Objective Development Forums

One more file, usbconfig.h: /* Name: usbconfig.h * Project: AVR USB driver * Author: Christian Starkjohann * Creation Date: 2005-04-01 * Tabsize: 4 * Copyright: (c) 2005 by OBJECTIVE DEVELOPMENT Software GmbH * License: GNU GPL v2 (see License.txt) or proprietary (CommercialLicense.txt) * This Revis...

Hi everyone. I am having some weird keyboard problems with VUSB. My project is convert a BBC Micro keyboard to USB, a path well trodden by others who have converted keyboards. For some reason they keyboard only seems to work properly if plugged in after the OS has loaded. If plugged in when the mach...

Kash, I guess you are referring to the Linux HID driver. I will pass the info on to my Linux guy and report back if it works. Thanks for the hint!

I did understand what you were getting at and attempted to implement it. I must have done something wrong I suppose. It's not an elegant solution either way and presents further problems for me because my device (Retro Adapter) re-configures itself on the fly by disconnecting, changing the USB confi...

Well I think the short answer is that Linux users are boned.

Okay, so just for clarity, using two endpoints works and I should try implementing it? Oh, and I looked on Ralph's site but couldn't see where he did it that way. -- EDIT -- I tried it and could not get it to work. To further complicate things even if it can be made to work then I need to be able to...

I noticed one thing which may be of interest. V-USB reports USB version 0x10, 0x01. Most USB 1.1 devices seem to go with 0x01, 0x01. I will send a test firmware to my Linux tester to see if it makes any difference. Oh, and BTW, Windows gets confused if you switch HID descriptor too much. You can go ...

Ah, I see what you mean.

I'd just like to point out at this stage that I'm not trying to run V-USB down or anything. It's an amazing bit of kit and seems very robust and reliable in general use. You can't beat being able to use USB on a <€1 microcontroller.

Well the composite devices I have checked all report back USB version 1.1 so I don't think it is a problem with the protocol as such. It has to be something about the way V-USB works. I'm tempted to get in touch with the Linux USB stack guys and see if we can find a solution. It should be possible t...

Ah, okay, thanks for the clarification. It's interesting that we have both come to the same conclusion.

Still, I would like to know what the problem is. My design is fixed now with an ATmega168 so I can't stop using V-USB.

I tried to contact Ralf but his email address is bouncing. I can't see anything he does differently in his code anyway. As a test I tried copying his data (HID report, USB descriptors etc.) and it didn't work. Maybe if I find the time I will build his hardware but it doesn't seem worth it at the mom...

No-one?

I am trying to create a device that appears to the computer as two separate joysticks. I have it working in Windows but not in Linux. Windows can see two separate joysticks in the game controller applet, but Linux only sees one in /dev/input. I know Linux can show two as I have a dual Playstation to...

I am working on this too. So far no luck. My device is recognised by Windows and seems to be receiving HID reports, but no movement or button presses can be seen. I think part of the problem is the HID descriptor. The one used by the UPCB is not the standard Sixaxis one. I think the best option will...

Thanks for the reply. I am not the OP, but I have a similar problem. I am getting some strange results. The above code works, sort of. I get 50 bytes returned with the correct data when I use SimpleHIDWrite to check it. For some reason though, the HID gamepad interface just seems to ignore it. I get...