Date: Tue, 23 Aug 2005 03:54:25 -0500 (CDT) Subject: electronics are embarrassing X-UID: 81 Content-Type: IMAGE/JPEG; name="img1801.jpg" Content-Type: IMAGE/JPEG; name="img1805.jpg" Content-Type: IMAGE/JPEG; name="img1806.jpg" I'm not proud of the electronics anymore. It is the physical embodiment of "spaghetti code" except in wire. You shouldn't need to have a hemostat in each hand in order to assemble simple electronics. It's a crazy design, just about impossible to work on. I feel that I'm developing surgical skills. I realized over the weekend that the design could have been greatly simplified. It uses two full H-bridges for the drive motors. This gives three benefits: 1. solid state, no moving parts 2. generative braking 3. independent forward/reverse on each wheel The cost is fairly high parts count, complex mechanical construction and lots of wiring. It is a very expensive and intricate design mechanically. It's just like bad software. An alternative would use the existing heatsinks and two DPDT relays. One relay would allow switching between forward/reverse. The other would allow for braking. They'd be driven using bipolar transistors. Relays would be sensitive to shock. But the huge advantage would be greatly reduced parts count and consequently much lower complexity. Taking stock of the positives: 1. lessons learned about practical electronics - don't build electronics in three dimensions! - this also happened with the borg appliance! - why do I keep doing this? 2. more balanced view of this project My experience is that when engineers start projects, they begin in an ivory tower. Then as enthropy increases, eventually they abandon the tower and start walking across the field without quite knowing where they are going. Eventually, it begins to rain and storm so they start running for shelter. If the project is successful, they reach shelter but are completely soaked. I feel that right now I've abandoned the ivory tower with this robot. The electronics box still needs: 1. transient voltage suppressors (high power breakdown diodes) for protecting against inductive voltage spikes from the motors 2. the sense lines from the motor driver ICs connected to the motor outputs 3. the cooling fan connected to terminals 4. connect the piezo gyro to the microcontroller Number 4 is an oversight - I may have to undo all of the work done tonight in order to wrap some wires around I/O pins on the underside of the microcontroller socket. This prospect is so horrible that I'm going to try connecting the wires from the top of the socket.