USB 3.0 Q's

[RJK]

Hi Long story short ....ish

Windows Ultimate / Asus M5A78L_USB3 motherboard with two USB 3.0 ports at
the rear, nothing plugged into them whilst machine was on my bench, and
after running a few MS Mr. Fixit's to fix several things, e.g. IE and Aero
theme ...had to reinstall Nvidia drivers , reinstalled his internet security
suite and updated it, lots of houskeeping etc. etc. .... machine was lovely,
....robust and lovely, and I returned it to it's owner.
Who, the next day, informed that it was refusing to boot, ...so earlier
tonight I drove over there to find that an external usb 2.0 hard-disk - for
backups , and a Dell D140? usb 2.0 printer were both plugged into the two
blue usb 3.0 ports. After removing them, and plugging them into USB 2.0
ports, and correcting some bios settings, that had changed, annoyingly,
Sata ports 1-4 mode had changed back to IDE mode, so had to change it back
to AHCI again, and the boot order had changed again !! I set it to LG dvd
drive, then Sata 1 boot hd, and dvd drive had dropped to the bottom !
(cleared cmos settings while it was here - and checkd CR2032 voltage.

Anyhooo, I've just been rummaging through the M5A78L_USB3 manual, and it
says in there to not connect a kb or mouse to a usb 3.0 port whils
tinstalling windos, and only to use them for data storage only, and usb3
drivers have to be installed to use usb3 mode.

I thought that usb3 ports were backwards compatible with usb v2.0 and v1.1,
and that drivers obviously are required to operate at the higher usb3 speed,
e.g. if one had forked out for a usb3.0 external hd, and was lucky enough to
get usb3 working - (been reading horror stories on the web again !)

So it looks like having his Dell usb printer plugged into a usb3 port was
causing all the trouble. i.e. with both usb3 ports empty, machine is booting
properly every time !!

....any tips gladly appreciated, as I obviously only have rudimentary
knowledge on usb 3.0

regards, Richard

 

[Paul]

Hi Long story short ....ish

Windows Ultimate / Asus M5A78L_USB3 motherboard with two USB 3.0 ports at
the rear, nothing plugged into them whilst machine was on my bench, and
after running a few MS Mr. Fixit's to fix several things, e.g. IE and Aero
theme ...had to reinstall Nvidia drivers , reinstalled his internet security
suite and updated it, lots of houskeeping etc. etc. .... machine was lovely,
...robust and lovely, and I returned it to it's owner.
Who, the next day, informed that it was refusing to boot, ...so earlier
tonight I drove over there to find that an external usb 2.0 hard-disk - for
backups , and a Dell D140? usb 2.0 printer were both plugged into the two
blue usb 3.0 ports. After removing them, and plugging them into USB 2.0
ports, and correcting some bios settings, that had changed, annoyingly,
Sata ports 1-4 mode had changed back to IDE mode, so had to change it back
to AHCI again, and the boot order had changed again !! I set it to LG dvd
drive, then Sata 1 boot hd, and dvd drive had dropped to the bottom !
(cleared cmos settings while it was here - and checkd CR2032 voltage.

Anyhooo, I've just been rummaging through the M5A78L_USB3 manual, and it
says in there to not connect a kb or mouse to a usb 3.0 port whils
tinstalling windos, and only to use them for data storage only, and usb3
drivers have to be installed to use usb3 mode.

I thought that usb3 ports were backwards compatible with usb v2.0 and v1.1,
and that drivers obviously are required to operate at the higher usb3 speed,
e.g. if one had forked out for a usb3.0 external hd, and was lucky enough to
get usb3 working - (been reading horror stories on the web again !)

So it looks like having his Dell usb printer plugged into a usb3 port was
causing all the trouble. i.e. with both usb3 ports empty, machine is booting
properly every time !!

...any tips gladly appreciated, as I obviously only have rudimentary
knowledge on usb 3.0

regards, Richard

Was there a report on the screen of "overclocking failed" ?

