or what didn't quite go too well

Cornell University EE 476 Final Project

Fertilizer Feed Rate Controller

Warren Scott


When you have a lot of soldering to do, it is very worthwhile to invest in a good soldering iron. The cheap things at Radio Shack that just get hot when they are plugged in aren't worth much when you're under the iron for a few hours at a time.


CodeVision AVR doesn't deal with comparisons between unsigned ints and ints as I expected. This cost a few days of debugging. In hindsight I should have been thinking more in a microprocessor world than a Microsoft world.


I should have thought more about using the serial port. PORTD.0 and PORTD.1 were left unused so that the serial port could be used if one wanted, but I should have started putting features on it. I don't know what, but having a laptop in your tractor is not out of the question, especially if one is going to implement GPS later.


The control system still has quite a bit to be desired. Unfortunately, the system it is controlling is so unknown nothing can be done except guess and check at this point, especially since the system does not exist to do experimental response testing on. We can only wait to see how things turn out.


No extensive testing of the radar unit was done. It was tested in that when powered and moved it gave out a 12V square wave, and that the frequency varied as one walked faster with it. However, its calibration was never tested. I should have made an adapter for a cigarette lighter and drove around campus with a weird black thing pointed out of the window and seen how many sorority chicks in Grand Cherokees dropped their cell phones because their radar detectors went off :)






Go Home