Version 2.6 of the ArduCopter code is now available in the AP Mission Planner and in the downloads area!
Updates to MavLink 1.0 means you will need to use "ArdupilotMegaPlanner10.exe" to connect. If you've updated your mission planner recently you should find this executable in the directory where the mission planner is installed.
The above video is done using a prototype 3dr ublox GPS which seems to have better accuracy than the standard mediatek.
Improvements over 2.5.5 include:
- MavLink 1.0 support (use with ArdupilotMegaPlanner10.exe) [Tridge, Craig]
- Stability improvements especially during level hover [Jason]
- throttle range improvement (higher min and max) [Jason]
- improved standard Loiter PIDs [Alan, Heino, Jason, Angel]
- dataflash erase speed up ('+' messages removed but it only takes 6 seconds now) [Tridge]
- Copter LEDs [Robert Lefebvre]
- RTL loiter stage target set to home to improve final landing position [Jason]
- flip & acro improvements [Jason]
- circle mode target improvement for ground station [Jason]
- Auto Approach [Adam Rivera / Marco]
Bug fixes include:
- UBLOX driver fixes (lock should now be more reliable) [Tridge]
- enable mavlink messages during dataflash erase which resolves issue in which new APMs fresh from the factory appeared unresponsive [Tridge]
- proper printing of lat/lon values in dataflash logs [Randy]
- removed duplicate GPS reads [Jason]
- resolve flooding of telemetry link with low-battery warnings [Tridge]
- RTL bug would land if rtl_approach_alt was more than 1 [Jason]
- WP Radius could not be set larger than 1.3m [Jason/Randy]
PIDs are optimised for the 3DR/Jdrones quad with 850 motors and 10" props. If you're using more powerful motors/props and are seeing bad flight behaviour in stabilize, start by turning down Rate Roll P in 25% steps.
This time we spent some time optimising the loiter PIDs. Tuning loiter can be tricky so please refer to the discussions which will appear below for more community feedback on what parameters work best.
All feedback welcome below. Enhancement requests and bug reports can be put into the arducopter issues list. When possible please include logs (tlog and/or dataflash) and tell us whether you're using APM1 or APM2 and what version of the software you're using (presumably 2.6 but tell us anyway!).
this video is STABILZED mode NOT acro
hehe, yes, it does not make sense, but it does :)
Flipping in ACRO mode mean that the motors/props are always spinning, but it's full manual control with NO stabilization, thus a lot harder to fly for most people, that also means you can do full manual rolls and loops, BUT when you flip a copter with spinning motors (unlike a variable pitch heli) as soon as it's upside down it want to move tawards the ground veen faster than gravity will let us free fall as it's powered, with STABILIZE bode it's just a lot easiers..
I simply pump the throttle and then hard stick right/left/forward or back aat the same time (like half a second before) I can the power fully, thus the momentum of the roll makes the copter flip, I then wait for it to do the full loop and then up throttle again, as soon as the props start spinning the APM 'catches' the quad as STAB mode is again activated, I basically 'cheat' the STAB mode, as with no powere the is no STAB mode even though I'm flying in STAB mode. Get it ?
It's FAR EASIER to do flips like this than in ACRO mode. As you can see even with a huge octo, you just need enough momentum to flip it.
It's simply a matter of rolling or pitching hard on full throttle then as the frame approaches full tilt (40 degrees in stab mode) killing the throttle for a second. With no throttle applied the APM will not stabilise anything, and the momentum of the hard roll/pitch will carry it over for a flip. Then simply power up again and the APM will stabilise the frame level again. Doing this in Acro is a bit diffferent as you have to be more precise with the timing and ready with roll and pitch to bring the frame back in shape.
My favourite is a tri copter front flip, it looks great. I'd do a video, if it wasn't for the severe weather warning...
Dave, so you have a Tri? Do you have a problem with yaw oscillation? Want to test something for me? I can give you something to patch in, pretty safe and simple to start with. Basically creating a dead-band between the yaw angle controller and the yaw rate controller, so that if the yaw angle error is less than 2°, it just tell the yaw rate controller "Zero", so that it's not constantly sawing back and forth on that yaw servo.
I've tested it on my heli as a heli-only option, but I'm thinking of making it compile for heli and tri.
Just got in, am setting up the tri with 2.6 now, I like the sound of the deadband - as i mentioned I have a really poor analogue tail servo so its a great test bed, i.e. if it helps lock the tail of this tri, it should work with pretty much anything.
Send me the patch over, I'll tri with and without.
I've just had a play with the tri 2.6 in the lounge, concentrating on yaw.
At stock settings stab_yaw_p 7, rate_yaw_P 1.3. I get serious (90 degree) oscillations but only after a yaw input. Say I'd yaw 90 degrees and let go the nose would bounce back at least 45, then back again, slowly reducing until it settles at the mid point.
So I chnl6 tune stab_yaw_p, the lower i went the better it got, I have ended up at 3, this gives me quite slow yaw, not too bad though, but no bounce back, it just turns on input and stops where it is when i let go. I've yet to try this in wind though, I fear i may need more to stop the tail being blown around. We'll see when it stops raining.
Robert you mentioned that stab_yaw basically controls the rate_yaw at the moment, could you elaborate a little please? Once I'd settled on my low stab_yaw_p, I tried tuning rate_yaw_p but found anything below 1.3 slowed it down, and anything above made it unpredicatable.
Conclusions so far, I have to lower stab_yaw_p to 3 from 7 on my tri (with a shoddy tail servo) to stop bounce-back oscillations after a yaw input. This does have the effect of slowing the turning rate down, could this be seperated? And also I wonder if I will be able to increase my stab_yaw_p once my new superquick digital servo arrives.
I'd include some logs, but mine don't seem to be working (i got two logs from 2 batteries, each one only going as far as APM initialisation.)
EDIT: I should mention that i tried all of this fiddling about with and without the compass enabled. Results the same.
Good news, I've found a fast digital servo for my tri in a box of bits!
Results however are exactly the same, at Stab_yaw_p of 3, the tri's tail is very obedient, but very slow. At 7, it is almost unusable, ask the tri to turn to the west from north and sometimes it will bounce from south to north almost before settling
Is there a way that i can increase the rate, but keep the stab_p lower?
@ R_lefebvre - could you send me over the dead-band patch please, I'd like to see if it has an influence.
You really tested the robustness of 2.6 :D nice flights. Also congratulations on your machines. MK and DJI repent your sins and fear the AC ehehehh
HAHA, yip, lol my aim ;) MK/DJI look out, opensource bunch of 'noobs' will soon take over (in my opinion already have)
Duran, thank you so much for these videos! They really show off the systems, and give the Devs something to be proud of. Sometimes it seems like all we hear are the problems. But this shows what the system is capable of in the right hands.