On an Asus motherboard, the motherboard can reset the BIOS settings,
as part of recovering from overclocking. Even though, the user has
not been overclocking. Any event with similar symptoms to an
overclocking failure, will trigger the feature.

When this happens, it implies something is not stable, or perhaps
the user turned off the power in mid session. The BIOS may record
that as a crash event, and respond by resetting the BIOS.

On the most improved versions of such coding in BIOS code, the
BIOS only corrects the CPU clock, and leaves the other settings
alone. There is no reason to be changing the boot order, storage modes,
or whatever. The actual changes to recover from overclocking, should
be a bit more restricted than that - CPU clock, memory speed,
that sort of thing.

I have one Asus motherboard here, which did that with annoying
regularity. And it would change a couple storage settings, that
I would have to correct each time (it basically didn't have
very good default settings). Needless to say, I was pretty
fed up with the feature after a while.

A very few motherboards, include the ability to store the "profile"
to flash. Such that it can be recovered, if the settings are ever
lost. This is great, if the CMOS battery is flat. Only problem with
the feature, is when the BIOS code is flash upgraded, there is no
procedure in the BIOS to convert the old BIOS settings, into the
newer BIOS format. Requiring the user to enter the settings again.
So the "profile" idea, is of limited value. It doesn't really
memorize things, in a portable way.

The Asrock approach on this, is a bit better. In that, three
consecutive reset events must happen, before overclocking
recover is attempted. If you actually overclock it too far,
pressing the reset button several times in succession,
will trigger a return the hardware to nominal speed.
That method, didn't cause me quite as much grief.

Paul

 

[Ken1943]

Hi Long story short ....ish

Windows Ultimate / Asus M5A78L_USB3 motherboard with two USB 3.0 ports at
the rear, nothing plugged into them whilst machine was on my bench, and
after running a few MS Mr. Fixit's to fix several things, e.g. IE and Aero
theme ...had to reinstall Nvidia drivers , reinstalled his internet security
suite and updated it, lots of houskeeping etc. etc. .... machine was lovely,
...robust and lovely, and I returned it to it's owner.
Who, the next day, informed that it was refusing to boot, ...so earlier
tonight I drove over there to find that an external usb 2.0 hard-disk - for
backups , and a Dell D140? usb 2.0 printer were both plugged into the two
blue usb 3.0 ports. After removing them, and plugging them into USB 2.0
ports, and correcting some bios settings, that had changed, annoyingly,
Sata ports 1-4 mode had changed back to IDE mode, so had to change it back
to AHCI again, and the boot order had changed again !! I set it to LG dvd
drive, then Sata 1 boot hd, and dvd drive had dropped to the bottom !
(cleared cmos settings while it was here - and checkd CR2032 voltage.

Anyhooo, I've just been rummaging through the M5A78L_USB3 manual, and it
says in there to not connect a kb or mouse to a usb 3.0 port whils
tinstalling windos, and only to use them for data storage only, and usb3
drivers have to be installed to use usb3 mode.

I thought that usb3 ports were backwards compatible with usb v2.0 and v1.1,
and that drivers obviously are required to operate at the higher usb3 speed,
e.g. if one had forked out for a usb3.0 external hd, and was lucky enough to
get usb3 working - (been reading horror stories on the web again !)

So it looks like having his Dell usb printer plugged into a usb3 port was
causing all the trouble. i.e. with both usb3 ports empty, machine is booting
properly every time !!

...any tips gladly appreciated, as I obviously only have rudimentary
knowledge on usb 3.0

regards, Richard

If the bios settings changed twice, I wouldn't check the battery, I would
replace it.

Must have something to do with the usb3 chip/driver. I just got a Asus
mobo (Intel). No such warnings in the manual.


KenW

 

[Robin Bignall]

I thought that usb3 ports were backwards compatible with usb v2.0 and v1.1,
and that drivers obviously are required to operate at the higher usb3 speed,
e.g. if one had forked out for a usb3.0 external hd, and was lucky enough to
get usb3 working - (been reading horror stories on the web again !)

