Experimenter is a magazine created by EAA for people who build airplanes. We will report on amateur-built aircraft as well as ultralights and other light aircraft.
Issue link: http://experimenter.epubxp.com/i/126719
ber of buttons, and use off-the-shelf components wherever possible. The components needed to be integrated such that the same data is never entered twice. The logical approach was a distributed system design with smart components interfaced to a Linux PC (display) via serial lines. Since the display computer sees all data, it can integrate information to create higher-level metrics. As an example, to accurately calculate fuel at destination you need remaining route distance, ground speed, fuel levels, and fuel flow. This requires data from three different components. Scotty, my instrument instructor, would always tell me the autopilot could save my life. I remembered his words and powered the autopilot from a separate switch and connected it directly to the Garmin 430W. This independence will maximize the autopilot's usefulness during an emergency. modifying the ADC component. My idea was to output a voltage pulse from the ADC to the trim servos that became shorter as the airspeed increased. After two or three iterations, I could feel no sensitivity change from landing configuration to high cruise. It was one of those satisfying moments of elegantly solving an unforeseen problem within the original design. I did, however, add a switch to bypass the ADC trim motor pulses with a fixed-bus voltage in case the ADC failed. The first panel was a simple prototype to check out the system. I continued to fly this panel for another six months while I matured the software. Most of the modifications pertained to calibration routines and rendering yet another primary screen. The personal reward of making software modifications and then observing the changes in flight was immeasurable. In March 2007 I ripped out this panel and started over. I performed all the hardware engineering and programming for the nav interface and air-data computer (ADC). The nav interface box organizes the many Garmin 430W outputs into packets and sends them over the serial link, and the ADC calculates all the air values and becomes the garbage collector for many miscellaneous inputs and outputs. It was an interesting task working through the air data and digital smoothing mathematics. I settled into a regimen of working on the airframe during the day and programming at night. Ergonomics This is how our RV-10 panel looks today with square primary and round backup instrumentation. As you can see, except for backup instruments, the panel is clean, symmetrical, and designed for easy maintenance. Stein Air produced a panel blank that was very well thought out and became a fine foundation. First Flight In August 2006, the panel and airframe were ready for the first flight. The actual flight was uneventful and all systems worked fairly well. After a few flights, though, it was obvious that the elevator trim system was too sensitive as the RV-10 reached cruise speed. I implemented a solution by I chose two 10.4-inch sunlight-rated displays because nothing larger would fit. They are coupled with a VGA amplifier and show the same images. Yes, the same images, and it was a brilliant decision. I've seen panel designs completely oriented toward the left seat, even using canted segments to optimize pilot viewing. This design may provide some comfort for the pilot but psychologically demotes the right seat. Using the co-pilot display, Sara can conveniently see how the flight is progressing, can help look over weather issues ahead, has good input into the alternatives, and feels involved in the EAA Experimenter 21