Saturday, March 31, 2007

Flight Testing - TH1

After working with Dave remotely from Argentina, staying updated on all of the project details, I finally had to opportunity to come back home to California for a short visit. As luck would have it, I ended up staying on the boat right next to him in the Oakland marina.

Our first glider flying outing was at Tilden park in Berkeley where we walked up to Vollmer Peak and around to its backside which looks directly at Mt. Diablo to east. The winds were extremely light at about 0-5mph but Dave hucked the robot off anyways and flew it around a couple of times on R/C, my first time to see the TH1 in action. Due to the extremely light winds we called it a day after only a couple of short flights and didn't switch into autopilot once.

Our second trip out was a repeat up to Vollmer peak where the conditions where much stronger. We got the glider up and away from the hill and flipped it back and forth from manual to R/C control using the onboard autopilot robot. It works! TH1 hasn't quite learned how to seek and climb in thermals yet but it can maintain smooth and level flights even in the turns as it hunts for pre-programmed way points while in autopilot.

Since these first few flights a couple of weeks ago, Dave has been working on monitoring the power levels (voltage) for the onboard batteries powering the gumstick computer and the servos. He is also working on converting the PWM (pulse wave modulation) digital signal, used to control the switching from manual R/C control to autopilot, to an analog value.

"If the glider flies out of range of the R/C transmitter, the receiver stops receiving the R/C signal. Since there is no pulse at the PIN, the controller board times out, while waiting for a pulse, introducing a delay in the loop that is used to read and write PWM values. Now I'm reading analog values so that the onboard computer can decide if its on autopilot or not," says McCutcheon.