USB OverCurrent Status Detected

[Man-wai Chang]

I removed one USB device and the error went away.

1. Could I conclude that it's just that device not the motherboard?

2. How could I check whether that USB device short-circuited?

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.37.1
^ ^ 22:12:01 up 1 day 5:58 0 users load average: 1.04 1.09 1.13
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

 

[Flasherly]

I removed one USB device and the error went away.

1. Could I conclude that it's just that device not the motherboard?

2. How could I check whether that USB device short-circuited?

Pin diagram and maybe breakout for getting to appropriate leads for
one of these...?

http://www.dealextreme.com/p/multimeter-xiole-dt9205a-yellow-black-large-619

assuming these's no strobes to work w/ or anything else off the scale
- should be logic.

(Took 3 months, but got mine free, like that one, except I've a freq
counter Fx and temperature probe).

 

[Paul]

I removed one USB device and the error went away.

1. Could I conclude that it's just that device not the motherboard?

2. How could I check whether that USB device short-circuited?

I don't know all of the workings of the USB interface. For example,
each USB device has a current rating stored in its configuration
information. I don't know if that is used to determine whether
a port is overloaded.

In terms of hardware protection, the Southbridge has a number
of OC# signals. These are active low signals. Grounding one of
those signals, should cause a USB overcurrent dialog to appear.

On a laptop, a "power bug" chip is used. That is an 8 pin DIP
chip. It has a precision current sense inside. If the current
exceeds 500mA, some FET transistors inside turn off the power
feeding the USB port. At the same time, a logic signal from the
power bug, is sent to one of the OC# pins on the Southbridge.
As a result, the laptop doesn't need a fuse.

On a desktop, instead of spending money on a power bug, they
use a Polyfuse. (Cooling is better inside a desktop computer,
so high operating temperatures aren't an issue.) Typically,
one fuse is shared by two USB ports. The Polyfuses on my motherboards
here, have "1.1" printed on them, implying they will allow
about 1 ampere to flow, to a total of two ports. If only one
USB port has a device plugged in, it could draw up to 1 ampere
before the fuse opens. Polyfuses automatically recover when
they cool off, so the fuse does not need to be replaced. It
has a conductive polymer inside.

For an example of how it works, download BXDPDG10.PDF (if
it's still available).

http://www.intel.com/design/chipsets/designex/BXDPDG10.PDF

Go to PDF page 26. In the upper left hand corner, you can see
fuse F1. On a modern motherboard, that would be a Polyfuse,
with a lower current rating. The Polyfuse is automatically
resetting, so doesn't need to be replaced.

Notice on page 26, how the OC#0 signal works. If the fuse
is intact and working, both sides of the fuse are lifted to +5V.
The resistor network is lifted to +5V and a logic 1 is sent to
OC#0. That tells the overcurrent sensor that everything is
OK.

Now, pretend to overload the circuit. The left hand side of the
fuse is +5V. The right hand side (due to the short circuit), is
now at zero volts. The fuse goes open circuit. Now, the resulting
voltage value at OC#0 is zero volts as well. That is the
active condition for the OverCurrent signal, and now there is
a dialog on the computer screen, indicating an overcurrent has
occurred.

Therefore, to test a desktop motherboard, check the voltage on
either side of the Polyfuse. If it's 5V on both sides, the
fuse is intact. If it's 5V on one side, and 0V on the other,
you've got an overload.

Many of my motherboards, use green Polyfuses like this one.
A "dimple" on the end, helps differentiate the Polyfuse from
other component types. The Polyfuse should be near the USB stack
or USB 2x5 header. A motherboard with ten USB ports, should have
five Polyfuses total, protecting the ports. Make sure you're probing
the appropriate one when testing. A Polyfuse may also protect
the Parallel I/O port, and the keyboard and mouse PS/2 ports.

http://i01.i.aliimg.com/photo/v0/215265281/Poly_Fuse.jpg

Best guess,
Paul

 

[Man-wai Chang]

Now, pretend to overload the circuit. The left hand side of the

fuse is +5V. The right hand side (due to the short circuit), is
now at zero volts. The fuse goes open circuit. Now, the resulting
voltage value at OC#0 is zero volts as well. That is the
active condition for the OverCurrent signal, and now there is
a dialog on the computer screen, indicating an overcurrent has
occurred.

Therefore, to test a desktop motherboard, check the voltage on
either side of the Polyfuse. If it's 5V on both sides, the
fuse is intact. If it's 5V on one side, and 0V on the other,
you've got an overload.

