[PD] Re: [PD-announce] usb iobox comming - testers needed !

[Cyrille Henry]

hello christian,


>0.) a cimputer with usb and hid drivers (that means: if your comp can use a 
>mouse/keyboard/joystick... on usb)
>
>  
>
ok

>1.) a programmer that is cappable of burning the microchip pic 16c765 
>microcontroller
>
>  
>
ok, but I don't have a uv errazer : does a compatible chip exist in 
Flash memory?

>2.) of course one if that microcontrollers, hard to get in germany, sorry.
>be sure to get the -jw version, which is uv-erasable. otp-chips are a waste at 
>the current state of my development.... remember, its just a test 
>firmware....
>  
>
ok

>3.) the current test-firmware from http://mamalala.de/iobox-test.zip
>  
>
ok

>4.) a small prototype board, a usb-b connector, a cable from usb-b to usb-a 
>(to you computer) and led's (with 1k resistors), buttons and faders as 
>needed. take look at the microcontrollers datasheet 
>(http://www.microchip.com/1010/pline/picmicro/category/perictrl/14kbytes/devices/16c765/index.htm)
>we need a 6 mhz xtal, portb is used to display led's, portd is used for 8 
>buttons, an0-an7 as analogue inputs plus the two pwm outputs on rc1 and 
>rc2... 
>  
>
ok

>5.) pd or jmax with a event-device object (on jmax i made ev_dev, i think h.c. 
>steiner has made a similar object for pd on linux)
>  
>
hum, I don't know.
I use pd / windows .
does a such object exist?


I don't have much time and dont really know how to write hid object for 
pd (could learn : I already write object for pd, but don't have lot's of 
time).

So, I think I a pd for windows exist, I could make a try. otherwise, I 
think I'll not have time.

>after burning and assembling, the thing does the following :
>on connecting to the computer, it enumarates to the host as hid device.
>it sets up 64 led outputs, 64 single key inputs, 88 adc inputs, 64 dac outputs 
>and a 3-byte in-/output. the adc out's are declared (by the device) as 
>absolute events, the buttons and led's as what they are, respectively.
>the dac is accessed as led, just with a 16 bit value range instead of the 
>usual 1-bit on/off state .... the 3-byte in/out is done as one 3-byte wide 
>button and a 3-byte wide led.
>  
>
ok, really nice for a single low cost chip...

nice work.
please tell me if I can help you.
(restriction are : no uv errazer, and not lot's of time, I'm on windows 
(and a bit of linux) )...

cyrille henry





>currently used by the chip (as this are the built-in capabilities) are 8 adc, 
>2 dac (done by using the pwm outputs of the chip) , 8 buttons and 8 led's.
>(see http://mamalala.de/iobox-test.jph how that looks in jmax)
>  
>

>that makes up the following events :
>
>inputs: (from device to computer)
>a:	type 1 (key) 
>	code 256 - 263 (would be 256 - 319 in full version)
>	range 0/1 (on/off)
>b:	type 3 (abs. movement)
>	code 6-13 (would be 6-92 in full version)
>	range 0-255 (would be 0-255|1023|4095 in full version, depending on used adc)
>
>outputs (all are type 17:led) 
>a:	0-7 (would be 0-63 in full version)
>	range 0/1 (on/off)
>b:	64-65 (would be 64-127 in full version)
>	range 0-256 (would be 0-255|1023|4095 in full version, depending on used dac)
>
>the rawbyte in/out's do nothing yet, but will be used for thing like buffered 
>midi in/out, serial in/out, rotary encoders, small lcd displays, [put your 
>ideas here]....
>
>now comes the point where i need your assistance :
>
>please test if the device connects to your computer without error.
>
>if you are on linux, test it's in/output. to send to the device in a propper 
>way, you need a recent linux-usb driver. older ones (for example 2.4.18) dont 
>include the report-id to the device, but we need that ;-) newer drivers do 
>(as in linux 2.4.24 and up....)
>all the i/o should be do-able by the standard linux event input device 
>interface (evdev) and the linux usb-hid driver compiled to use that 
>interface....
>
>more important to me is, if you are on mac/windows/.... :
>i cant supply objects for pd/jmax/... on these platforms.
>but, from what i expect they should be easy to write. just access the hid 
>interface of the host, it will give/take the reports as needed.....
>for example, a usb-hid conform "driver" in source-form for a mouse, keyboard, 
>joystick, tablet.... should provide a good basis for that.....
>if you plug it into window, it already announces itself as unknown hid device.
>
>please help, provide some objects for that thing !
>all the linux stuff i will do myself, ev_dev is already there. what i will do 
>next is a alsa-driver that wraps the device into a midi stream, for use wit 
>other sequencer/synth apps. such a thing would be nice on other platforms 
>too. imagine all midi-conform apps could use that thins faderbox, for 
>example...
>
>ok, thats all for now. if you want to try that and help me out on that part,
>drop me a note and "just do it". report back your experience....
>
>thnaks a lot,
>
>chris
>
>
>_______________________________________________
>PD-announce mailing list
>PD-announce at iem.at
>http://iem.at/cgi-bin/mailman/listinfo/pd-announce
>
>  
>