics generally but I don't know how
to locate the part. Does anyone have any clues ?
Graham

look for a Polyfuse.

http://en.wikipedia.org/wiki/Resettable_fuse

They're typically green in color, with a "dimple" on
either end of them. I see three in a picture of the KT7A,
up near the I/O connectors. They can be used to protect
PS/2 mouse, PS/2 keyboard, and USB devices.

Loading Image...

Don't reach for the soldering iron just yet.
Use a multimeter, and check to see whether
+5V is available on both sides of the fuse or
not. Both before and after plugging in a USB
item.

A Polyfuse resets itself, after the overload
is removed.

Paul

[Eeyore]

Post by Paul

Post by Eeyore
I have a (several actually) KT7A mobos including RAIDS in a few PCs.
Including one running a 2400XP CPU ! That FLIES and I picked up the
2600XP-M CPU too which has much the same settings but a larger cache
that gives it the extra speed.
Anyway, a while back, my scanner fell of the table it was on and landed
on its backside. It would no longer work, so I though I must have killed
it, even though there were volts present on all the USB connector pins.
Later, trying another USB device, the power LED would 'blip' on for a
second and then go out.
I twigged I'd probably blown a protective fuse/fusible resistor in the
USB 5V supply most likely. Probably gone high resistance and could
charge a capacitor on the mobo enough just to get that 'blip' on the
power LED.
I'm good with soldering and electronics generally but I don't know how
to locate the part. Does anyone have any clues ?

look for a Polyfuse.
http://en.wikipedia.org/wiki/Resettable_fuse
They're typically green in color,

Ah, I've only seen them in Orange.

Post by Paul
with a "dimple" on
either end of them. I see three in a picture of the KT7A,
up near the I/O connectors. They can be used to protect
PS/2 mouse, PS/2 keyboard, and USB devices.

Yup. You didn't see the PCB legend did you ?

Post by Paul
http://www.phoenix1.co.uk/images/pptc_sm.jpg
Don't reach for the soldering iron just yet.
Use a multimeter, and check to see whether
+5V is available on both sides of the fuse or
not. Both before and after plugging in a USB
item.
A Polyfuse resets itself, after the overload
is removed.

So it's not a Polyfuse then.

Graham

[Paul]

Post by Eeyore
So it's not a Polyfuse then.
Graham

When I suggest the multimeter, I'm suggesting to you
that you observe the voltage on either side of the
fuse. Chances are the fuse is not at fault. What
I cannot tell you, is if there is any active device
upstream of that point. If the USB header is fed from
+5VSB and the +5VSB winked out, the computer would
reset or power off. And you don't mention that in
your symptoms. The computer would similarly crash,
if the raw +5V was dying (but that rail can offer
many more amperes, and can burn whatever gets in the
way). That leaves something like a "power bug" in the
path, an eight pin chip which monitors the current
flow closely. It has MOSFETs inside to cut off the
current flow. But if a Polyfuse is present, there is
no need for a power bug. A power bug might be seen
in a laptop for example. I haven't seen any in
desktop motherboards.

I guess what I'm saying is, I don't have any real good
suggestions for a match between your symptoms and a cause.
But you can check the fuse, by using your ohmmeter across
the Polyfuse, and verifying it reads zero ohms. (Do that
with all power off in the computer.) Then, when the computer
is running, observed the +5V levels on either side of the
fuse. For example, if you plug in a USB device, and
both sides of the fuse no longer have 5V on them, then
something upstream has switched off the power. And that
suggests there is some other kind of power management
device.

(Clip the ground lead of your multimeter, to an I/O screw
on the back panel of the computer. Then, you'll only
need one hand to probe the ends of the Polyfuse, while
the computer is running.)

If the Polyfuse has 5V on one side, and nothing on the
other, just after the USB device is plugged in. then
it has opened for some reason. It is going to be
pretty hard to put an ammeter in series with the
circuit, to measure the actual current flow, and
figure out whether the Polyfuse is tripping at
a lower than normal current.

I'm also not sure whether there are any special
soldering instructions for Polyfuses, and whether
hand soldering for them is recommended. If it was
my motherboard, I'd want to be absolutely sure
the thing was bad, before attempting a repair.

Paul