Dev said:
We are system integrators. Recently discovered a new bug in a wide
series of Intel Motheboards here in India. Maybe this problem will
persist for all these boards across the globe.
Problem:-
Portable hard disk without external power (e.g. Western Digital
Passport, Elelemets) are not recognised.
Motherboards in concern are:-
DG33FB
D945GCNL
As the "DG" series is a recent series so we suspect the the entire
range recently launched will be effected.
Against our ticket, INTEL HAS ACCEPTED THAT THERE IS A FAULT IN THE
BOARDS WHERE AS THE EXPECTED VOLTAGE OF 5V AS REQUIRED BY THESE HARD
DISKS ARE NOT SUPPLIED.
No solution is yet provided to us our complaint is more than a month
old, perhaps they have no solution apart from a recall. As portable
hard disks are the order of the day these days, this is a very
critical bug. I guess buying an Intel Board with this defect is not
suggested.
Regards,
Dev
Have you tested with a USB Y-cable ?
http://www.compusa.com/applications/searchtools/item-details.asp?EdpNo=3617627&Sku=C184-28107
The connector on the left here, has one USB with both data and power
connections (+5V,D+,D-.GND). The second connector on the left
has +5V and GND and no data connections. The purpose of the
+5V and GND connector, is to allow extra current to flow.
http://secure.serverlab.net/shop/graphics/00000001/yCable.jpg
Say the back of the computer has four USB connectors, arranged
in two stacks. You would place only one connector per stack, and
do your external disk test. Like this -
+------+ +------+
| XXXX | | XXXX | Plug one connector per stack
+------+ +------+ Don't place both connectors on
the same stack, for your first test.
+------+ +------+
| | | |
+------+ +------+
USB stack USB stack
of two of two
The problem is, some hard drives draw more than 500mA
during the spinup of the hard drive. For example, if
you are using a 7200RPM 2.5" drive, maybe the current
is higher than a single USB port can sustain.
To substantiate your claim of a "global design issue", you
should do a load test, and discover just exactly what
the load line for the USB port looks like. USB ports
do have current limiting devices on them, to prevent
accidents, and the fuse is usually set to a reasonable
value like 1.1 amps or slightly higher. You can examine
the motherboard for a Polyfuse, and read the fusing value
off the top of the component.
On a laptop, the situation may be slightly different, in
that an 8 pin protection chip is connected to the ports.
Those can accurately measure 500mA, and cut off the port
when the current is exceeded. But those are not typically
used on desktop motherboards - instead a Polyfuse is used.
http://en.wikipedia.org/wiki/Resettable_fuse
On some of the ones on my motherboard here, they have "110"
marked on the top of them. Or perhaps "1X1", which also
means 1.1 amps.
http://web.archive.org/web/20040804122001/http://www.wickmann.com/products/SMD1812.pdf
AFAIK, Littelfuse bought the Wickmann company, and the
marking here uses a letter, like the letter H equals
1.1 amps. The 1.1 amps is shared by two USB ports in a
stack.
http://www.littelfuse.com/data/en/Data_Sheets/Littelfuse_1210L.pdf
That is the basic concept - well designed USB ports have
protection against short circuits. And that can prevent
an external 2.5" hard drive from working.
The DG33FB is pictured here. There is a RJ45-USB-USB stack
and then two USB stacks to the right of it. Below each
stack is a green Polyfuse, but I cannot make out the letter
printed on the top. That Polyfuse should limit the current.
http://www.gtweb.net/images/MB705-T-900.jpg
HTH,
Paul
