xbox360 chatpad, awsome backlit mini keyboard

**justchat_1** · 08-24-2010, 01:44 PM

Originally Posted by w239

I have decided to just post anything I find that may be of interest regarding the Chatpad driver.

A guy is trying to create a windows 7 driver(link below), pretty recent posts. 1st and last post the are most interesting.

http://www.autoitscript.com/forum/in...owtopic=116941

Cheers

Its not an HID device....so no you certainly can't create a script for it using autoit

**sirbod** · 08-26-2010, 04:30 AM

Originally Posted by Stateamuse

I Found Something that Maybe of help, Just Download the object and read the spin with notepad or an editor, it states that you have to keep telling the pad to stay on.

http://obex.parallax.com/objects/510/

It also contains scan codes a hardware examples.

Hope this helps!

This is actually rather useful, as they've figured out the communication protocol. ie.

- Send 87,02,8C,1F,CC,00 to initialise and await a 12 byte response
- Send 87,02,8C,1B,D0,00 as a keep alive every 2 seconds
- When awake it sends an 8 byte status message every 70mS
- When a key is pressed it sends an 8 byte message, beginning B4, C5 that contains the keycode

**CrazyIvan** · 08-29-2010, 11:45 AM

Excellent!! This keep alive must be the missing puzzle piece that stumped the folks who got it to work for 1 keypress.

I wonder if you need to constantly send keep alive's to teh chatpad or would you just send the keep alive before sending a key so you don't waste the battery when idle?

**Grumbel** · 08-29-2010, 12:10 PM

Originally Posted by CrazyIvan

Excellent!! This keep alive must be the missing puzzle piece that stumped the folks who got it to work for 1 keypress.

A little to early to celebrate. That the pad needs keep alive messages is already known, in the case of an original Xbox360 it will send 1F, 1E, 1F, 1E, ... each second, 1B is used to activate keyboard input and a bunch of other commands can be used to make the LEDs light up. The problem is that the whole controller dies (i.e. will not respond to anything and will need a usb reset to be brought back to live) whenever you try to send those keep alive messages. This might be a Linux specific problem or maybe it needs some specific timing or such, which is why it would be rather useful if somebody would try to write a Windows driver with the given knowledge to see what the pad does in Windows.

**CrazyIvan** · 08-29-2010, 12:22 PM

Originally Posted by Grumbel

A little to early to celebrate. That the pad needs keep alive messages is already known, in the case of an original Xbox360 it will send 1F, 1E, 1F, 1E, ... each second, 1B is used to activate keyboard input and a bunch of other commands can be used to make the LEDs light up. The problem is that the whole controller dies (i.e. will not respond to anything and will need a usb reset to be brought back to live) whenever you try to send those keep alive messages.

Gotcha. Have you tried sending the 1B keepalive at least every 2 seconds like the parallax guys say?

**Grumbel** · 08-29-2010, 01:37 PM

Originally Posted by CrazyIvan

Gotcha. Have you tried sending the 1B keepalive at least every 2 seconds like the parallax guys say?

If I send 1F, 1E, ... like the real Xbox360 the controller dies, if I send 1F, 1B, 1B, ... the controller stays alive, but doesn't send more then the very first keypress, only toggling the LEDs works after that.

**sirbod** · 09-02-2010, 02:18 PM

The fact the XB sends 1E / 1F alternately every second may be significant, timing may well be an issue.

Also, although the wValue* entry is mostly known (1E / 1F etc), what's the wIndex value for? I've seen various analyzer logs where the wIndex value is different - is that significant?

What are the full 8 byte Setup Packets you're sending? The XB trace uses a bmRequestType of 41 (ie Direction: Host to Device, Type: Vendor, Recipient: Interface) and yet the code on the previous page uses 87 (ie Direction: Device to Host, Type: Standard, Recipient: Device, Interface, Endpoint). They can't both be right as they're going in different directions.

* USB Setup 8 byte Packet:
0 - bmRequestType
1 - bRequest
2/3 - wValue
4/5 - wIndex
6/7 - wLength

**Grumbel** · 09-02-2010, 02:30 PM

Originally Posted by sirbod

What are the full 8 byte Setup Packets you're sending?

I am sending:

reqtype: LIBUSB_REQUEST_TYPE_VENDOR | LIBUSB_RECIPIENT_INTERFACE
request: LIBUSB_REQUEST_GET_STATUS
value: <here goes the 1e, 1f, 1b, etc.>
index: 0x02
data: 0
length: 0

I think in raw bytes that should be something like (if I didn't mess up the endianess):

0x41 0x0 {value} 0x00 0x02 0x00 0x00 0x00

**langarn** · 09-30-2010, 04:14 PM

I have followed the work some of you have put down in this project.
Cant honestly say I understand much. Is it a possibility that someone is near a solution?

Great work u guys, thanks.

**royalcs** · 11-11-2010, 07:50 AM

hey, i found this OSX driver:
http://tattiebogle.net/index.php/Pro...ller/OsxDriver

ChatPad

I have got the Microsoft ChatPad working with my wired controller. The latest release of the driver includes support, and I'll be updating the USB information section of this website shortly. I've not yet checked the wireless receiver for compatibility.

source is available, so maybe this helps