Many of my motherboards, use green Polyfuses like this one.
A "dimple" on the end, helps differentiate the Polyfuse from
other component types. The Polyfuse should be near the USB stack
or USB 2x5 header. A motherboard with ten USB ports, should have
five Polyfuses total, protecting the ports. Make sure you're probing
the appropriate one when testing. A Polyfuse may also protect
the Parallel I/O port, and the keyboard and mouse PS/2 ports.

http://i01.i.aliimg.com/photo/v0/215265281/Poly_Fuse.jpg

Way too complicated for me.... There were some sudden blue-screen since
I attached that USB device. That USB device is actually Mygica TD312
5-in-1 TV+Radio tuner.

Without plugging the tuner back to the motherboard (taking the risk of
over-current), should I just take that tuner back to the distributor and
let their PCs test it?


--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.37.1
^ ^ 00:32:01 up 1 day 8:18 0 users load average: 1.10 1.08 1.11
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

 

[Man-wai Chang]

With the error message you got it sounds like the port protected itself

from the overload the device was providing. You can try using an hub
that has an external power supply and see if it will allow a the device
to function.

If there was really a short-circuit in the USB device, plugging it back
to the motherboard might risk killing the USB port if not the whole board...

What was the device giving you problems?

Mygica TD312 5-in-1 TV/Radio Tuner.

I have had some USB hard drives that were powered by the port that
worked initially but as they aged they drew more current than the port
allowed. Using an external 5 volt power supply on them "fixed" the
problem for me.

I have never had an "USB over-current" problem. This is the first time I
met one.

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.37.1
^ ^ 00:32:01 up 1 day 8:18 0 users load average: 1.10 1.08 1.11
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

 

[Man-wai Chang]

http://www.dealextreme.com/p/multimeter-xiole-dt9205a-yellow-black-large-619

assuming these's no strobes to work w/ or anything else off the scale
- should be logic.

(Took 3 months, but got mine free, like that one, except I've a freq
counter Fx and temperature probe).

So I could just check the 4 pins in the USB device for short-circuit?

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.37.1
^ ^ 00:32:01 up 1 day 8:18 0 users load average: 1.10 1.08 1.11
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

 

[VanguardLH]

I removed one USB device and the error went away.

1. Could I conclude that it's just that device not the motherboard?

2. How could I check whether that USB device short-circuited?

The maximum current across a pair of USB ports is 500mA *total*. One
USB controller manages 2 USB ports, so its 500mA maximum is shared
across the 2 ports that it manages. Plugging in a high-current USB
device, like a headset, in one port may work if the other USB port is
connected to a low-current USB device, but it may not. Sometimes the
current is so high for one USB device that you cannot use the other
paired USB port.

Self-powered (not port-powered) USB hubs with their own power adapter
should provide 500mA to each port so you don't run into the current
limitation with USB ports connected to the motherboard in the system
case; however, then you have the mess of cabling for the USB cable
(system unit to hub) along with all the other USB device cables along
with the power cable and transformer brick.

You never mentioned WHERE is the USB port. Is it in the system case or
as a separate hub? If a hub, is it a USB-powered hub (that gets its
power from the case port) or is it a powered hub with its on AC adapter
and power cord? You never mentioned WHAT device you were plugging into
the USB ports, especially for the paired USB ports. Headset, speakers,
external hard disk, or what?

The manual should tell you if the device is a high- or low-power unit
but I've found enough examples where they didn't mention that fact which
is necessary to determine how to distribute the 500mA load across the
paired USB case ports. http://www.mygica.com/download/d_manual.asp
doesn't show "312" so I can't find info on that model. Guess you'll
have to contact them to get help. I looked at one of their manuals
(http://www.mygica.com/down-eng/manual/atsc/u6850amanual.pdf) but it
never mentioned the power load on the USB port. I looked at some of
their product specs but they still didn't mention the power load. Guess
you'll be visiting http://www.mygica.com/index-5.asp.

 

[Paul]

Way too complicated for me.... There were some sudden blue-screen since
I attached that USB device. That USB device is actually Mygica TD312
5-in-1 TV+Radio tuner.

Without plugging the tuner back to the motherboard (taking the risk of
over-current), should I just take that tuner back to the distributor and
let their PCs test it?

Polyfuses can open more than once.

First, test that the USB port still works, by using a USB mouse or other light load.

*******

The advertisement here, is quite insistent that power consumption is low.
But they don't state exactly how low it is.

The description suggests it's a two chip solution, a silicon tuner plus
a processing chip of some sort. They use the term "realtek", which is
a company that makes a few different kinds of chips, but I see no
evidence that Realtek makes TV/radio stuff. Your product is also not
listed on the Mygica web site.

http://translate.googleusercontent....le.com&usg=ALkJrhgLcS-j-2aObvO1Loh1cI4Tr5GqZg

"Finger-type plug-in USB2.0, extremely low power consumption, the use of persistent"

Take it back to the distributor and have them test it.

