[pvrusb2] SOLVED: 29xxx pvrusb2 fails on "Device being rendered inoperable" after firmware load

[Mike Isely]

On Fri, 27 Nov 2009, Antti Pyykko wrote:

> On Thu, 26 Nov 2009, Mike Isely wrote:
> 
> > Can you try a different USB port, controller, hub, and / or cable?
> 
> Unbelieveable.
> 
> I went through every single USB cable I have nearby or in use and the device
> worked with ONE (1) cable out of the 8 cables I have at hand.
> 
> That PVR is one picky sucker!

Hmm, well I have not had that level of problems here.  Not that this 
would make you feel any better about it...  I've got a junk box full of 
USB cables - they all work fine.  One time I did a 27 hour continuous 
capture just to prove to myself that everything really was stable...

This might be that the signal quality was already marginal (perhaps a 
bad solder joint inside the USB A-socket or maybe the hub - if you're 
using one - has issues) and without an absolutely flawless cable that 
you'd have problems.  You said it works on a laptop - have you tried the 
various cable combinations there too?

For what it's worth I've had a fair amount of USB "fighting" to do at my 
day job.  We built a self-contained box that used an industrial-grade 
PCI-104 form factor single board PC (running Linux, of course).  The 
board has two USB ports.  We wired those out to several devices and it 
is simply amazing just how sensitive that wiring can be - and it is even 
more amazing the wierd fickle problems that can happen if you get it 
wrong.  The contractor that assembled our first few prototypes didn't 
know how to deal with 480MHz differential pair signalling and seriously 
messed up the USB wiring.  The wires weren't twisted right, incorrect 
wire was used, and in some cases the wires in the signal pair were 
routed different ways - including intersecting the signal pair with 
other unrelated signals.  And in spite of all that abuse, a 10 pin USB 
flash module worked perfectly at the far end of one of those bad wire 
runs.  Yet another video capture device suffered a long litany of 
reliability problems - about %0.001 of the transfers would be corrupted 
- and that's enough to ruin USB connectivity.  We went through long 
series of ECOs before that mess got straightened out, and a number of 
people learned hard lessons about just how tricky it is to maintain 
proper USB signal integrity!  We found that bad wiring on the 10-pin 
module's port could even screw up the signal integrity on the *other* 
port.


> 
> Thank you for your help, Mike! Sorry for wasting your time with a stupid
> hardware problem.
> 

No problem.  "Stuff" happens :-)

  -Mike


-- 

Mike Isely
isely @ isely (dot) net
PGP: 03 54 43 4D 75 E5 CC 92 71 16 01 E2 B5 F5 C1 E8