My first gluonpilot flights - a grand success

[Mitch]

First flights today! I've been having sooo much fun playing with the board with different test firmware that I didn't want to stop and really use it. Tom's circle navigation really got my attention so I decided to install the board in an airplane I already had and do some flying with it.

I used a precision aerobatics katana, very popular in the states three or four years ago. Hacker outrunner and ESC, Hitech servos and Hitech receiver modified for PPM. It was only about 30 minutes of work to retrofit the airplane for the gluonpilot board and GPS module. I mechanically reduced all control surface travels to about 1/2 of where they were set for aggressive 3D flying. Then I spent a few days checking everything multiple times and looking at all of the failsafe issues. That wasn't really necessary because I think Tom has already considered most of the issues I was looking at. I used the same initial gains Tom provided from his funjet - I will eventually do some more tweaking - because it is very docile now. The rudder is only manual and is connected directly to the receiver via PWM.

The first flight I did was just two trips around the pattern in manual mode only. I wanted to establish exact trims and find a stable flight power setting. I then downloaded the flash and checked to make sure everything was tracking correctly.

During the second flight I tested the "stabilized mode" during a straight and level pass and then made a large oval. Perfect!!! Then I did some basic acro so I could check the imu performance. Too cool!!! I'm going to spend some time reviewing the data now!!!

I've attached some photos, youtube re-creations, and the KMLs below

Short trim flight

Link


First stabilized flight - then some acro!

Link

[lukasz]

Very Nice

I also have a Katana as you do .
I might try putting my gluonpilot board to this plane.
Did you do any modifications to Tom's code or you used latest firmware?

Which Hitec receiver did you modify to pull out PPM signal?

[Mitch]

I used a Hitech Electron 6 that I had disected quite some time ago. Instructions are here http://paparazzi.enac.fr/wiki/Other_Hardware#Futaba_FASST_7-channel_receiver except rather than run a separate wire outside of the case, I severed the pwm trace to the channel 6 connector, and jumpered the ppm signal to that pin. Thus the pwm signals for channels 1-5 are still intact and the channel 6 connector functions as gnd,pwr, and ppm. Made for a very clean setup.

I used version .3 firmware directly from Tom's hex code and gluonconfig .3 = worked perfectly and not a glitch!!!

A few notes -
I calibrated the sensors with the esc bec power rather than the external pack I used for testing
I leveled the airplane by blocking up the tail when I calibrated the sensors, rather than let it sit three point. I was curious to see how the three point attitude during boot would affect things - worked fine

[Tom]

Fantastic!!!

Glad to see you did everything "by the book" and that it is working so fine!
Next step is to try navigation, but the circle navigation around the home position will also work fine.

My next step is to make the navigation programmable from the config tool and a gcs

But you can try to break my height record in navigation mode (current version allows you to control the pitch angle with the stick in navigation mode)

If you have any input about the documentation wiki, please let me know!

[Tom]

It's also a delight to see your clean installation that shows how simple it *can* be

And you see to have good nerves: first stabilized mode flying 320meters away from your home position

[Mitch]

I was still paranoid. I only went as far as I could see it well enough to do a manual bailout if I had to!

[Mitch]

Tom,

I really didn't use the wiki - mostly I used the code and circuit diagram, but I just reviewed it a bit. Lucasz just had some very pertinent questions that I think could be clarified on the wiki. I think that clarification of all of the connections and jumpers for a "basic installation" would be helpful so one didn't have to refer to the circuit diagram. I have printed out a photo of the board and marked it up with notes over the past few months to pick up where the silk screen overlay can't be specific.