If you want a safer, more precise way to test, try plugging it into
a laptop. The laptop, using the "power bug" chip, will cut off the
power safely and precisely at 500mA. It is much more accurate than
the Polyfuse. The Polyfuse is a bit forgiving of overcurrent,
because the typical desktop computer will allow 1 ampere to flow
before the fuse opens. If the thing is drawing 1 ampere, that
suggests it's well out of spec.

You might also visually examine the USB connector, for flaws or for
broken conductors. USB connectors don't have a good strain
relief, between the connector and the PCB, and snap easily.
Many people have ruined USB flash sticks, because of connector
problems. I've got one USB flash stick now, where the connector
is loose on it.

Paul

 

[Paul]

The maximum current across a pair of USB ports is 500mA *total*. One
USB controller manages 2 USB ports, so its 500mA maximum is shared
across the 2 ports that it manages. Plugging in a high-current USB
device, like a headset, in one port may work if the other USB port is
connected to a low-current USB device, but it may not. Sometimes the
current is so high for one USB device that you cannot use the other
paired USB port.

Self-powered (not port-powered) USB hubs with their own power adapter
should provide 500mA to each port so you don't run into the current
limitation with USB ports connected to the motherboard in the system
case; however, then you have the mess of cabling for the USB cable
(system unit to hub) along with all the other USB device cables along
with the power cable and transformer brick.

You never mentioned WHERE is the USB port. Is it in the system case or
as a separate hub? If a hub, is it a USB-powered hub (that gets its
power from the case port) or is it a powered hub with its on AC adapter
and power cord? You never mentioned WHAT device you were plugging into
the USB ports, especially for the paired USB ports. Headset, speakers,
external hard disk, or what?

The manual should tell you if the device is a high- or low-power unit
but I've found enough examples where they didn't mention that fact which
is necessary to determine how to distribute the 500mA load across the
paired USB case ports. http://www.mygica.com/download/d_manual.asp
doesn't show "312" so I can't find info on that model. Guess you'll
have to contact them to get help. I looked at one of their manuals
(http://www.mygica.com/down-eng/manual/atsc/u6850amanual.pdf) but it
never mentioned the power load on the USB port. I looked at some of
their product specs but they still didn't mention the power load. Guess
you'll be visiting http://www.mygica.com/index-5.asp.

It's 500mA per port on a port on the computer.

Which is why the fuse feeding a dual USB stack
is set to 1 ampere or so. If you use only one of
the two ports in a stack, that scheme also allows
you to exceed the 500mA stated in the standard.

Laptops on the other hand, police power on a per port basis, so
it's harder to draw an excessive amount of current there. Exactly
why they bother using the 8 pin chip to police power, instead
of a PolyFuse, isn't clear. Perhaps the PolyFuses don't do
well when the surrounding temperature is elevated ?

Paul

 

[VanguardLH]

It's 500mA per port on a port on the computer.

500-900 mA per *controller* depending on USB version (which was NOT
mentioned by the OP). Each controller manages 2 ports across which the
total controller current draw is regulated. Since the OP never
mentioned the USB version, it is likely that he has USB 1.1/2.0 ports in
his case from his motherboard. That means 500 mA max across the 2
ports.

Up to USB 2.0, a unit-load was 100 mA and up to 5 loads were supported
for output power from the controller chip. In USB 3.0, a unit-load is
150 mA and up to 6 loads are supported. The total unit-loads are
measured across the paired USB ports managed by the controller.

USB 2.0: 5 unit loads x 100 mA/unit-load = 500 mA

USB 3.0: 6 unit loads x 150 mA/unit-load = 900 mA

Low-power devices are supposed to draw a maximum of 1 unit-load (100
mA). High-power devices can draw the full maximum load from the
controller, so one of the paired ports could suck up all 5 unit-loads
for USB 2.0 or all 6 unit-loads for USB 3.0 leaving nothing available
for the other paired port.

You would be correct if the OP had USB 3.0 ports. I doubt the OP has a
motherboard that provides USB 3.0 headers to connect to the case ports.
Can't tell what the OP has. No details.

 

[Paul]

500-900 mA per *controller* depending on USB version (which was NOT
mentioned by the OP). Each controller manages 2 ports across which the
total controller current draw is regulated. Since the OP never
mentioned the USB version, it is likely that he has USB 1.1/2.0 ports in
his case from his motherboard. That means 500 mA max across the 2
ports.