I thought that too, Richard, but my W7 Ultimate / Gigabyte board also
shows such weird problems if I plug anything into a USB3 port other than
an external USB3 HDD.

 

[Ken1943]

I also would check the Asus site for an updated usb3 driver and bios
update.

KenW

 

[RJK]

Was there a report on the screen of "overclocking failed" ?

On an Asus motherboard, the motherboard can reset the BIOS settings,
as part of recovering from overclocking. Even though, the user has
not been overclocking. Any event with similar symptoms to an
overclocking failure, will trigger the feature.

When this happens, it implies something is not stable, or perhaps
the user turned off the power in mid session. The BIOS may record
that as a crash event, and respond by resetting the BIOS.

On the most improved versions of such coding in BIOS code, the
BIOS only corrects the CPU clock, and leaves the other settings
alone. There is no reason to be changing the boot order, storage modes,
or whatever. The actual changes to recover from overclocking, should
be a bit more restricted than that - CPU clock, memory speed,
that sort of thing.

I have one Asus motherboard here, which did that with annoying
regularity. And it would change a couple storage settings, that
I would have to correct each time (it basically didn't have
very good default settings). Needless to say, I was pretty
fed up with the feature after a while.

A very few motherboards, include the ability to store the "profile"
to flash. Such that it can be recovered, if the settings are ever
lost. This is great, if the CMOS battery is flat. Only problem with
the feature, is when the BIOS code is flash upgraded, there is no
procedure in the BIOS to convert the old BIOS settings, into the
newer BIOS format. Requiring the user to enter the settings again.
So the "profile" idea, is of limited value. It doesn't really
memorize things, in a portable way.

The Asrock approach on this, is a bit better. In that, three
consecutive reset events must happen, before overclocking
recover is attempted. If you actually overclock it too far,
pressing the reset button several times in succession,
will trigger a return the hardware to nominal speed.
That method, didn't cause me quite as much grief.

Paul

Mnay thanks Paul,

As always you hit the nail on the head,
(the "overclocking failed" was what I used to occasionally get on my M3N78 /
Phenom II 955 with 2 x 1gb Crucial Ballistix set to 4-4-4-12-24 2.0v ,
until I gave up manually setting some known hardware values, and left
everything set to Auto, including the memory voltage !)

"something is not stable, or perhaps

the user turned off the power in mid session"

...pretty sure he did that when it stopped responding ! ...he didn't report
an "oveclocking failed" ...the thing was more hanging up than anything else.

It seems that simply having the printer usb lead connected to a usb3 port,
was upsetting several bios settings, ( now I'm wondering if usb 3.0 drivers
are installed, ....i would of course have run the motherboard cd when it
was built ....a year or more ago, ...didn't check usb 3.0 drivers as the
owner has no 3.0 ext. disks ...oh well ! ).
....and considering that the Asus manual says, in as many words, that USB 3.0
is for connecting storage devices only, he should not have had his printer
connected to a USB 3.0 port, I suppose, but, I didn't know that anyway until
I spotted it in the manual.

I did several cold starts while I was there, (with empty usb3 ports, and it
was fine every time. Here's hoping that the correct bios settings hold, if
not I'll be replacing the battery of course.

regards, Richard

 

[RJK]

If the bios settings changed twice, I wouldn't check the battery, I would
replace it.

Must have something to do with the usb3 chip/driver. I just got a Asus
mobo (Intel). No such warnings in the manual.


KenW

Hi,

At the start of the manual is states that USB3.0 is backwards compatibel
with USB 2.0, and then later on in the manual it states,
"USB 3.0 devices can only be used as data storage only"
and that they're unavailable for use under DOS envronments.

....anyhooo, I seem to have learnt the hard way, that I should have disabled
the usb3 ports in bios, in case someone plugs a printer into them !

regards, Richard

 

[Paul]

Hi,

At the start of the manual is states that USB3.0 is backwards compatibel
with USB 2.0, and then later on in the manual it states,
"USB 3.0 devices can only be used as data storage only"
and that they're unavailable for use under DOS envronments.

