Flew the Lego Autopilot UAV and the cellphone autopilot UAV (sans cellphone) on Sunday to test some new stabilization calibration. Again, if the definition of a successful test is useful data and no crashing, this was a success, but it was a little tense at times. Lessons from these flights:
1) The Lego autopilot suffered a mechanical malfunction. One of the gear linkages came off in midair, freezing the autopilot in the "on" position. This meant that I had to fight a hard-left rudder all the way back to the ground. I'll superglue that part going forward.
2) The FMA stabilization units on both UAVs were very squirrelly on a hot, sunny day over concrete. At high sensitivity, they porpoised and wing-rocked badly. I think the solution is to move the IR sensors to the top of the fuselage body (as the instructions suggested).
3) The directional 2.4 Ghz antenna for the video download worked pretty well, but only when it was pointed right at the plane. Plane tracking is full-time job on the ground, which is definitely a limiting factor. Will try the omidirectional antenna once more (elevating it from the ground) before reconciling myself to a full-time antenna person during flights.
4) It turns out that we still have the Alameda air field site for now. Politics and red tape has slowed the dredging that was supposed to make the field unavailable, so we'll continue using it as long as we can.