This is a simple setup cobbled together to give a basic 2D testing harness for the microcontroller code. ATX power supply, AVR ISP mk II, RadioShack USB/Serial cable and a simple frame to hold the steppers in position.
The steppers are connected to the drawing surface using two potentiometer knobs attached to the usual drawing handles. The pot knobs have small screws which make it easy to detach the steppers, lift the frame away and erase the picture. The wiggly lines are manual testing, the straight lines are made using the steppers and a simple movement algorithm.
This is an almost standard RepRap Stepper Control board, except it's built on a prototyping breadboard. Interestingly it has a smaller form factor than the actual board, the components are packed quite tightly together. The L298 does not fit into a standard board because the second line of pins are off by 1/2. To fix this the pins were tweaked just slightly and the chip crammed into the board. You can see a row of drill holes from the first attempt at placing the diodes on the right.
Here is the reason building on prototype boards will never take off on a scale. While the component placement was pretty neat and straightforward connecting everything together is a mess. Making this single board took five long evenings and half a foot of desoldering wire. Some placement decisions which seemed good initially turned out to be horrible during the wiring part. The same device took one hour to solder onto a PCB.
The standard RepRap Stepper Motor Driver v1.1. easy placement of components, quick to assemble.
I've built several microcontroller boards but none of them has stuck yet. Instead I ended up wiring up a simple breadboard layout. I'm using an ATmega8, mainly because I happen to have several of them around. On the left is a MAX232 for serial connection and you can see the ISP programming header on the top. To make it easier to attach to the ports the RESET pin is wried next to the SPI interface. The breadboard is divided into a 5V section on the left and a 3V section on the right. The ATX power supply is connected to the board power input. On the right, in the 3V area, is a 512MB SanDisk SD card.
I didn't have SD card holders handy but wanted to start experimenting with the card during a weekend. My first option was to solder the wires to the card directly, but that had the drawback of making the card impossible to use in a computer. I tried building a holder out tape and wire, several different versions using the original case the card came in, but nothing really provided the necessary physical interface. This was supposed to be a last ditch effort to get the card temporarily hooked up, but it turned out to be a very versatile setup. The wires put downward pressure on the connectors which assured a good electrical connection plus it keeps the card in place. The breadboard spacing is ideal for the card and a number of resistors under the card keep it at a slight angle. If you are careful you can slide the in and out of the socket which makes for easy pushing into the laptop.
And here is a snapshot of the firmware terminal. It's a simple command line oriented interface that lets you run the steppers and poke around the SD card. The two sectors dumped here are the FAT16 Boot Sector and the Directory Table for the root directory on the card. The last two lines of output before the prompt are after a restart, you can see some details of the SD card in the probe line. Given the card was bought in July 2008 it took 1 year and 8 months until it found a home. As test data there are three installers on the card (largest files I could find): The Java JDK plus eval versions of Alibre and Maya.