Re: USB Functional RNDIS timeouts
- From: Henrik Viklund <henrik.viklund@xxxxxxxxx>
- Date: Thu, 7 Feb 2008 01:21:13 -0800 (PST)
Hi Trond,
A USB device indicate its speed by pulling either the D+ or D- line
high to 3.3 volts. A full speed device, will use a pull up resistor
attached to D+ to specify itself as a full speed device. The pull-ups
at the device end will also be used by the host or hub to detect the
presence of a device connected to its port. Without pull up, the USB
host assumes there is nothing connected to the bus.
The usual way of achieving what you want (assuming PXA27x) is to hook
up the pull up to a GPIO so you control the pull-up with the GPIO pin
(hopefully the HW guy put one in for you to use).
It should be sufficient to activate/deactivate the pullup state in the
Start(...) and Stop(...) methods of the usbfn driver to let the host
detect the usb function once the driver has initiated.
Henrik Viklund
http://www.addlogic.se
On 6 Feb, 15:55, Bølla <B...@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote:
Thanks for your support!
Problem is solved.
It was only a registry configuration that was needed.
I had to change the following line of code from:
[HKEY_LOCAL_MACHINE\Comm\RNDISMINI1\Parms\TcpIp]
to:
[HKEY_LOCAL_MACHINE\Comm\RNDISFN1\Parms\TcpIp]
and then it worked!
Another related issue:
How can I get my device to be detected by the PC (host) when the USB-cable
is already plugged in during boot of my device? (PC is on all the time.)
My users do not like to remove and then insert the cable again. It should
automatically detect my device. I have forced an interrupt detecting the
attach cable, but it looks like it comes to late. The PC tells Unknown device
before my device is fully booted.
I thought there should be some registry setting delaying or wait with USB
startup until after my device is fully booted.
/trond
"Silver" wrote:
The static zap possibly appears as an interrupt at the PXA, or enough of a
charge to indicate that USB presence is detected.
This stinks a small leakage current from some kind of detector circuit.
Which piece of hardware that is being affected will require some debugging.
When you zap the board when changed? That is was the system stuck in a low
power state then woken up? Try adding debug output around the OEMIdle code
and sleep code (if any). Is it possible that attach/detach are setup to
prevent the system going into idle/sleep but on the other hand it is not
capable of bringing it out of that state?
- Try stubbing out OEMIdle and sleep code. Does it still happen?
- is this your own hardware design? Is there a GPIO line or similar that
detects USB presence but could slowly discharge. However it is somehow
'reset' by insert remove if it is still above a certain threshold. I realize
how vague that that sounds.
- Could you tell us a bit more about the USB hardware, are you using the PXA
USB client controller? Doesn't that have a bunch of errata?
- Can you get some debug output from the OS about what state it is in? If
you connect a rs232 serial cable does the problem go away - that tells you
something?
Geoff
--- Dölj citerad text -
- Visa citerad text -
.
- References:
- Re: USB Functional RNDIS timeouts
- From: Bølla
- Re: USB Functional RNDIS timeouts
- Prev by Date: TFAT issue
- Next by Date: Landscape mode on Windows CE 5.0
- Previous by thread: Re: USB Functional RNDIS timeouts
- Next by thread: USB charging feature on WindowsCE 5.0
- Index(es):
Relevant Pages
|