...anyhooo, I seem to have learnt the hard way, that I should have disabled
the usb3 ports in bios, in case someone plugs a printer into them !

regards, Richard

The USB3 chip will not have as complete a BIOS code module,
as the USB ports on the chipset. For example, if you had a
NEC or Etron chip for USB3, and other USB2 ports on the
Southbridge, the Southbridge is likely to have the best support.

There is nothing architecturally to prevent it, but it's a lot
of work for a company making a USB3 chip, to provide that
kind of code. That code is best written by AMI/Award/Phoenix etc.

There are a few chipsets, with "Native" USB3. An example would be
A75 from AMD, and there are some recent Intel chipsets with USB3
as well. But generally, native USB3 is rare, if you look at
all the chipsets that have shipped with USB3. Most will have
used a separate chip.

To support DOS, the USB3 chip support code in the BIOS, would
need the equivalent of INT 0x13 routines for USB Storage Class.

You could see something similar, back in the transition from
USB 1.1 to USB 2. The USB 2 chip back then, was separate, and
behavior differed between the Southbridge USB 1.1 ports, and
the USB 2 chip that was added. Now, USB 2 is native, and
is that way on everything. Some day, USB 3 will be like that
as well.

*******

If you look at USB 3 connector pinouts, there are two sets
of pins. The USB 3 pins are separate from the USB 2 pins.
It means the add-on chip, has aspects of USB 2 and USB 3.
The USB 3 pins must be kept separate, because of the
gigabit speeds involved. (It's possible the USB 3 pins
use low-amplitude signaling, and they might not tolerate
5V levels of legacy devices very well.)

http://usb3expresscard.com/images/usb/usb3_pinout.jpg

In principle, inside the NEC or Etron chip, the path
for the USB 2 pins and USB 3 pins is separate enough,
that they could have been run in parallel. But in practice,
the software is not designed to run them that way. There
is a negotiation phase, just as there was with USB 1.1/2.0,
and after the decision is made to operate at a particular
speed, then only one of the two sets of pins inside the blue
connector, is used for data transfers.

Paul

 

[Dave]

I also would check the Asus site for an updated usb3 driver and bios
update.

KenW

Sometimes I'll read a post and just don't believe what I've read.
Specifically for this post, I moved my usb printer cable from it's normal
position to the usb 3.0 port. Just as I would have expected, it works fine.
Now, maybe someone can tell me why Dell and others put two usb 2.0 ports
on the front of the machine but hide the usb 3.0 ports on the back, thus
making one reluctant to use the things (even if I had something that would
benefit from such use), but I suppose only Dell can answer that one.

 

[Paul]

Sometimes I'll read a post and just don't believe what I've read.
Specifically for this post, I moved my usb printer cable from it's normal
position to the usb 3.0 port. Just as I would have expected, it works fine.
Now, maybe someone can tell me why Dell and others put two usb 2.0 ports
on the front of the machine but hide the usb 3.0 ports on the back, thus
making one reluctant to use the things (even if I had something that would
benefit from such use), but I suppose only Dell can answer that one.

USB3 runs at 5 Gbit/sec.

The connector on the back of the computer, mounts directly
on the motherboard itself. It means there can be due attention
to controlled impedance and electrical reflections. At that
speed, you can't even afford sharp corners on the copper tracks.
Also, static discharge is handled properly back there. The
"USB stack" presses against tabs on the I/O plate, to provide
a route to chassis ground. Something that is typically missing
on the plastic front of the computer.

They do make cabling for USB3, for front panel setups.

http://www.wtotoy.com/images/watermarked_images/detailed/14/_DSC8800.jpg

Notice there is no bare wire on that thing. It's likely shielded,
right inside the molded connector housings. Dressing of the wires
is important, to keep the electrical impedance constant.

So it can be done. You'd need the appropriate 2xN connector on the
motherboard, as a place to plug the cable. If the computer only has
two USB3 ports, they would probably not want to split them, and
put just one on the front and one on the back.

