Prev: p5q-e + ai suite
Next: Looking for a new video card
From: Don Lope de Aguirre on 23 Nov 2009 19:50
"Paul" <nospam(a)needed.com> wrote in message
news:heegie$s4e$1(a)news.eternal-september.org...
> The removal of PS/2 is part of legacy port removal. I've heard of
> at least one motherboard being legacy free, so that would mean no
> PS/2 port at all,

Removal of mouse PS2 but leaving keyboard PS2 is not a legacy free mb, it's
just a boneheaded decision by someone at Asus. My latest Asus P7P55D mb has
both mouse and keyboard PS2 so it isn't dead yet because this mb is for
i5/i7 which are latest and greatest CPUs.

From: Ross Ridge on 23 Nov 2009 22:14
Paul <nospam(a)needed.com> wrote:
>Sometimes, when a port is chopped for being "legacy", the
>actual electrical interface is still there, and it simply
>isn't hooked up to the I/O area. An example of this
>kind of silliness, is what happened with floppy. At one
>point, a SuperI/O floppy interface could control two
>floppy drives, A: and B:. Then, the legacy nonsense was
>interpreted by the manufacturers as, "hey, let's remove
>the control signals for one of the floppy drives". The
>chip itself may have still had the circuitry for
>two floppy drives (for people who transfer files from
>floppy to floppy), but if you hooked up the floppy
>cable, only one would work. ...

Yah, I don't get that. All Super I/O chips I've seen support two
floppy drives. I can understand saving a few pennies by not including
a floppy connector at all, but by only supporting a single floppy drive
all they're saving is a couple of traces on the motherboard.

Ross Ridge

--
l/ // Ross Ridge -- The Great HTMU
[oo][oo] rridge(a)csclub.uwaterloo.ca
-()-/()/ http://www.csclub.uwaterloo.ca/~rridge/
db //
From: JClark on 24 Nov 2009 04:46
On Mon, 23 Nov 2009 12:25:30 -0500, Paul <nospam(a)needed.com> wrote:

> plug in a PCI to
>USB card, and try testing how well USB works with that.
Paul,
You are a storehouse of information. I'm copying all of your messages
for storing as "textbook" files.

I'll try that test. It's cheap enough. Certainly USB reliability is
critical.
I have about two more weeks to return the MB to the vendor, so I'll
work with it for a bit longer.

Things do seem to have changed a bit. My original assessment (only IO
based USB ports not working) was in error. Now all of them are working
intermittently. For example, I just booted the system with the USB
mouse in an IO port, and it was dead after boot. I unplugged it and
replugged it in the same port, and it lit up and worked.
Strange ... at least to me.
Many thanks.

Jack
From: Paul on 24 Nov 2009 05:28
JClark wrote:
> On Mon, 23 Nov 2009 12:25:30 -0500, Paul <nospam(a)needed.com> wrote:
>
>> plug in a PCI to
>> USB card, and try testing how well USB works with that.
> Paul,
> You are a storehouse of information. I'm copying all of your messages
> for storing as "textbook" files.
>
> I'll try that test. It's cheap enough. Certainly USB reliability is
> critical.
> I have about two more weeks to return the MB to the vendor, so I'll
> work with it for a bit longer.
>
> Things do seem to have changed a bit. My original assessment (only IO
> based USB ports not working) was in error. Now all of them are working
> intermittently. For example, I just booted the system with the USB
> mouse in an IO port, and it was dead after boot. I unplugged it and
> replugged it in the same port, and it lit up and worked.
> Strange ... at least to me.
> Many thanks.
>
> Jack

Not that it matters, but do you see the correct number of
"Enhanced" entries in the USB section of Device Manager ?
That should be two. And for the other type, either four
or five of them. At least, according to the ATI docs for SB750,
which are rather generic (i.e. treated as SB7XX). I'm trying
to determine here, whether what is inside the chip is being
detected OK.

This example actually contains two devices, a Southbridge
and a NEC USB card.

http://www.usbman.com/Guides/checki4.jpg

24D2,24D4,24D7.24DE are USB 1.1 blocks (enough for eight ports on Southbridge)
Standard Enhanced PCI to USB Host Controller (USB2 for all eight ports on Southbridge)

NEC PCI to USB Open Host Controller (two of them for four PCI USB card ports)
NEC PCI to USB Enhanced Host Controller (USB2 for all four ports on PCI USB card)

In your case, I'm expecting two Enhanced entries from SB750.
And four or five others.

*******

Other than USB power, early BIOS screens had a "USB drive strength"
setting. I haven't seen such a setting, since the first generation
of USB2 devices were introduced. And in the few instances where
that was adjusted, nothing seemed to happen. So it's not like
there was a marginality evident.

Do you have a complete list of what USB devices are connected
in your computer ? Is there anything in that list which is
unusual or ancient ? Do you have any complex networks of
external hubs ?

On the one hand, in terms of the physical implementation,
there is a certain degree of sharing. For example, your
USB 1.1 logic block being shared over three ports. The
+5V power coming through a Polyfuse, being shared over two
ports. If you observed a clustering to the problems,
that would make it easier to blame something. I can't explain
random dropouts, short of something being wrong with
some aspect of the power in the system. That could even
include a marginal low voltage supply to the Southbridge
logic. But if that was the case, you'd likely see more than
USB dropping out, and some other ports would be affected,
like SATA.

You could also test with an alternate OS, but there the problem
might be, getting all the ports recognized to the same degree
as Windows manages to do. So that is unlikely to be a good
enough test, to add anything to the diagnosis.

If you want a USB item display, you can try UVCView. There is
an archived copy on here (click any of the dated links). This
reports config info as delivered by the USB devices, and indicates
whether endpoints have been set up and so on. When your mouse
goes missing, see if you can see anything here. With this tool,
you might see up to 24 lines in the display - 12 entries for the
ports running at USB 1.1 rate, 12 entries for the ports running
at USB 2.0 rate. A plugged device can only run at USB 1.1 or USB 2.0
and not both, so of the 24 lines, a max of 12 can be occupied. So
this tool is more "port oriented", unlike the Device Manager
USB entries. I don't think this tool works for things connected
through a hub, and it only reports on things directly
connected to the computer.

http://web.archive.org/*/http://download.microsoft.com/download/e/b/a/eba1050f-a31d-436b-9281-92cdfeae4b45/UVCView.x86.exe

Paul
From: JClark on 25 Nov 2009 06:21
On Tue, 24 Nov 2009 05:28:48 -0500, Paul <nospam(a)needed.com> wrote:

>correct number of
>"Enhanced" entries in the USB section of Device Manager ?
>That should be two.
Yes, there are two (view screenshot here):

https://secure.storegate.com/user/share.aspx?id=07e38411-a55e-4003-abf0-644bd93a1d27

>Do you have a complete list of what USB devices are connected
>in your computer ?
The only thing connected now is the USB mouse.
I did have connected an external hard drive, which I used to copy a
lot of files from the computer this one is going to replace, but it's
currently disconnected.

Incidentally, or parenthetically, and not to confuse the issue, but my
BIOS (latest one 1708) may be flaky. Whenever I connect the USB hard
drive, BIOS changes all my previously set boot sequences and boot
device configurations and makes the external hard drive the first boot
device, which causes boot failure. I have to re-do the BIOS settings
if I ever disconnect then reconnect the USB external hard drive.
"Curiouser and curiouser".

>try UVCView

Looks interesting. I'll work on in tonight (work calls!)

Many thanks again, and I will report back.

Jack


First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5
Prev: p5q-e + ai suite
Next: Looking for a new video card