uint8_t readDigital8() {
uint8_t temp = 0;
uint8_t result = PIND; // Read feedback of six digital outputs, pins 2 through 7
// The following swap the bit order as needed by liberlab

if (result & 128) temp = temp + 1;
if (result & 64) temp = temp + 2;
if (result & 32) temp = temp + 4;
if (result & 16) temp = temp + 8;
if (result & temp = temp + 16;
if (result & 4) temp = temp + 32;
result = temp;
return result;
}

Cheers

on line 50, should that be a “G” or a “g”. I ask because line49 is “G”.

Arduino and Liberlab are both two very interresing project.
I believe that some improvements could be made.

To increase the bandwidth, the uC should store data in memory (a buffer) and send the buffer to the computer using the serial com.
The speed for acquiring data and the speed for sending data to the computer will be differents so if buffer is full, the aquistion will stop (or the beginning of the buffer overwritten)

This kind of mode is called “burst” by opposition of “sample” mode
http://www.google.com/search?hl=en&q=data+acquisition+burst+sample&spell=1

An over usefull feature could be to add a triggered scope
http://en.wikipedia.org/wiki/Oscilloscope#Trigger

About Arduino… instead of using a FTDI chip they should use a USB harwared capable uC such as PIC18F2550, PIC18F4550, ATMEL AVR USB http://www.atmel.com/products/AVR/usb/

an other idea could be to use a software implementation of USB such as http://www.obdev.at/products/avrusb/index.html

Best regards

Sebastien CELLES

Liberlab is not really active these days (lack of time) and since Arduino is now such an great project, and easily available, I “”"planed”"” (a while ago) to adapt Liberlab to it. So, I’m really glad Randall actually did it

For information, the Liberlab firmware (I wrote 2.5 years ago) is not the most clean/efficient code and there is plenty of space for improvement. In particular I think the protocol is way too long to increase the actual data acquisition speed (100/150 measures/sec on the Liberlab board at 57600 bauds). For example there’s no real need for these extra “A1″ “A2″ “A3″ “A4″ characters (if we know where the data is in the string) and also the firmware shouldn’t need to process and send the four ADC channels if the user just use one.
I began few months ago to make these changes in a firmware for the Arduino (and reading the data sent using the Liberlab python module). It was much quicker (over 1000 measures/sec, not sure of the baudrate anymore) but each few seconds I had a slowdown for a fraction of a second and didn’t investigate more. I’ll try to digg up that code for those interested.

Apart from the GUI software there are interesting things to do with the Python module:
http://pcsm1.u-strasbg.fr/liberlab/englishversion/logiciels/module-liberlab/document_view

A fun example is to have three photo-resistors and one thermistor linked to the four ADC so you can play with the light to deform the x,y,z length of a 3D cube and heat the thermistor to change the color of the cube:
http://www.flickr.com/photos/frenchy/100834508/
http://liberlab.cvs.sourceforge.net/liberlab/liberlab-module/liberlab-demo-vpython.py

Another example is to see a board live on the web and have the measures in real time and command the output in the browser (I’ll try to digg up the code and put it online):
http://www.flickr.com/photos/frenchy/175546610/in/set-72057594057698650/

Anyway, thanks, I think Arduino is the way to go. I’ll try to make some time for Liberlab software on the Arduino

francois