They also make USB3 hubs.

http://img.diytrade.com/cdimg/13234.../USB3_0_4-Port_3_5_Front_Panel_USB3_0_HUB.jpg

You use an ordinary USB3 cable, run it from the single back connector on that
thing, out through a card slot hole, then plug into the back USB3. That
one gives 1:4 fanout. I have no idea how well those hubs work, as there
haven't been any reports of them that I've read.

Startech makes one, but the single review of it, says it broke
after 7 or 8 uses.

http://images10.newegg.com/NeweggImage/productimage/11-996-039-03.jpg

http://images10.newegg.com/NeweggImage/productimage/11-996-039-05.jpg

http://www.newegg.com/Product/Product.aspx?Item=N82E16811996039

When they put enough chips to have four USB3 ports, then there
is room to put two on the front and two on the back.

Paul

 

[R. C. White]

Hi, Paul.

FYI: My MSI 990FXA-GD80 motherboard, new in December 2012, has 2 USB 3.0
ports on the backplane. It also has a connector for two more USB 3.0 ports,
and comes with a cable and bracket to mount the two jacks onto the back of
the computer, where a PCI card would usually be. (I've not seen this 20-pin
mobo connector (NEC D720200) before; I bent some of the pins trying to plug
the cable into the jack and had to replace the mobo.)

This gives me 4 USB 3.0 ports (in addition to several USB 2.0 ports), all at
the back of the case. The 2-port cable could be mounted on the front of the
case - but my old case has no provision for such.

So far, the only USB 3.0 devices I have are a 3 TB Seagate GoFlex Desk
external HDD and a couple of 8 GB Lexar USB 3.0 flash drives. They all work
fine. And my USB 2.0 devices work, too, at the slower speed when I test
them in the 3.0 ports. The USB 3.0 speed is much better than 2.0 - but SATA
3 (6 Gb/s) is still faster.

RC
--
R. C. White, CPA
San Marcos, TX
(e-mail address removed)
Microsoft Windows MVP (2002-2010)
Windows Live Mail 2012 (Build 16.4.3508.0205) in Win8 Pro


"Paul" wrote in message

Sometimes I'll read a post and just don't believe what I've read.
Specifically for this post, I moved my usb printer cable from it's normal
position to the usb 3.0 port. Just as I would have expected, it works
fine.
Now, maybe someone can tell me why Dell and others put two usb 2.0 ports
on the front of the machine but hide the usb 3.0 ports on the back, thus
making one reluctant to use the things (even if I had something that would
benefit from such use), but I suppose only Dell can answer that one.

USB3 runs at 5 Gbit/sec.

The connector on the back of the computer, mounts directly
on the motherboard itself. It means there can be due attention
to controlled impedance and electrical reflections. At that
speed, you can't even afford sharp corners on the copper tracks.
Also, static discharge is handled properly back there. The
"USB stack" presses against tabs on the I/O plate, to provide
a route to chassis ground. Something that is typically missing
on the plastic front of the computer.

They do make cabling for USB3, for front panel setups.

http://www.wtotoy.com/images/watermarked_images/detailed/14/_DSC8800.jpg

Notice there is no bare wire on that thing. It's likely shielded,
right inside the molded connector housings. Dressing of the wires
is important, to keep the electrical impedance constant.

So it can be done. You'd need the appropriate 2xN connector on the
motherboard, as a place to plug the cable. If the computer only has
two USB3 ports, they would probably not want to split them, and
put just one on the front and one on the back.

They also make USB3 hubs.

http://img.diytrade.com/cdimg/13234.../USB3_0_4-Port_3_5_Front_Panel_USB3_0_HUB.jpg

You use an ordinary USB3 cable, run it from the single back connector on
that
thing, out through a card slot hole, then plug into the back USB3. That
one gives 1:4 fanout. I have no idea how well those hubs work, as there
haven't been any reports of them that I've read.

