Attaching USB header to P4P800

[Peter Kronenberg]

I just got an Asus P4P800. I'm trying to figure out how to hook up
the front USB ports (what the manual refers to as the USB Header on
page 2-30) on USB_56 and USB_78. The manual has the pins labeled as
USB+5V, USB+5V, USB_P5-, USB_P6-, USB_P5+, USB_P6+, GND, and GND.
However, there are only 4 leads from each USB port on the case. They
are +5V, +D, -D, and GND.

So why do I have 4 leads to go onto 8 pins and what goes where?

thanks,
Peter

 

[Paul]

I just got an Asus P4P800. I'm trying to figure out how to hook up
the front USB ports (what the manual refers to as the USB Header on
page 2-30) on USB_56 and USB_78. The manual has the pins labeled as
USB+5V, USB+5V, USB_P5-, USB_P6-, USB_P5+, USB_P6+, GND, and GND.
However, there are only 4 leads from each USB port on the case. They
are +5V, +D, -D, and GND.

So why do I have 4 leads to go onto 8 pins and what goes where?

thanks,
Peter

There are two USB interfaces on the USB_56 header and
two USB interfaces on the USB_78 header.

Connect your four wires to the right hand column of four
pins (plus NC = no_connect) or to the left hand column
of four pins.

The two "sides" of the USB56 header.
One four wire USB interface goes on each side.

1 +5v USB+5V USB+5V 2 +5v
1 -D USB_P5- USB_P6- 2 -D
1 +D USB_P5+ USB_P6+ 2 +D
1 ground GND GND 2 ground
^ NC
|
Missing pin ---+

HTH,
Paul

 

[Peter Kronenberg]

(e-mail address removed) (Paul) wrotd:

There are two USB interfaces on the USB_56 header and
two USB interfaces on the USB_78 header.

Connect your four wires to the right hand column of four
pins (plus NC = no_connect) or to the left hand column
of four pins.

The two "sides" of the USB56 header.
One four wire USB interface goes on each side.

1 +5v USB+5V USB+5V 2 +5v
1 -D USB_P5- USB_P6- 2 -D
1 +D USB_P5+ USB_P6+ 2 +D
1 ground GND GND 2 ground
^ NC
|
Missing pin ---+

HTH,
Paul

Thanks! That helps a lot

Peter

 

[Darkfalz]

I just got an Asus P4P800. I'm trying to figure out how to hook up
the front USB ports (what the manual refers to as the USB Header on
page 2-30) on USB_56 and USB_78. The manual has the pins labeled as
USB+5V, USB+5V, USB_P5-, USB_P6-, USB_P5+, USB_P6+, GND, and GND.
However, there are only 4 leads from each USB port on the case. They
are +5V, +D, -D, and GND.

So why do I have 4 leads to go onto 8 pins and what goes where?

You have USB 1.1 connectors on your case, but your board uses USB 2.0.

 

[Peter Kronenberg]

You have USB 1.1 connectors on your case, but your board uses USB 2.0.

Does that mean I can't use them? Is there any physicall difference
between USB 1.1 and USB 2.0?

 

[Darkfalz]

Does that mean I can't use them? Is there any physicall difference
between USB 1.1 and USB 2.0?

Actually I'm not sure. I think maybe the second row of pins is just to
connect two USB ports at a time.

 

[Paul]

Does that mean I can't use them? Is there any physicall difference
between USB 1.1 and USB 2.0?

USB 2.0 is backward compatible with USB 1.1.

Both carry +5V down the cable, to power bus powered devices with
a maximum of 0.5 amps.

A USB 2.0 device will talk at USB 1.1 rates for a short interval
during the setup of the device.

A USB 1.1 device uses differential signalling (that is why there
is a D+ and a D- signal). The signals on the wires are the inverse
of one another. The amplitude is based on up to 3.3V signal swing
(i.e. the I/O pins are powered by what looks like a regulated 3.3V
source, possibly regulated from the bus power +5V signal).

A USB 2.0 device, when talking at 480Mb/sec, pumps current into the
line, such that 0.4V is developed. In a sense, it uses differential
signalling as in USB 1.1, but the signals are smaller in amplitude.
The smaller amplitude reduces electrical interference with devices
external to the cable.

About the only question you have to ask of your equipment, is whether
the components are rated to carry a 480Mb/sec signal. Due to the
legacy of USB 1.1 devices, there are still components around,
such as the wiring on the case of your computer, that may only
be suitable for carrying the lower speed signals. To carry 480Mb/sec
requires good transmission line quality cable that controls the
impedance and loss over the length of the cable.

For example, if the case has long lengths of loose wires for the
D+ and D-, that spells doom for the USB 2.0 signals. Loose wires
must be kept very short, and the majority of the wire should
be shielded.

For example, look at the way this Asus USB/Game Port bracket is
constructed. There is only a short length of unshielded wire
near the blue connector. The rest of the wire is carefully
shielded.

http://www.newegg.com/app/Showimage.asp?image=13-131-206-01.JPG

If your USB wiring has a lot of loose looking wires like some
of the cables in this picture, it is less likely to work at
USB 2.0 rates.

http://www.jmallard.com/MyComputer/LianLiUSB.htm

In the case of Antec, people have been contacting Antec and
getting a replacement USB assembly rated for use at USB 2.0
rates.

When the wiring is bad, I presume what happens is the equipment
tries to run at USB 2.0, get a lot of CRC errors, and downshifts
to USB 1.1 rates. Obviously, your chipset has to support USB 2.0
to even try to use the protocol, and you need an "Enhanced" entry
in the USB entries of the Device Manager to indicate that a
USB 2.0 driver is installed.

The standards are here (I didn't read the whole thing today...)

http://www.usb.org/developers/docs/usb_20.zip (650 pages)
http://www.usb.org/developers/docs/usbspec.zip (327 pages)

HTH,
Paul

 

[Peter Kronenberg]

A USB 2.0 device will talk at USB 1.1 rates for a short interval
during the setup of the device.

So, is there anyway to test whether your USB ports are running at full
speed, or if they will properly support 2.0?

 

[Paul]

So, is there anyway to test whether your USB ports are running at full
speed, or if they will properly support 2.0?

I haven't heard of a way to do it.

The question is, whether the USB interface on the Southbridge or
on some PCI USB chip, has a counter that can count the CRC errors
detected on the line. If you could gain access to that counter,
that would be the key to testing the cable. For example, if
a USB2 disk drive is connected to a cable, and then you transfer
lots of data to the disk, the USB CRC counter would be a good
way to evaluate the quality of the interconnect.

I suspect this is just a matter of no one bothering to write
an application to do this.

Perhaps the USBHSET toolkit can do this ? Give it a try and
tell us whether it can do such a test. (I don't have any USB2
devices to test right now.)
http://www.usb.org/developers/tools#usbhset

This site has some info on USB2:
http://usbman.com/USB 2 News.htm

Paul