Up to USB 2.0, a unit-load was 100 mA and up to 5 loads were supported
for output power from the controller chip. In USB 3.0, a unit-load is
150 mA and up to 6 loads are supported. The total unit-loads are
measured across the paired USB ports managed by the controller.

USB 2.0: 5 unit loads x 100 mA/unit-load = 500 mA

USB 3.0: 6 unit loads x 150 mA/unit-load = 900 mA

Low-power devices are supposed to draw a maximum of 1 unit-load (100
mA). High-power devices can draw the full maximum load from the
controller, so one of the paired ports could suck up all 5 unit-loads
for USB 2.0 or all 6 unit-loads for USB 3.0 leaving nothing available
for the other paired port.

You would be correct if the OP had USB 3.0 ports. I doubt the OP has a
motherboard that provides USB 3.0 headers to connect to the case ports.
Can't tell what the OP has. No details.

500ma per port. (The OPs TV stick is USB2)

http://www.usb.org/developers/compliance/check_list/compchksys080205.pdf

"Can the system supply 0 to 500mA on each of its downstream ports..."
"Can the system maintain Vbus between 4.75 and 5.25V at all of its
downstream connectors for DC loads between 0 and 500mA per
downstream port?"

There is a separate specification for hubs. I was specifically discussing
the implementation of motherboards, not hubs. A motherboard has an
oversized fuse, and to save money, one fuse is used per two ports.
The fuse has a capacity slightly larger than 2x500mA. The ones I have
here are stamped 110 or 1.1. Actually, my newest motherboard has one
marked 160, implying slightly more capacity.

There is a discussion about sizing of current protection devices here,
where they recommend oversizing the Polyfuse, to prevent Vbus droop.
So Asus could have gone even higher with their fuse choice if they
wanted. (It's for safety, not policing.) The Asus implementation looks
like Figure 2 on page 4. At least with respect to providing a Polyfuse
and bypass cap per port pair. The cap provides holdup current during
downstream inrush.

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

Paul

 

[Paul]

500ma per port. (The OPs TV stick is USB2)

http://www.usb.org/developers/compliance/check_list/compchksys080205.pdf

"Can the system supply 0 to 500mA on each of its downstream ports..."
"Can the system maintain Vbus between 4.75 and 5.25V at all of its
downstream connectors for DC loads between 0 and 500mA per
downstream port?"

There is a separate specification for hubs. I was specifically discussing
the implementation of motherboards, not hubs. A motherboard has an
oversized fuse, and to save money, one fuse is used per two ports.
The fuse has a capacity slightly larger than 2x500mA. The ones I have
here are stamped 110 or 1.1. Actually, my newest motherboard has one
marked 160, implying slightly more capacity.

There is a discussion about sizing of current protection devices here,
where they recommend oversizing the Polyfuse, to prevent Vbus droop.
So Asus could have gone even higher with their fuse choice if they
wanted. (It's for safety, not policing.) The Asus implementation looks
like Figure 2 on page 4. At least with respect to providing a Polyfuse
and bypass cap per port pair. The cap provides holdup current during
downstream inrush.

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

Paul

One other thing just came to mind. Your computer may be small enough and
low power enough, that they're using the power supply fan to cool the
whole computer. If that is the case, you'd better be careful about
your system cooling. If the new power supply doesn't move enough air
for you, you may need to add a second (case) cooling fan.

My first computer is like that. There is no room on the back of the
computer, for an 80mm or 120mm fan for case cooling. The power supply
provides the only cooling fan. To fix that, I bolted an intake fan to
the front of the computer.

Paul

 

[Man-wai Chang]

Take it back to the distributor and have them test it.

Sure!

If you want a safer, more precise way to test, try plugging it into
a laptop. The laptop, using the "power bug" chip, will cut off the
power safely and precisely at 500mA. It is much more accurate than
the Polyfuse. The Polyfuse is a bit forgiving of overcurrent,
because the typical desktop computer will allow 1 ampere to flow
before the fuse opens. If the thing is drawing 1 ampere, that
suggests it's well out of spec.

Really?

You might also visually examine the USB connector, for flaws or for
broken conductors. USB connectors don't have a good strain
relief, between the connector and the PCB, and snap easily.
Many people have ruined USB flash sticks, because of connector
problems. I've got one USB flash stick now, where the connector
is loose on it.

Didn't notice anything. Anyway, that device had earlier caused a few
blue-screens in Window$.