Startech makes one, but the single review of it, says it broke
after 7 or 8 uses.

http://images10.newegg.com/NeweggImage/productimage/11-996-039-03.jpg

http://images10.newegg.com/NeweggImage/productimage/11-996-039-05.jpg

http://www.newegg.com/Product/Product.aspx?Item=N82E16811996039

When they put enough chips to have four USB3 ports, then there
is room to put two on the front and two on the back.

Paul

 

[RJK]

So far, the only USB 3.0 devices I have are a 3 TB Seagate GoFlex Desk
external HDD and a couple of 8 GB Lexar USB 3.0 flash drives. They all
work fine. And my USB 2.0 devices work, too, at the slower speed when I
test them in the 3.0 ports. The USB 3.0 speed is much better than 2.0 -
but SATA 3 (6 Gb/s) is still faster.

....but, what happens if you plug a usb printer signals lead into one of them
?

I'm still pondering on the statements in the Asus M5A78L_USB3 manual :-

"USB 3.0 ports 3 and 4 . These two 9-pin Univeral Serial Bus (USB) ports
connect to USB 3.0/2.0 devices."
"DO NOT connect a keyboard / mouse to any to any USB 3.0 port when
installing Windows operating system."
"USB 3.0 devices can only be used as data storage only."

I'm sure there's a contradiction in there somewhere, and I'm still studying
Pauls posts on this thread that details differences in design, esp.
differences in manufacturers implementations of the two types of ports, USB
2.0/1.1 and 3.0

Until spotting in Asus M5A78L_USB3 manual that USB 3.0 ports are for data
sotrage devices only, and only after encountering a friends PC that went
"tits up" when a USB printer lead was connected into a USB 3.0 port, I would
not have guessed that plugging a USB printer lead into a USB 3.0 port was a
"no go." ....also, for some strange reason, I would instinctively never
have attempted to connect a usb printer to a usb 3.0 port, e.g. had there
been any present on my motherboards back panel !!! ...or maybe I would
have, and had to problem solve that the same hard way !

regards, Richard

 

[Gene E. Bloch]

...but, what happens if you plug a usb printer signals lead into one of them
?

I'm still pondering on the statements in the Asus M5A78L_USB3 manual :-

"USB 3.0 ports 3 and 4 . These two 9-pin Univeral Serial Bus (USB) ports
connect to USB 3.0/2.0 devices."
"DO NOT connect a keyboard / mouse to any to any USB 3.0 port when
installing Windows operating system."
"USB 3.0 devices can only be used as data storage only."

I'm sure there's a contradiction in there somewhere, and I'm still studying
Pauls posts on this thread that details differences in design, esp.
differences in manufacturers implementations of the two types of ports, USB
2.0/1.1 and 3.0

Until spotting in Asus M5A78L_USB3 manual that USB 3.0 ports are for data
sotrage devices only, and only after encountering a friends PC that went
"tits up" when a USB printer lead was connected into a USB 3.0 port, I would
not have guessed that plugging a USB printer lead into a USB 3.0 port was a
"no go." ....also, for some strange reason, I would instinctively never
have attempted to connect a usb printer to a usb 3.0 port, e.g. had there
been any present on my motherboards back panel !!! ...or maybe I would
have, and had to problem solve that the same hard way !

regards, Richard

I can't speak for your motherboard, since I have an Asus P8 H67-M Evo,
but so far (fingers crossed) I haven't had a problem plugging a USB 2
device into a USB 3 port.

I don't recall if I've ever done the opposite, so I won't remark on
that.

Never mind, I'll try a USB 3 drive in a USB 2 port now...

With my unerring instinct for overkill, I tried it with both a USB 2
cable and a USB 3 cable.

OK, no problem. I created, edited, saved, and reread a text file, then
deleted it. Then I did the same with the other cable.

Granted, this test only shows that it can work on some systems, and of
course it isn't a very thorough test...

