Other software sniffers like HHD USB monitor fail to catch the
enumeration sequence, when the sequence doesn't succeed. Then, they
are not helpful for debug of enumeration.
Anyway, SourceUSB and BusHound are expensive.
If it is your personal project, and it's just for low- and full-speed
device, Beagle is a good choice.
The strength of hardware bus analyzers is that you can see NAK and SOF
on the bus. It is useful for performance tuning.
When you see many NAKs on the bus, the performance problem lies on the
device response. Instead, you see no NAK on the bus, the problem lies
on the host side.
Tsuneo
--- In l..., Michael Gray wrote: >
> Thanks for all the comments on the hardware and software solutions for USB > monitoring.
>
> The Beagle looks like a good device and at $400 is pretty low risk even if > it doesn't do everything. The Ellisys Tracker
110 is a great device, just > costs 3X more by the time you are done.
>
> I tried several versions of the software monitoring tools. They weren't > all that useful for low level stuff. I was most
interested in watching
> the enumeration sequence to make sure the LPC2148 USB driver was answering > all the requests properly.
> On Thu, 27 Dec 2007, Paul Curtis wrote:
>
> > Hi,
> >
> > > Get a Beagle. Cheapest hardware protocol analyzer for $400.
> > >
> > > http://www.totalphase.com/products/beagle/usb12/
> >
> > If you can believe the specs, that looks like an all-time great device! I > > have an Eliisys Tracker 110 and, although it's
a great device, it has had a > > couple of times where it's lost the plot and
sent me on a wild goose
> > chase--a reboot fixed that.
> >
> > What really attracts me to this Beagle thing is the real time
> > display--that's what I asked Ellisys about a long time ago and their
> > response was that it was hard to do (i.e. don't expect it). I do
wade
> > through logs after capturing them, but real time display would be a
> > fantastic way to scratch-n-sniff device behaviour.
> >
> > I also sprung for the Ellisys class decoder which makes it a pretty
> > expensive solution.
> >
> > So, if the hardware and software are to be believed and stable, this looks > > like a better option than my favoured Ellisys
box.
> >
> > (I did try using host-side USB stuff on Windows, it was terrible, much
> > better to get a hardware device to see what's happening on the bus).
> >
> > --
> > Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
> > CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
> >
> >
> >
> >
> >
> >
> >
Reply by s.ro...@yahoo.com●December 30, 20072007-12-30
Thanks for all the comments on the hardware and software solutions for USB
monitoring.
The Beagle looks like a good device and at $400 is pretty low risk even if
it doesn't do everything. The Ellisys Tracker 110 is a great device,
just
costs 3X more by the time you are done.
I tried several versions of the software monitoring tools. They weren't
all that useful for low level stuff. I was most interested in watching
the enumeration sequence to make sure the LPC2148 USB driver was answering
all the requests properly.
On Thu, 27 Dec 2007, Paul Curtis wrote:
> Hi,
>
> > Get a Beagle. Cheapest hardware protocol analyzer for $400.
> >
> > http://www.totalphase.com/products/beagle/usb12/
>
> If you can believe the specs, that looks like an all-time great device! I
> have an Eliisys Tracker 110 and, although it's a great device, it has had
a
> couple of times where it's lost the plot and sent me on a wild goose
> chase--a reboot fixed that.
>
> What really attracts me to this Beagle thing is the real time
> display--that's what I asked Ellisys about a long time ago and their
> response was that it was hard to do (i.e. don't expect it). I do wade
> through logs after capturing them, but real time display would be a
> fantastic way to scratch-n-sniff device behaviour.
>
> I also sprung for the Ellisys class decoder which makes it a pretty
> expensive solution.
>
> So, if the hardware and software are to be believed and stable, this looks
> like a better option than my favoured Ellisys box.
>
> (I did try using host-side USB stuff on Windows, it was terrible, much
> better to get a hardware device to see what's happening on the bus).
>
> --
> Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
> CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
>
>
>
Reply by sig5...@hotmail.com●December 27, 20072007-12-27
I have one. Had it for over a year. It works fine. Never had any problem with
it. But I have to admit that I never really ran into much of a need for
troubleshooting my USB designs. All my USB MCUs so far have been Cypress chips.
Soon I will be doing a USB/Enet design with the LPC2468 and perhaps I may end
up using it more on that.
If you can believe the specs, that looks like an all-time great device! I
have an Eliisys Tracker 110 and, although it's a great device, it has had
a
couple of times where it's lost the plot and sent me on a wild goose
chase--a reboot fixed that.
What really attracts me to this Beagle thing is the real time
display--that's what I asked Ellisys about a long time ago and their
response was that it was hard to do (i.e. don't expect it). I do wade
through logs after capturing them, but real time display would be a
fantastic way to scratch-n-sniff device behaviour.
I also sprung for the Ellisys class decoder which makes it a pretty
expensive solution.
So, if the hardware and software are to be believed and stable, this looks
like a better option than my favoured Ellisys box.
(I did try using host-side USB stuff on Windows, it was terrible, much
better to get a hardware device to see what's happening on the bus).
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
Reply by sig5...@hotmail.com●December 26, 20072007-12-26
Get a Beagle. Cheapest hardware protocol analyzer for $400.
all software monitors are monitoring the communication between drivers.
So anything that happens on the lowest driver level will not be seen by
them. The start of the enumeration is at this level. Thus if your device
is completely not working, you will see nothing. These monitors are only
useful for higher level monitoring (like to monitor class specific stuff).
On the other hand USB has some error recovery mechanisms at the hardware
level which can not be seen by software at all. So you can have a device
that seems to work ok, but it is not just the hardware retries and
solves the problem most times..
I'm developing an LPC2148 USB application and want to capture the USB
traffic on the host for debug. Any recommendations for USB Host
monitoring software? I've found the following:
They all seem to have the same feature sets for the most part.
However, I'm not sure if they'll capture the device enumeration
sequence. Something with hardware support, such as Ellisys, may be
required.
