Ides of… Oh Crud, I’m Overdue


My quadro­tor heli­copter project has had some rough times. I am now two years into its devel­op­ment. I designed a lot of it from the ground up, most of it in elec­tron­ics and soft­ware. The cost so far has been some­where between $500 and $1000, but I have no way of recu­per­at­ing that cost or explain­ing it to my par­ents. How­ev­er, I expe­ri­enced some­thing extreme­ly aggra­vat­ing in the past week.

Two weeks ago, I decid­ed to switch the main con­troller from a PIC 18F2550 with 28 pins in SPDIP to its big­ger broth­er, the 18F4550 with 40 pins in PDIP. It’s sup­posed to have iden­ti­cal hard­ware, but when I port­ed over my Wii Nunchuk read­er code (I’m actu­al­ly not sup­posed to link the pub­lic SVN, but I can do what­ev­er I want >:D), which uses the MSSP mod­ule for hard­ware I2C com­mu­ni­ca­tions with the Nunchuk, the com­mu­ni­ca­tions ceased to work. I added some debug­ging code, and found out that the MSSP mod­ule on the 18F4550 sim­ply sucks com­pared to the one on the cheap­er 18F2550. Where­as I could get a rock-solid stream of data with the 18F2550, the big­ger chip, even after days of care­ful tweak­ing, was squir­re­ly and only read data out in unpre­dictable bursts.

Need­less to say, I am now stuck with few­er pins than I need, and expan­sion will be dif­fi­cult, if not impos­si­ble. The long range radio com­mu­ni­ca­tions mod­ule may need to be mul­ti­plexed with the Z-axis gyro’s SPI inter­face (which is soft­ware-based ATM), and I hon­est­ly have no idea how I’m going to deal with the ultra­son­ic rangefind­er (for main­tain­ing height).

Just had to rant for some rea­son. 😐