 

[Jeff Barnett]

RJK wrote, On 5/17/2013 2:51 PM:

"USB 3.0 ports 3 and 4 . These two 9-pin Univeral Serial Bus (USB) ports
connect to USB 3.0/2.0 devices."
"DO NOT connect a keyboard / mouse to any to any USB 3.0 port when
installing Windows operating system."
"USB 3.0 devices can only be used as data storage only."

I'm sure there's a contradiction in there somewhere, and I'm still studying
Pauls posts on this thread that details differences in design, esp.
differences in manufacturers implementations of the two types of ports, USB
2.0/1.1 and 3.0

I'd assume that the manual is literal: don't connect a KB or mouse WHEN
INSTALLING WINDOWS OPERATING SYSTEM. It might be that sufficient driver
support isn't available until after the OS is on its own two feet.

Jeff Barnett

 

[Paul]

RJK wrote, On 5/17/2013 2:51 PM:


I'd assume that the manual is literal: don't connect a KB or mouse WHEN
INSTALLING WINDOWS OPERATING SYSTEM. It might be that sufficient driver
support isn't available until after the OS is on its own two feet.

Jeff Barnett

That sounds reasonable.

The thing is, a lot of the usage scenarios described, would be using
the USB2 side of the USB3 chip, and likely all the standard USB "classes"
are supported on that. You really should be able to connect a printer.

Now, are there any printers using the USB3 wires ? Probably not. I
would not expect them to waste the time doing that. USB2 would be fast
enough (compared to the page print rate). So while you plug the USB2
printer, into the USB3 hole, only the USB2 pins touch.

You are allowed to connect other things to USB3. A company made a video
capture box, that relies on the USB3 rate being available. And I haven't
seen any complaints about it.

Paul

 

[Jeff Barnett]

Paul wrote, On 5/17/2013 5:26 PM:

That sounds reasonable.

The thing is, a lot of the usage scenarios described, would be using
the USB2 side of the USB3 chip, and likely all the standard USB "classes"
are supported on that. You really should be able to connect a printer.

Now, are there any printers using the USB3 wires ? Probably not. I
would not expect them to waste the time doing that. USB2 would be fast
enough (compared to the page print rate). So while you plug the USB2
printer, into the USB3 hole, only the USB2 pins touch.

You are allowed to connect other things to USB3. A company made a video
capture box, that relies on the USB3 rate being available. And I haven't
seen any complaints about it.

Paul

I find this discussion interesting since I'm soon to embark on an
adventure - generate two fairly powerful Win 7 64bit machines to move
from XP 32bit. I'm particularly interested in USB 3.0 capability since
it could, if manufacturers nail the specs, be a major improvement in my
life, e.g., backups, loading raw files from cameras, etc. I've even been
thinking of buying a USB 3.0 via PCI-E card and an external disk USB 3.0
disk for an XP machine to see if this technology is as useful as the ads
say.

Jeff Barnett

 

[Paul]

Paul wrote, On 5/17/2013 5:26 PM:

I find this discussion interesting since I'm soon to embark on an
adventure - generate two fairly powerful Win 7 64bit machines to move
from XP 32bit. I'm particularly interested in USB 3.0 capability since
it could, if manufacturers nail the specs, be a major improvement in my
life, e.g., backups, loading raw files from cameras, etc. I've even been
thinking of buying a USB 3.0 via PCI-E card and an external disk USB 3.0
disk for an XP machine to see if this technology is as useful as the ads
say.

Jeff Barnett

The people here say they had a lot of fun testing USB3. It's too bad Intel
was so late to the party, and this article was done before Intel delivered
theirs.

http://vr-zone.com/articles/usb-3.0-speed-tests--7-way-host-controllers-roundup/13358-11.html

Etron made a name for themselves a couple years ago.

http://forums.tweaktown.com/asrock/44094-usb-3-0-problem-3.html

Paul

 

[Jeff Barnett]

Paul wrote, On 5/18/2013 3:03 AM:

The people here say they had a lot of fun testing USB3. It's too bad Intel
was so late to the party, and this article was done before Intel delivered
theirs.

http://vr-zone.com/articles/usb-3.0-speed-tests--7-way-host-controllers-roundup/13358-11.html


Etron made a name for themselves a couple years ago.

http://forums.tweaktown.com/asrock/44094-usb-3-0-problem-3.html

Paul

Thank you for the pointers. Doesn't sound promising. The motherboard I
was thinking of using for my builds was the ASROCK Extreme6 with a 2011
socket CPU. (The second article seemed relevant.) I just scanned the
first article (need to go to the store before spending enough time to
read the whole thing) and it also doesn't look promising. Do you know if
the Intel chip sets did any better than those tested?