--
@~@ Might, Courage, Vision, SINCERITY.
/ v \ Simplicity is Beauty! May the Force and Farce be with you!
/( _ )\ (x86_64 Ubuntu 9.10) Linux 2.6.37.1
^ ^ 15:34:01 up 1 day 23:20 0 users load average: 1.10 1.09 1.09
ä¸å€Ÿè²¸! ä¸è©é¨™! ä¸æ´äº¤! ä¸æ‰“交! ä¸æ‰“劫! ä¸è‡ªæ®º! è«‹è€ƒæ…®ç¶œæ´ (CSSA):
http://www.swd.gov.hk/tc/index/site_pubsvc/page_socsecu/sub_addressesa

 

[VanguardLH]

500ma per port. (The OPs TV stick is USB2)

http://www.usb.org/developers/compliance/check_list/compchksys080205.pdf

"Can the system supply 0 to 500mA on each of its downstream ports..."
"Can the system maintain Vbus between 4.75 and 5.25V at all of its
downstream connectors for DC loads between 0 and 500mA per
downstream port?"

There is a separate specification for hubs. I was specifically discussing
the implementation of motherboards, not hubs. A motherboard has an
oversized fuse, and to save money, one fuse is used per two ports.
The fuse has a capacity slightly larger than 2x500mA. The ones I have
here are stamped 110 or 1.1. Actually, my newest motherboard has one
marked 160, implying slightly more capacity.

There is a discussion about sizing of current protection devices here,
where they recommend oversizing the Polyfuse, to prevent Vbus droop.
So Asus could have gone even higher with their fuse choice if they
wanted. (It's for safety, not policing.) The Asus implementation looks
like Figure 2 on page 4. At least with respect to providing a Polyfuse
and bypass cap per port pair. The cap provides holdup current during
downstream inrush.

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

Paul

http://www.usb.org/developers/docs/

USB 2.0 specification:
http://www.usb.org/developers/docs/usb_20_021411.zip
Open the usb_20.pdf file.
Page 206, 7.3.2 Bus Timing/Electrical Characteristics
High-power output: 500mA
Low-power output: 100mA

Now head over to section 7.2.1.2 on self-powered hubs (which includes
the root hub). Yes, there are multiple data *ports* listed. That's for
the data lines. They get all their *power* through the [downstream]
Vbus. In this diagram, it shows 5 unit-loads (500 mA) for USB 2.0 for
the power delivered to EACH port, so it *looks like you are correct*.

I don't know when I got misled regarding the power output of EACH port.
usb.org doesn't have the USB 1.0/1.1 spec for me to see if it was
something from that era that I forget to update in my memory. I managed
to find http://www.scaramanga.co.uk/stuff/qemu-usb/usb11.pdf where it
said, section 7.2.1, for root hubs, "Systems that obtain operating power
externally, either AC or DC must supply at least five unit loads to each
port." So it's been that way since USB 1.1. According to
http://www.usbman.com/Guides/USB_version1.0_version1.1.htm, what I read
for 1.1 also applies to 1.0.

My only guess at this point is that I mixed in bus-powered and
self-powered hubs together. In the USB 1.1 pdf file, section 7.2.1.1,
it shows that the 5-load upstream power is getting distributed as 1-load
power across the output ports. Self-powered hubs (which include root
hubs or external powered hubs) jump back to 5-load power per port.

While I've found some USB controller chip specs, so far they just list
the purpose of each pin, not the power output of each pin (or the total
power for VSB across all downstream ports). I did find one spec sheet
at http://www2.renesas.com/maps_download/pdf/S16265EJ5V0DS00.pdf where
section 2.2 shows VBUS (from Power switch output) shared by all
downstream ports. Searching on "vbus" didn't find the "power switch
output" mentioned in the diagram for me to find its max power output. I
found the voltage ranges listed for the Vo outputs but not their power.
Page 8 lists the pinout yet nothing sprang out as the VBUS output from
the controller chip (and no Vo listed there, either). Page 11 mentiones
Ioh (high-level output current) yet everywhere Ioh is mentioned doesn't
come close to 500 mA or a 5-load measure. Guess it's a secret.

I don't know how often the above USB controller chip is used. It was
just one that I found in a Google search. I've seen Via chips on mobos
(http://www.via.com.tw/en/products/peripherals/usb/vt6212/) but they
don't provide the technical details for their pinouts and electrical
characteristics. No luck there.

I found an example USB schematic at http://www.alanmacek.com/usb/ using
the PIC 16C745 controller and a spec sheet for it at
http://ww1.microchip.com/downloads/en/devicedoc/41124c.pdf. Page 127
lists max sink current to the ports which is 200 mA and which looks to
be a combined draw across multiple ports. Maybe I happen to pick
controllers that rely on an external regulator circuit/chip for the Vbus
line to the USB ports. Geez, they don't make it easy to figure out.