Hi All,

I'm doing my walk around testing, I thought I might list the sequence/experiences that I had. I have MatrixNav (Red, 1.7) loaded but referred to the Aileron assist document as it describes the initial set up and where the servos plug in a little better. My LED on the EM406 glows a very faint red while powered up, I would like to know if others have observed this, it still seems to operate correctly.

Firstly, I set up my 3 position flap switch (JR 2720 radio) as follows:
Position 'Norm': Flaps at 25% down
Position 'Mid': Flap at 10% up
Position "Land': Flap at 100% up

The trims on my transmitter are digital and therefore awkward to move full left or right, the flap switch works a treat.

If you power up in an electrically noisy environment i.e. next to my PC then the green LED will come on and after a few seconds the servos will go crazy. The green LED reflects a TX signal (valid or otherwise).

If you power on in a quiet environment, then green LED stays off and the servos don't move. Switching on the TX results in the green LED coming on and the servos responding correctly.

The red should flash for about a second and then go off. On GPS lock the aileron servo will move back and forth seven times and then it seems set. This takes around 30s if the GPS has been off for awhile. If the GPS does not have a lock (i.e. you are indoors), the mode switch is ineffective.

In a 'normal' mode the green LED is on and the red off, in assist mode the green is on and the red is on, in RTL mode the red flashes and the
What I did find that I cannot explain, the switches do not change the direction of the servos! I tried moving the switches while powered up and after resetting the board but the directions didn't change. I haven't tried changing the gains to a negative value.

Once, after switching off the transmitter and switching it on again, the board would not return from RTL mode! I had to reset the board to get it working again, I haven't been able to duplicate that and so I have put it down to something that I did.

Manual stick movements will still move the servos even in RTL mode. I am not sure if this is correct?

The RTL is fairly untested as I didn't walk that far and every step I took resulted in some servo jitter. I think it was behaving correctly but I will try later at the field where there is more space.

I hope the above is useful to the first timers.


Views: 246

Reply to This

Replies to This Discussion

Hi Bill,

Quick reply, it is a three position switch.

I'll try seting the plane up in stand where I can measure the attitude on reset. Trying different angles I'll be able to measure the results. As it is I set it up on a stand and hold it about as level as I can see. The plane flies with a 'nose up' attitude and by placing one finger under the fuselage it seems about right in the air. I read about disabling the attitude hold this afternoon but I still had good results.... Must be good software!

To add the throttle cable will require a little surgery but I'll do it for the weekend. We intend flying the T3 course, I have a friend who has a GPS logger and we can log the flights.


Maybe you and I are getting closer to the "same page". Also, I just realized that I was not clear enough about (at least) one thing.

The idea behind the zeroing of the accelerometers (and the gyros) on power up, is that the firmware can remember the power up attitude. So, I was not 100% accurate when I said you should have the plane perfectly level when you power up. It is more accurate to say that you should orient the plane the way you want it for zero trim. In your case, you mentioned that you want the nose one finger up. That being the case, when you power it up, the nose should be one finger up.

What the firmware does is record the trim settings and the attitude of the plane on power up, and uses both of them as a baseline for in flight calculations. When the joysticks are in the neutral position, and the trim settings on the transmitter are the same as those on power up, the firmware will attempt to maintain the attitude in the power up position.

Best regards,
Hi All,

I’ve flown some more this weekend and must say that I am having a great deal of joy!

This project has worked flawlessly since I started, given that I didn’t always quite understand what I was doing! I flew Saturday with ‘Altitude hold’ but as there was a nearby political rally I couldn’t really hear the motor coming on and off (these rallies usually have a great deal of singing). One part that I did notice is that on the upwind leg, the motor would come on, the plane would climb until the motor shut off and then the wind would blow it back while it lost height, it took quite some time to make the upwind leg (I need a faster plane).

Sunday allowed me to fly in fairly good conditions and the performance was great, I can’t measure it but the altitude hold was constant within a few meters with the motor being ‘proportional’ in some legs. Programming new waypoints was dead easy and worked flawlessly (if you remember to build the project each time!)

My points of concern are this:

The turns remain fairly aggressive, some were good though. I estimate the turning radius at about three meters. The plane would sometimes almost form a spiral dive turning about 270-300 deg’s. in the process, it would then continue in a very large arc to the next waypoint (I was flying a triangle with equal 100m legs) and then repeat itself, the resulting path looking much like a clover leaf (with twisted turning points?). Deviation was around 30m from the desired path at the mid point. I tried upping the YAWKP (0.2) constant later and it didn’t seem to improve this waypoint tracking much. One thing that has confused me since the beginning is that the YAWKP refers to the proportional feedback (OK) but YAWKD is not the differential term of the same loop. It was initially confusing to me, I know that I am being pedantic but it might confuse others too. Upping the YAWKD to 0.25 would cause the rudder to oscillate not be more damped.

It is quite possible that you will simply watch your plane flying along by itself and not remember to keep a check on the flight time! The battery went flat twice with me and I had to quickly switch to manual mode and fly it back. With my airframe, closing the throttle for RTL mode would result in a very long walk. You notice the plane getting lower and lower before you realize what’s happening.

We really need some of the clever ‘chaps’ on this forum to port Ardustation code to UAVDev or perhaps another. I wish I was clever enough to do it. Having telemetry to be able to quantify the results would be great. Having something that would transmit the battery voltage would be fantastic! The new version (I forget who’s writing it) would open so many new possibilities especially the ability for flying wings, can’t wait.

Thanks to Billu Bhaiya and the rest for such a robust and well thought out project, not once have I powered it up or flashed it, built a project that it hasn’t simply worked. I have done something that I have dreamt about since childhood.

I just read Greg's latest post. Is it true that you need to rebuild or make the project every time you make changes to waypoints (or the options.h file)?
Hi Bryan,

The waypoints are compiled along with the code, the compiler counts them up and puts them into program memory space (not RAM), same goes with the options.

Hot dog!

Ben has done the code already and it's all here: http://groups.google.com/group/uavdevboard
Its true you need to reprogram every time you make changes to the waypoints, but it does not take long, it only recompiles the waypoints.
The waypoints are stored in program space, so you can have plenty of them, on the order of 1000.

By the way, because the waypoints are measured relative to the power-up point, the waypoint pattern is portable. Once you program it in, you can use it anywhere you fly, without having to reprogram the board. This has some advantages.

For example, I fly at a local park that has football, soccer, and baseball fields that might be in use. I can make a last minute decision where to fly at the park without even touching the board. Also, there are 4 different fields where I can fly. I can use the same program at any of the fields.

Another advantage of this approach is that it kept Greg's plane from flying away the first time he engaged the waypoint function with the pattern that came with the firmware. If I had used absolute coordinates, his plane would have headed out to my field. With relative coordinates, it flew around in his neighborhood instead.

Best regards,

The link you gave is to the discussion group.

There is a separate link to the project: http://code.google.com/p/gentlenav/

Yes You r right billu bhaiya but it will bw much better to add a small application which sends waypoints to dspic in meters... and we can add few lines in already existing interrupt loop of serial , it will get data and save in eeprom ..
in waypoint file it reads waypoint data from eeprom ..
It may help to change the waypoint on the fly itself ..
and can add few more features once two way telemetry works up

Reply to Discussion



Season Two of the Trust Time Trial (T3) Contest 
A list of all T3 contests is here. The current round, the Vertical Horizontal one, is here

© 2019   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service