USB Problem P4C800 Deluxe

  • Thread starter Thread starter J.Botelis
  • Start date Start date
J

J.Botelis

Front of case USB plugged into USB header on board. Won't recognize my
camera and gives me unknown device warning.
Back of case everything works great. Front works with old joy stick.
Help!!!
 
Front of case USB plugged into USB header on board. Won't recognize my
camera and gives me unknown device warning.
Back of case everything works great. Front works with old joy stick.
Help!!!

I hope you are aware of the "latchup" problem with Intel ICH4, ICH5,
ICH5R based motherboards. The problem has to do with plugging in USB
devices while the computer is running. The end result is the
Southbridge chip (the ones listed above), gets red hot and dies.
It gets red hot, because a phantom path has formed inside the chip
that conducts amps of current, causing destruction of the chip.
This is basically an Intel design defect, as chips should not
normally be as sensitive to this problem, as the Intel chips
appear to be.

http://tw.giga-byte.com/Motherboard/Support/FAQ/FAQ_456.htm

If you have a USB port that is not functioning properly, do not
torture it. Either RMA the board and get it fixed. Or, use the
ports on the back of the computer. Or, for greater safety, use
a separate PCI USB2.0 card for your USB devices. By using a
separate PCI card, you no longer gamble by using the
Southbridge USB ports. (Also, for greater safety, I would
disconnect the USB wires from the Asus 2x5 USB headers, to
reduce the chances of induced static discharge into the
USB headers.)

It is unclear what the failure mechanism is. Whether failure
is being causes by static discharge into the data lines on the
USB port, or is somehow related to the power path. I looked
at an Intel reference schematic, and power to the USB ports
is switched by a SI4501. That buffers the control path to
the USB power switch, and likely eliminates the power path
as the location of the failure. That leaves the USB data
inputs (D+ and D-) as the path being taken by the static
discharge. For static to do this, the static discharge can
either flow straight down the D+ and D- wire, and into
the Southbridge. Or, the D+ and D- wires can have current
flow induced into them, by a static discharge flowing in an
adjacent conductor.

HTH,
Paul
 
Is your case an Antec ?
If so you can email Antec for a free replacement of the front usb panel.
 
I have the same problem with usb 2.0 mem key and a Antec case. I think
that is the configuration of usb front panel cable.
 
Thanks Paul, would the same apply to the fire wire connector?

Firewire is done by the VT6307 chip, a chip separate from the
Intel Southbridge. It won't suffer from the exact same issue.

There are occasionally problems with Firewire ports. The claim
is, that some of the failures are caused by bad cables, or
plugging in the connector wrong. The only analysis attempted
on Firewire problems, was done by Weibetech. This article
should explain some of the exposures.

http://www.wiebetech.com/pressreleases/FireWirePortFailures.htm

HTH,
Paul
 
Malam said:
Hi Paul:

Does the ICH5R used by the P4C800E Deluxe have the same problems ?
Thanks.

All three chip types are possible candidates for failure.
ICH4 (an older Southbridge), and the ICH5 and ICH5R are actually
the same chip, just with enumeration differences. Any motherboard
(doesn't matter what brand) with one of those Southbridge chips,
could suffer from the problem.

I have not heard any stories about ICH6 chips yet, but there
are not as many of those in circulation.

And from the stories being recounted here, only a separate
USB2 PCI card, is going to make USB ports a safe thing to
use, on those motherboards. That means I will be picking up
a USB2 PCI card for my P4C800-E Deluxe.

Paul
 
I have the same problem with usb 2.0 mem key and a Antec case. I think
that is the configuration of usb front panel cable.

There are several Antec cases with wiring errors on the front
port connectors. On one of my Antec cases, the (+) and (-)
data lines are reversed on the Firewire port.

If you own an Antec case, I recommend finding a picture of
the pinout of the ports, then using a multimeter set to the
ohms range, to verify the name printed on each wire, corresponds
to the correct pin on the front of the computer. Even a
$20 meter is good enough for this kind of work.

In the wiring errors I've found so far on Antec cases, no
errors involved power and ground wires (fortunately). Mixing
up power and ground can be a very expensive error, damaging
the devices you plug into the computer. Burning up a camera
that connects to a Firewire port, can cost hundreds of dollars
to fix, and should offer an incentive to check the wiring.