Jeff

 

[Paul]

Thank you for the pointers. Doesn't sound promising. The motherboard I
was thinking of using for my builds was the ASROCK Extreme6 with a 2011
socket CPU. (The second article seemed relevant.) I just scanned the
first article (need to go to the store before spending enough time to
read the whole thing) and it also doesn't look promising. Do you know if
the Intel chip sets did any better than those tested?

Jeff

I found an article here, but they didn't really put that much effort into it.

This (advertising heavy) web page, compares Z77 to A75 (Intel versus AMD).
But the test cases probably aren't actually evaluating the absolute
peak of performance (you need USB3 to SATA plus an SSD for that). Still,
the Intel does better than the others in the test.

http://xtreview.com/addcomment-id-21739-view-Intel-Z77-vs-AMD-A75-USB-3.0-performance.html

It's actually pretty difficult to benchmark USB3, with storage devices.
Not that long ago, the best USB3 to SATA could do around 200MB/sec,
while a theoretical calc said somewhere around 336MB/sec should be
possible.

One other issue with PCI Express bus lanes, is packet size. PCI Express
uses packets. Chipset makers, don't use very large packets for their
designs. This reduces PCI Express efficiency (ratio of header to payload
size). On my motherboard, with Intel chipset, this roughly cuts the bandwidth
in half on a video slot.

http://www.plxtech.com/files/pdf/technical/expresslane/Choosing_PCIe_Packet_Payload_Size.pdf

And I'm not aware of anyone testing for that issue.

*******

As for controller cards, you can find a lot of cards with x1 slot, and
those may be limited with the older 250MB/sec x1 lanes. Not all
motherboards have x1 slots with the 500MB/sec Rev.2 lane.

This one is an interesting design. It's not for sale any more.
It has a PCI Express x4 edge connector. A PEX switch chip, converts
the x4 (Rev 1.1 or Rev 2) bandwidth, into something the Asmedia chips
can use (probably x1 Rev 2). Since there is a chip per port, there
should not be a bottleneck. This is similar to something Asus did, when
they did the U3S6 card for $25 a piece. But I bet Highpoint wanted
a lot more than that for it. These are the kinds of products, you
cannot hope to keep in the market. They're one-offs.

http://www.newegg.com/Product/Product.aspx?Item=N82E16816115104

It's like, when another company made a bunch of interesting
bridged products - a PCI Express x1 conversion chip plus a PCI
chip for the function. Since the cards were always a bit more
expensive than a plain PCI version, the cards could not compete
on price, and the company discontinued the cards. As far as I
was concerned, such cards might have had a niche market even today,
but for the companies making this stuff, they discourage easily.
So anything which is the slightest bit exotic, gets dumped
pretty quickly.

Paul

 

[Yousuf Khan]

...and considering that the Asus manual says, in as many words, that USB 3.0
is for connecting storage devices only, he should not have had his printer
connected to a USB 3.0 port, I suppose, but, I didn't know that anyway until
I spotted it in the manual.

I've even had motherboards refuse to boot and other weird things when a
simple hub was installed into a bog-standard USB2.0 port. What it turned
out to be was that the USB hub was a powered hub, and once we removed
the external power, it worked fine. For some reason the power was
feeding back towards the motherboard's USB port, creating all kinds of
havoc. I try to avoid powered USB hubs these days, and avoid the hubs in
general. But USB has always been flaky.

Yousuf Khan