Successful one-hand flying

[Honk]

First successful flight with only one hand so I can finally have a beer in the other!

http://www.youtube.com/watch?v=vmXNU-FY8to

Now I'm waiting for some carbon fiber square tube to come so I can make a much needed new frame.

[RCvertt]

First successful flight with only one hand so I can finally have a beer in the other!

http://www.youtube.com/watch?v=vmXNU-FY8to

Now I'm waiting for some carbon fiber square tube to come so I can make a much needed new frame.

Very nice. Just goes to show how stable it will be if the accel gets integrated.

[Boolander]

Nice! Are you flying mode 1 or 2? Airplane or heli?

[Arthur P.]

Good work !

Does look as if this was in evening twilight and low/no wind conditions. So it might bounce a bit more if there is more wind. It might be worth defining a minimum throttle value when in altitude hold mode, this could reduce dropping too much / too fast and thus help stay closer to target.

[Honk]

RCvertt, Yes, I'm quite impressed of the BMP085 barometer now even though I have not tried another one yet. But you know, this is far from the potential it provides. My filtering is slow/bad and the P gain is not correctly trimmed (too high as you can see on the oscillations) and the I term is not yet used. I think that will reduce the bad effects a little.

Boolander, don't know what mode it is, but you can see my left thumb in the air, so I'm doing Roll/Pitch with my right thumb. Airplane mode is the one reluctantly used since I had to have the 2 extra channels individually controlled, in heli mode they were relying on like throttle input and such, but in heli mode I could also do a throttle curve that I can't have in airplane mode, but none of that will matter when I have successfully coded the throttle channel (now linear) will regulate vertical speed instead of motor PWM.

Arthur P. , well you make a good point. Do you think wind affect climb/fall speed/variations much? I have not been able to monitor that proportion very good. A minimum throttle value is a good idea, but I have defined such a thing so i'll just have to trim it. The barometer is not really giving absolute throttle commands now, but only an offset, but I can change my "windup guard" to not go as far down as -150 in motor PWM, but then I risk the quad of going upwards instead... Actually the next approach will be based on velocity instead.

I will simply do this: if the throttle stick is in the middle and alit-mode activated:

*0 value directly from Throttle channel to motors
*strive for 0 value from accelerometer if throttle stick is in the middle (1500us + offset)
*strive for 0 pressure delta if throttle stick is in the middle (1500us + offset)

else if throttle is above/below threshold, regulate vertical velocity (like 2000us = +1m/s, 1000us = -1m/s, linear proportion)

Simple, but not fool-proof

Thanks for the responses!

[Arthur P.]

With the MK there is this "dynamic stability" setting which can be used to reduce the effect of wind gusts. Normallly, just like heliis, multikopters balloon up in wind. This is also noticeable when you land. With the correct setting that can be supressed to be near neutral or even to the level of actually dropping a bit. I never did dig into the code to figure out how this was done. Most likely through the Z-ACC responses. I.e. if the throttle is not changed then there should not normally be a sudden change in vertical velocity. If it does occur, then most likely this is wind effect and the FC should adjust the throttle slightly to negate the effect and stick as close as possible to the Z-ACC measured prior to the effect as long as the throttle stick has not been moved.

[Mikro]

Woo hoo! Promising start! I tested the BMP085 baro on my proto AQ v2 board by hand, just moving it at an arms length is pretty exciting to see discernable changes. I can't wait to start playing with this sensor with you!

[Arthur P.]

Sorry Mikro, you do have to wait: GO ON VACATION (and ENJOY). This is doctor's prescription !!! Gotta keep you healthy ;=))

[Honk]

Nice, I can't wait either to get this work forward, probably nicer interactions/implementations when you look over it Mikro! The structure of the original Wii code is kinda messed up now since I've played so much with the code back and forth. I noticed the RAW value was changing like 10-20 pretty quickly when looking at them as I moved it up and down in the air, but then I added filter to it and woop!, it went from so-so to really good! Now it seems to have more accuracy than the datasheet actually states, but it may be an illusion. You can see some graphs in the BMP thread, latest test result shows very accurate dynamic reading when moving it down 1m by hand, using LesNewell's IIR filter implementation that he described to me.

But yeah, enjoy your vacation first! I took "vacation" today by visiting a friend at work, hehe...

Arthur, yes, that is what I intend to do: If throttle stick is not moved, the quad should not move basically. This slow yet after a while oscillating regulation I am sure to help with Z accelerometer reading, maybe not as refined as Mikrokopter, but that's the point, everyone can pitch in so it will not be up to one single person even though it's fun working