[PD] Re: [PD-announce] usb iobox comming - testers needed !
cyrille.henry at la-kitchen.fr
Tue Feb 10 13:14:48 CET 2004
>0.) a cimputer with usb and hid drivers (that means: if your comp can use a
>mouse/keyboard/joystick... on usb)
>1.) a programmer that is cappable of burning the microchip pic 16c765
ok, but I don't have a uv errazer : does a compatible chip exist in
>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
>3.) the current test-firmware from http://mamalala.de/iobox-test.zip
>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
>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
>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
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...
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) )...
>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
>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
>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
>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,
>PD-announce mailing list
>PD-announce at iem.at
More information about the Pd-list