Generic USB 2.0 hub showing up as USB 1.1

I have two generic 4-port USB hubs that are supposed to be USB 2.0
compliant. The older one is fine, works as advertised. Meanwhile the
newer one sometimes shows up under the USB 2.0 root hub (i.e. "Standard
Enhanced PCI to USB Host Controller"), or usually it shows up under the
slower "Standard OpenHCD USB Host Controller" (USB 1.1).

My motherboard (Asus M2NPV-VM) USB ports are all USB 2.0 compliant, and
they individually auto-detect whether they need to switch down to USB
1.1 speeds. I'm using a Microsoft tool called UVCView under Windows XP
to display the details of the USB devices, including the hubs and roothubs.

According to UVCView, the older hub and newer hub seem to have the same
chipset vendor (idVendor = "Genesys Logic, Inc."), although externally
they look quite different and have different brand names. So I'm not
sure why one would be consistently USB 2.0 compliant, while the other
one is not.

Now another interesting thing I noticed is that the inconsistent hub
will only show up as USB 2.0-compliant after I boot into Windows after
having previously rebooted from Ubuntu 7.10 Linux (dual-boot system). So
maybe Linux does something to the device that puts it right. But when I
list the devices while in Linux I see that it is listed under USB 1.1
just like when in Windows. I have no idea why Linux leaves the hub fixed
for Windows, but doesn't fix it for itself?!

Any idea what's going on with this hub?

Yousuf Khan

On Mon, 31 Dec 2007 02:34:02 GMT, "nobody@nowhere.net"
<mygarbage2000@hotmail.com> wrote:

>Taiwan is OK, it's the mainland that produces cheap crap. Yet what
>else could be expected? Nobody looks at China as a producer of
>quality goods - that place is reserved by EU, Japan, and USA, with
>countries like S.Korea and Taiwan trailing pretty close.

Once upon a time, if it was "Made in Taiwan" it was known as crap too


I don't think it's really fair to say that just because the hub is
made in China, it must be of poor quality. As you noted yourself, even
a $30 Belkin hub is likely to come from a China factory. The key
difference I find is whether the company that holds the brand, are
they willing to pay for better quality and more stringent controls.

I met a non-PRC owner of a factory in China via my partner once. He
mentioned to us very frankly that he has no choice but to cut corners
in order to stay competitive with less scrupulous factories. But
usually it's also made clear to the customers if they want cheap, what
they should expect and if they want better quality stuff, it won't be
dirt cheap. I'm not sure if all factories make that point to their
customers but ultimately, the fact remains you do get what you pay
for.

--
A Lost Angel, fallen from heaven
Lost in dreams, Lost in aspirations,
Lost to the world, Lost to myself

On Sat, 29 Dec 2007 16:39:50 -0500, Yousuf Khan wrote:

> I have two generic 4-port USB hubs that are supposed to be USB 2.0
> compliant. The older one is fine, works as advertised. Meanwhile the
> newer one sometimes shows up under the USB 2.0 root hub (i.e. "Standard
> Enhanced PCI to USB Host Controller"), or usually it shows up under the
> slower "Standard OpenHCD USB Host Controller" (USB 1.1).
>
> My motherboard (Asus M2NPV-VM) USB ports are all USB 2.0 compliant, and
> they individually auto-detect whether they need to switch down to USB
> 1.1 speeds. I'm using a Microsoft tool called UVCView under Windows XP
> to display the details of the USB devices, including the hubs and roothubs.
>
> According to UVCView, the older hub and newer hub seem to have the same
> chipset vendor (idVendor = "Genesys Logic, Inc."), although externally
> they look quite different and have different brand names. So I'm not
> sure why one would be consistently USB 2.0 compliant, while the other
> one is not.
>
> Now another interesting thing I noticed is that the inconsistent hub
> will only show up as USB 2.0-compliant after I boot into Windows after
> having previously rebooted from Ubuntu 7.10 Linux (dual-boot system). So
> maybe Linux does something to the device that puts it right. But when I
> list the devices while in Linux I see that it is listed under USB 1.1
> just like when in Windows. I have no idea why Linux leaves the hub fixed
> for Windows, but doesn't fix it for itself?!
>
> Any idea what's going on with this hub?

I can almost guaranty that an operating system can't leave any USB hub
(fixed).

Can you compare the data transfer rate of the hub in question while in XP
to that in linux? Take a large file say 100 megs and transfer it from an
external drive and time it. I'd like to see which is faster, XP or linux
or if it;s the same.

