Recent Posts

Pages: 1 ... 8 9 [10]
91
USB / Type-C cables connect GND and Shield, but adapters do not?
« Last post by jb on February 12, 2021, 12:25:03 pm »
USB Type-C Spec R2.0 - August 2019.pdf

For all the cables with plugs at both ends, it says

  • "All Ground return pins shall be connected together within the USB Type-C plug."
  • "Shield and GND grounds shall be connected within the plugs on both ends of the cable assembly."

For the adapters with receptacles on one end, it just says:

  • "All Ground return pins shall be connected together within the USB Type-C plug."

So the Shield should not be connected to the grounds in these adapters?  Anyone know why?
92
USB / Re: How to programatically force enumeration
« Last post by Renate on February 03, 2021, 01:38:16 pm »
Well, you could force enumeration:
By sending something to your device to make it disconnect.
By sending something to the controlling hub to make it disabled/enabled.

But it's better to leave everything connected.

Doesn't Windows scale whatever the touch screen is to what the screen is (at least for a single screen)?
Can't you just report the range as [0, 0] to [65535, 65535]?
Then you can scale/rotate the touches on your side.
93
USB / Re: seeing unwanted endpoints
« Last post by Renate on February 03, 2021, 10:49:44 am »
Mmm, hubs usually transmit upstream data to all downstream ports.
Is this what you're talking about?
Your hardware should be ignoring anything not on its own address.
Usually for hardware debugging you want to use a host controller with nothing else connected (just so it's quieter).
94
USB / Re: How to programatically force enumeration
« Last post by Jan Axelson on February 02, 2021, 05:14:26 pm »
I believe a new report descriptor would require enumerating as a new device or version in the device descriptor. This would require emulating detach and reattach.
95
USB / How to programatically force enumeration
« Last post by stivo77 on February 02, 2021, 05:04:44 pm »
I have an interesting problem with a touch screen device that interoperates with windows by responding to queries for touch HID report descriptors.  The project I am working on requires that the device be put into 'rotation mode' via a USB control message.  The device firmware is handling the command and dynamically does the rotation transform when reporting touches.  But the problem is that windows has already queried for the report descriptor and now the logical x/y limits in the descriptor are wrong.  Is there a windows API that can be called to force windows to request the report descriptor again? The device firmware knows that rotation  is in effect and will answer back with the right descriptor with the rotated x/y limits.  The control code that initiates the rotation is a C# library that uses LibUsbK, so I have a logical place in code to make windows API calls. Just not sure if there is a way to make windows ask for the report descriptor using the C# library.  Any help appreciated.
97
USB / Re: isochronous transfer in device-winusb_lvr and winusb_cs_31
« Last post by greglapt on January 31, 2021, 05:44:21 am »
The above example uses 3 endpoints for bulk transfers and unfortunately none isochronous. Is there an example of a configuration descriptor for WinUSB device uses isochronous  endpoints for your .net app (winusb_cs_31)?
99
USB / Re: seeing unwanted endpoints
« Last post by Jan Axelson on January 29, 2021, 08:54:42 pm »
The mechanism for filtering by address is specific to the device hardware/firmware. See your device's data sheet for details.
100
USB / isochronous transfer in device-winusb_lvr and winusb_cs_31
« Last post by greglapt on January 29, 2021, 04:39:55 pm »
Hello Jan.

Thanks for your book and work.

At the latest winusb_cs_31 there is the isochronous transfer test, but in the firmware source code (device-winusb_lvr.zip) I did not found any isochronous endpoints in the device descriptors and any code for isochronous transfers. What shuld I use for test isochronous transfer via WinUSB?
Pages: 1 ... 8 9 [10]