If you are not sufficiently skilled or motivated to check
the computer case front wiring, simply use the ports on
the back of the computer. The wiring on those will be
safe to use.

The reference to contacting Antec and getting a replacement
front port assembly, has to do with the USB 1.1 versus USB 2.0
issue. There are a number of older Antec cases, where the
circuit board holding the ports was designed for USB 1.1 rates
only. If you contact Antec, they can send you a replacement
assembly known to work at USB 2.0 rates. Antec did have a
web page, listing all the models that only had USB 1.1 ports
on the front of the computer, and the list was rather large.
Antec has since removed that web page, as I cannot find it
or any evidence of its existence. All I can tell you, is
my jaw dropped, when I saw how many cases had the crappy
USB board in them.

I've spent a couple of hours searching for the Antec web page
that lists which cases have USB 1.1 front ports and which have
USB 2.0 front ports. I still cannot find the link, but found
an alternate source of computer case info. If you go to
the www.intel.com web site, and enter your Antec computer case
model number, like "1080AMG", Intel has tested their motherboards
in various computer cases. A column in the Intel test result
table, shows whether the Antec case supports USB 2.0 or not.
Here is an example page, that happens to list the 1080AMG.

(Scroll down)
http://www.intel.com/cd/channel/res..._inst_info/proc_tested_source_lists/35832.htm

On my Antec Sonata case, the front port circuit board has
the words "USB 2" printed right on the board, so that
is an example of a case that is not affected. If your
circuit board says "USB 2" on it, give it a try and
see if it actually works at USB 2.0 rates.

HTH,
Paul
 
the Southbridge. Or, the D+ and D- wires can have current
flow induced into them, by a static discharge flowing in an
adjacent conductor.

Paul, I enjoy reading your posts. You really know your s*&t!

Thanks for your contributions to this newsgroup.
 
Paul, I enjoy reading your posts. You really know your s*&t!

Thanks for your contributions to this newsgroup.

I agree 100% with Tony. Paul's posts are very informative.
One of the few people on Usenet I actually filter to keep.
 
Paul, is there a speed or performance difference between the built-in
USB port or that from a PCI card?
Thanks.
 
Is this implying that not all the USB 2 ports on the motherboard is
risky to use ? Only the last 4 uses the connectors on the
motherboard, are these the risky ones ?
Thanks.
 
Malam said:
Paul, is there a speed or performance difference between the built-in
USB port or that from a PCI card?
Thanks.

No practical difference.

A USB 2.0 PCI card has 60MB/sec of transfer rate, to share
over all the USB connectors. This fits easily in the
100-110MB/sec practical transfer rate limits of the PCI bus.

While it may be possible to construct a pathological case,
where bandwidth limits might be an issue, I doubt you will
run into them. I think a USB disk enclosure I used, couldn't
transfer more than 30-40MB/sec, so that much bus traffic
should not be a problem. Many USB bridge devices have
internal limits that prevent their full potential from
being realised.

Paul
 
Malam said:
Is this implying that not all the USB 2 ports on the motherboard is
risky to use ? Only the last 4 uses the connectors on the
motherboard, are these the risky ones ?
Thanks.

There are two issues of risk here. The first risk, is that
static discharge will enter a USB port connected to the
Intel Southbridge. The ICH5/ICH5R Southbridge supports
a total of eight USB ports. Four ports are on the back of
the computer. Four ports are provided via USB56 and USB78
headers. Based on user reports, none of the ports is immune
to static induced latchup failure.

The second kind of risk, is associated with the wiring
provided on certain brands of computer cases. If there is
a wiring error, due to the wires in the computer being
mislabelled, then you could damage a device when it is
plugged into a miswired port. Since the connectors on
the back of the computer are pre-wired, there is never
a concern of a wiring error ruining a peripheral.

In other words, with P4C800/P4P800 boards, you should
not use any of the USB ports offered on the motherboard.
A separate PCI USB 2.0 card is the right solution to use
for such a board.

For models of motherboards that do not use ICH4/ICH5/ICH5R,
your main concern should be the correct wiring of the
computer case front ports. If you use the front computer
case ports, you should double check that the wiring is
correct, before using a front port on the computer. An
ohmmeter is a useful tool for checking the correctness
of the case manufacturer's wiring.

HTH,
Paul
 
Back
Top