On Sat, 29 Dec 2007 16:39:50 -0500, Yousuf Khan <bbbl67@yahoo.com>
wrote:

>I have two generic 4-port USB hubs that are supposed to be USB 2.0
>compliant. The older one is fine, works as advertised. Meanwhile the
>newer one sometimes shows up under the USB 2.0 root hub (i.e. "Standard
>Enhanced PCI to USB Host Controller"), or usually it shows up under the
>slower "Standard OpenHCD USB Host Controller" (USB 1.1).
>
>My motherboard (Asus M2NPV-VM) USB ports are all USB 2.0 compliant, and
>they individually auto-detect whether they need to switch down to USB
>1.1 speeds. I'm using a Microsoft tool called UVCView under Windows XP
>to display the details of the USB devices, including the hubs and roothubs.
>
>According to UVCView, the older hub and newer hub seem to have the same
>chipset vendor (idVendor = "Genesys Logic, Inc."), although externally
>they look quite different and have different brand names. So I'm not
>sure why one would be consistently USB 2.0 compliant, while the other
>one is not.
>
>Now another interesting thing I noticed is that the inconsistent hub
>will only show up as USB 2.0-compliant after I boot into Windows after
>having previously rebooted from Ubuntu 7.10 Linux (dual-boot system). So
>maybe Linux does something to the device that puts it right. But when I
>list the devices while in Linux I see that it is listed under USB 1.1
>just like when in Windows. I have no idea why Linux leaves the hub fixed
>for Windows, but doesn't fix it for itself?!
>
>Any idea what's going on with this hub?
>
> Yousuf Khan

Crappy cable? I've seen a few times a 2.0 device gets downgraded to
1.x when connected with a substandard cable. A good quality cable
(try Belkin) solved it for me.
Oh, one more thought - check where the hub in question was made. If
it's China, that explains it. If they deliberately use led paint for
children's toys (saves a fraction of a penny per toy vs. non-toxic
one), and conveniently forget to put cord into tires (saves both
material and labor - a few bucks total per tire - who cares if people
die when it blows out), you can expect similar "quality" materials and
workmanship from the hub.

NNN

"Yousuf Khan" <bbbl67@yahoo.com> wrote in message
news:ipSdnROWzJM6I-vaRVn_vwA@giganews.com...
>I have two generic 4-port USB hubs that are supposed to be USB 2.0
>compliant. The older one is fine, works as advertised. Meanwhile the
>newer one sometimes shows up under the USB 2.0 root hub (i.e.
>"Standard Enhanced PCI to USB Host Controller"), or usually it shows
>up under the slower "Standard OpenHCD USB Host Controller" (USB 1.1).
>
> My motherboard (Asus M2NPV-VM) USB ports are all USB 2.0 compliant,
> and they individually auto-detect whether they need to switch down
> to USB 1.1 speeds. I'm using a Microsoft tool called UVCView under
> Windows XP to display the details of the USB devices, including the
> hubs and roothubs.
>
> According to UVCView, the older hub and newer hub seem to have the
> same chipset vendor (idVendor = "Genesys Logic, Inc."), although
> externally they look quite different and have different brand names.
> So I'm not sure why one would be consistently USB 2.0 compliant,
> while the other one is not.
>
> Now another interesting thing I noticed is that the inconsistent hub
> will only show up as USB 2.0-compliant after I boot into Windows
> after having previously rebooted from Ubuntu 7.10 Linux (dual-boot
> system). So maybe Linux does something to the device that puts it
> right. But when I list the devices while in Linux I see that it is
> listed under USB 1.1 just like when in Windows. I have no idea why
> Linux leaves the hub fixed for Windows, but doesn't fix it for
> itself?!
>
> Any idea what's going on with this hub?
>
> Yousuf Khan


Have you visited the manufacturer's web site to get the specs on those
USB hubs? Having the same hardware (i.e., chipset) says nothing about
how the manufacturer utilized that hardware. Lots of analog modems
use the same Conexant chip but the *cards* don't have the same feature
set. Same subsystem components do not enforce the same system
features. Could be one of the hubs really only is 1.1 compliant and
that using it as a 2.0 device is not recommended. Of course, the
device could be just a crappy low-grade cheap unit that doesn't
properly respond to report itself correctly, or you need a better USB
cable.

Are both of these a self-powered hub (i.e., they have a power adapter)
or a low-power hub that relies on the current supplied from the USB
controller at the motherboard? If they are bus-powered hubs, they
CANNOT be on the same USB controller. Each controller provides 2
ports (channels) but these 2 ports still share the same controller and
so both are limited by a total amperage that can be supplied by the
same controller. That's why you see USB ports in pairs but you have
to watch how much current is drained by them together. Bus-powered
hubs or any other bus-powered devices will tax the low current
available from the USB controller, so instead make sure to use
self-powered USB hubs, especially considering that you are planning to
connect more than just 2 USB devices to the same controller (and
possibly not just low-powered USB devices). Initially a USB device is
allowed to draw 100 mA but that device may request more power for
upstream devices in increments of 2 mA but up to a maximum of 500 mA
(and that is across the pair of ports to the same USB controller).
For a bus-powered hub, the connected devices may only use a total of
400 mA (100 mA per port) so the hub is limited to 4 ports. If using
bus-powered hubs, make sure you are using low-power USB devices (or
the high-powered USB device provides its own power supply that is
connected to a bus-powered hub). USB devices rated for bus-power draw
can be used on a bus-powered hub (but watch the total draw across both
USB ports to the same controller). The number of bus-powered or
high-power devices that you connect to a self-powered hub depends on
how much current that hub can deliver.

