David,
 
Just so you know that it can be done, I used a 7 port USB hub with TT's on each port and filled it with 7 radio channels running simultaneously for a commercial trunking radio system and it works just fine.
The server was a 2.66 GHz Pentium 4 on an ASUS motherboard.
 
73,
 
Steve Henke, W9SH
 
On Fri, Jul 25, 2008 at 9:19 PM, David McGough <kb4fxc@inttek.net> wrote:

Hi Everyone,

Well, I just tried one of the usbgear hubs listed below. It ..almost..
works. The 2 URI devices I'm experimenting with are identified and
functional EXCEPT that it sounds like the audio sampling rate is
wrong--way too low with aliasing distortion (or something). ugh. Probably
a driver confusion issue??

Anyway, the USB hub and hardware otherwise seems to be fine.
Here is an output from the usbtree utility:

[root@kb4fxc-rpt ~]# ./usbtree
/: Bus 02.Port 1: Dev 1, Class=root_hub, Drv=ohci_hcd/8p, 12M
/: Bus 01.Port 1: Dev 1, Class=root_hub, Drv=ehci_hcd/8p, 480M
   |_ Port 4: Dev 2, If 0, Prod=USB2.0 Hub, Class=hub, Drv=hub/4p, 480M
       |_ Port 1: Dev 3, If 0, Prod=C-Media USB Headphone Set, Class=audio, Drv=snd-usb-audio, 12M
       |_ Port 1: Dev 3, If 1, Prod=, Class=audio, Drv=snd-usb-audio, 12M
       |_ Port 1: Dev 3, If 2, Prod=, Class=audio, Drv=snd-usb-audio, 12M
       |_ Port 1: Dev 3, If 3, Prod=, Class=HID, Drv=usbfs, 12M
       |_ Port 2: Dev 4, If 0, Prod=C-Media USB Headphone Set, Class=audio, Drv=snd-usb-audio, 12M
       |_ Port 2: Dev 4, If 1, Prod=, Class=audio, Drv=snd-usb-audio, 12M
       |_ Port 2: Dev 4, If 2, Prod=, Class=audio, Drv=snd-usb-audio, 12M
       |_ Port 2: Dev 4, If 3, Prod=, Class=HID, Drv=usbhid, 12M
       |_ Port 4: Dev 5, If 0, Prod=USB2.0 Hub, Class=hub, Drv=hub/4p, 480M


I'll investigate further since this is ultimately a config I may want to
use as well.

David KB4FXC


On Tue, 22 Jul 2008, David McGough wrote:

>
> So, I guess the next question is: Does anyone have a favorite multi-tt
> hub? ....I have used this one in the past for other linux applications and
> it seemed okay...I'll try one with the URI's....I guess the biggest issue
> here is that the hub you buy today won't necessarily be the one on the
> shelves tomorrow....A bit frustrating!
>
> http://www.usbgear.com/USBG-217C.html
>
> -David
>
>
>
> On Tue, 22 Jul 2008, David Kramer wrote:
>
> > I have successfully tested URIs with several high speed USB hubs.  But these
> > tests were performed without any other USB devices connected at the same
> > time so this may not be very usefull.
> >
> > Skip points out the difference between full speed (12Mbps) and high speed
> > (480Mbs) devices.  This terminology is confusing.  Not all USB 2.0 compliant
> > hubs support high speed, so be sure the package says HIGH SPEED.
> >
> > Very few hubs support multiple transaction translators, probably because the
> > chip sets are more expensive and almost nobody understands the need.
> > Theoretically, one TT is all that is needed provided you don't mix USB 1.x
> > and 2.0 devices on the same hub.  Keep in mind that USB supports several
> > data modes each with varying priorities and latency allocations.  Plus USB
> > physical layer is half-duplex so one-way communication never comes close to
> > 480Mbps.  The TT tries to handle all the traffic so don't expect good
> > results if you plug a bunch of web cams or other bandwidth hungry devices
> > into your hub.
> >
> > Hope this help.
> >
> > David Kramer
> > DMK Engineering Inc.
> > 310-544-1222
> >
> >
> > -----Original Message-----
> > From: app_rpt-bounces@lists.illiana.net
> > [mailto:app_rpt-bounces@lists.illiana.net] On Behalf Of Skip WB6YMH
> > Sent: Tuesday, July 22, 2008 1:22 PM
> > To: app_rpt@lists.illiana.net
> > Subject: [App_rpt] USB hubs & USB audio
> >
> > > Jim Dixon and I have had disapointing results with USB hubs during the
> > > testing of USB fobs with them. We don't know why they don't work
> > > properly and we don't have the time to figure it out right now. There
> > > may be hubs which work just fine, but we just don't know as the root
> > > cause of the problem has not been uncovered.  If Steve Henke W9SH
> > > knows something I don't them maybe he should speak up on the list.
> > > Steve  WA6ZFT
> >
> > The problem is probably complex.  USB 2.0 hubs come in two flavors single-
> > transaction translators (single-TT) and multi-transaction translators
> > (multi-TT).  Most USB audio devices are "full speed" devices (12
> > mbits/sec), not "high speed" devices (480 mbits/sec) so the USB hub must
> > translate the data rate between the host and the device.  The single-TT
> > devices only have a single translator which is shared between all ports.
> > A multi-TT hub has a translator per port.  So in theory a milti-TT hub is
> > what you should want for optimal bus usage.
> >
> > An additional complication is that the Linux kernel must be multi-TT hub
> > aware to properly schedule things.  I'm not current with Linux
> > development, but I've read discussions in the past about improving the
> > kernel's USB scheduler in this area.
> >
> > Here are a couple of links to info which may be useful:
> >
> > http://www.commsdesign.com/design_corner/showArticle.jhtml?articleID=1650536
> > 2
> >
> > http://www.linux-usb.org/usb2.html
> >
> > 73's Skip WB6YMH
> >
> > _______________________________________________
> > App_rpt mailing list
> > App_rpt@lists.illiana.net
> > http://lists.illiana.net/mailman/listinfo/app_rpt
> >
> > _______________________________________________
> > App_rpt mailing list
> > App_rpt@lists.illiana.net
> > http://lists.illiana.net/mailman/listinfo/app_rpt
> >
>
> _______________________________________________
> App_rpt mailing list
> App_rpt@lists.illiana.net
> http://lists.illiana.net/mailman/listinfo/app_rpt
>

_______________________________________________
App_rpt mailing list
App_rpt@lists.illiana.net
http://lists.illiana.net/mailman/listinfo/app_rpt