ChipKit

The Bigtrak Junior is a re-make of a classic toy apparently. I'd never seen one before, I remember school having a couple of floor turtles but I don't think I ever got to play with them (probably because I'd been messing around too much and not doing my work!). Basically the toy is a little programmable toy car that takes simple commands from a keypad with forward, back, turn left, turn right, pause etc. This is all very well but with only one memory location for storing your program between use and no editor only a "clear and start again" it's not much fun trying to get it to do things. Wouldn't it be better if it could be programmed from a web browser and you could actually use some sort of save/load program option to store your ingenious route design? The solution I've come up with could be argued to be over-engineered some what and quite expensive but I like to look at it as "expandable".

Having replaced the electronics in a BigTrak junior, I needed some way to control it from my BeagleBone. Initial experiments with driving the L293D directly from GPIO pins were quite successful but I quickly discovered the opto-encoders on the toy were extremely high resolution and needed some proper real-time interrupt based handling. Doing this from the high-level Linux based environment on the BeagleBone just sounded too much trouble to me so I looked around for something a bit lower level to handle the opto-encoder tick counting. I was torn between an Arduino Uno and the ChipKit Uno. In the end I went for the ChipKit for two reasons, the BeagleBone I/O is only 3.3V and not 5V tolerant at all. This would have meant level translators were required on any interface between the Arduino and BeagleBone, but the ChipKit is already 3.3V. The other reason was at full speed the motor encoder ticks come through about 1200 times a second, with two that means about 2500 interrupts a second. I've got no idea how big the interrupt overhead is for the Arduino but I thought if I handle 16 or 32 bit tick counters in the interrupt with that frequency it is going to get a bit close to the limits of available CPU cycles, the ChipKit is 32bit and has an 80MHz clock so plenty of time to do all I needed. The features I've used are all available on the Arduino, due to some pin function constraints you'll need to swap around the pin order a bit but the code should work okay on an Uno (provided it can keep up).

If you haven't already, check out my ChipKIT implementation of Conway's Game of Life because I'm not explaining all that again! In this version there are some subtle changes to it that make it more pretty to watch, although arguably more confusing to see what's going on.

After the success of the ChipKIT pong demo I started to think about the ways to improve the graphics output and came up with the idea of a coarse pixel display. This is a more traditional pixel based display where a rigid grid is displayed on the screen rather than the more vector based display that I had used previously. This demands more of the CPU time because it has to copy the pixels to the display manually as there is no DMA on the UNO32, rather than only having to display the colour for a tiny fraction of each display line like the pong game did.

I'm going to be at OggCamp again this year. I'm planning a talk for the barcamp on open source hardware, particularly hardware more powerful or less well known than the Arduino. The ChipKIT will feature in the talk and hopefully the Pong game will be there for you to play. I'm also going to be bringing a ChipKIT MAX32 for the raffle courtesy of Farnell, so for your chance to win this awesome board, you need to come along.