USB devices are supposed to report their power consumption. Maybe you
hubs don't. Or maybe they report too high a consumption to guarantee
USB 2.0 mode to work so the controller degrades to USB 1.1 mode.

Do you actually have any high-speed USB 2.0 devices connected to the
hubs when you boot the OS with the self-powered hubs already powered
up? Are they really high-speed USB 2.0 devices (USB 2.0 compliant
devices can report as low, full, or high-speed)?

For best setup, use self-powered hubs, or connect them to different
USB controllers (i.e., they don't share the same port pair coming from
the same USB controller).

On Sun, 30 Dec 2007 07:42:58 +0000, nobody@nowhere.net wrote:

> Oh, one more thought - check where the hub in question was made. If
> it's China, that explains it.

Do you know of a USB hub that isn't made in China?

Meat Plow wrote:
> I can almost guaranty that an operating system can't leave any USB hub
> (fixed).
>
> Can you compare the data transfer rate of the hub in question while in XP
> to that in linux? Take a large file say 100 megs and transfer it from an
> external drive and time it. I'd like to see which is faster, XP or linux
> or if it;s the same.


Well, that's not going to be likely to do. Since the newer hub is so
unreliable, I'm only trusting it with light duty at the moment, such as
mouse and keyboard connections, nothing data-heavy like external HDs, or
thumbdrives.

Yousuf Khan

nobody@nowhere.net wrote:
> Crappy cable? I've seen a few times a 2.0 device gets downgraded to
> 1.x when connected with a substandard cable. A good quality cable
> (try Belkin) solved it for me.

Worth a shot, I have tried moving the hub from one port to another, but
it made no difference.

> Oh, one more thought - check where the hub in question was made. If
> it's China, that explains it. If they deliberately use led paint for
> children's toys (saves a fraction of a penny per toy vs. non-toxic
> one), and conveniently forget to put cord into tires (saves both
> material and labor - a few bucks total per tire - who cares if people
> die when it blows out), you can expect similar "quality" materials and
> workmanship from the hub.

Isn't everything made in China these days? Avoiding Chinese made hubs
might be like trying to avoid any Swiss chocolate made in Switzerland.

Yousuf Khan

VanguardLH wrote:
> Have you visited the manufacturer's web site to get the specs on those
> USB hubs? Having the same hardware (i.e., chipset) says nothing about
> how the manufacturer utilized that hardware. Lots of analog modems use
> the same Conexant chip but the *cards* don't have the same feature set.
> Same subsystem components do not enforce the same system features.
> Could be one of the hubs really only is 1.1 compliant and that using it
> as a 2.0 device is not recommended. Of course, the device could be just
> a crappy low-grade cheap unit that doesn't properly respond to report
> itself correctly, or you need a better USB cable.

Well, as I said previously, these are "generic" hubs, very generic. I
doubt any of us have heard of the manufacturers' names: there isn't much
point in checking their websites, they probably sell tons of little
products. One is from Vantec and the other is Acrox. The Acrox is the
older more reliable one. Both of them are advertised as USB 2.0 hubs,
and both of them are identified as "USB2.0 Hub" internally, polled from
the USB configuration itself.

> Are both of these a self-powered hub (i.e., they have a power adapter)
> or a low-power hub that relies on the current supplied from the USB
> controller at the motherboard? If they are bus-powered hubs, they
> CANNOT be on the same USB controller. Each controller provides 2 ports
> (channels) but these 2 ports still share the same controller and so both
> are limited by a total amperage that can be supplied by the same
> controller.

Both can be self-powered or bus-powered, they have the power inputs.
Only one of them came with an included power cord though. And
surprisingly it's the less reliable one that has the power cord. The
more reliable one doesn't have one. I have tried that one with and
without the power cord, but it made no difference.

I have resorted to putting my fast peripherals on the older hub, such an
external hard drive, a digital camera, and a Skype phone. They all have
their own power cords so they don't need to be powered by the hub
anyways. The hard disk and camera show up under the mass storage device
class.

The slower hub is being used for slow peripherals like mice and
keyboards now.

> USB devices are supposed to report their power consumption. Maybe you
> hubs don't. Or maybe they report too high a consumption to guarantee
> USB 2.0 mode to work so the controller degrades to USB 1.1 mode.

Both are reporting 100mA.

Yousuf Khan

In comp.sys.ibm.pc.hardware.chips Yousuf Khan <bbbl67@yahoo.com> wrote:
> Isn't everything made in China these days? Avoiding Chinese made hubs
> might be like trying to avoid any Swiss chocolate made in Switzerland.

Depends on if by China, you mean both Taiwan and the PRC or just the PRC.
For that matter, since "made in" usually means final assembly, there are
probably some from Thailand and Malaysia too.

--
Nate Edel http://www.cubiclehermit.com/
preferred email |
is "nate" at the | "This is not a funny signature... or is it?"
posting domain |