Just loaded 2.0.33 into my AC (stock frame, well beaten- held together with tape, ragged props, no balancing, no antivibration, old board with no jumpers) and I cannot believe how far Jason has come! Stable mode flew great, no yaw probs, stayed flat and hovered well. I then did sonar alt hold, worked great at 1m and 2m - held without rising or dropping, kept within a few inches. I then tried baro height, hung well +- a meter. I even tried the former 'forbidden zone' right where it switches to baro (used to be I could never predict what happens at about 3m) and it handled it very well. I then loitered in my tiny front yard at 1m and it hung right in a 1m square (light wind) for a long time. Loiter at baro height similiar. No more NG for me!
Even better, I set a small mission (2 waypoints) at our local small park, screwed up my courage, and watched it actually do it!! It was fairly close to the plan, but a little close to the trees so I changed it to a simple takoff to 8m, loiter and land. Watching it repeat three times was amazing! One was a perfect thing of beauty as it loitered witin a tiny space at the exact right altitude and then landed right on target. All landed about the same place, some with a little searching that was fun to watch.
I think Jason has proven that it's not the processor but the programmer that really matters! Running short on code space is understandable, and I am on board with the 2650. This is a cheap and easy upgrade thanks to Arduino standards, and I can't wait to see what comes next!
Replies
John Hanson:
Hello just wanted to ask if you tried RTL also... I'm trying 2.0.33 and wonder if should go to 35 or stay here (33) doing more testing... which version do you prefer the most?
I was kinda hesitating to try that out.
Could you tell me a bit more about the auto take-off and landing mission ? Did you just put the copter there, switch it to auto and gave a little throttle ? Or what's the correct procedure to start it ?
Does it shut down its engines when auto-landed, or just stays at min throttle ?
Thx for any info !
can you share your motor type,prop, frame, weight of your QC and PID file parameter you use?
i hope you understands my bad language :)
Very nice news. I think that the community has done a great testing job, and with the help of talentuous programmers the project has rised to the Release candidate level.
Yes the processor is not the most important part. Smart programming , opening and board reliability is certainly more important than processor pure power. The wiki completeness with good assembling informations and recommandations is an important key of success as well.
I'm sure that perfection will come in the next weeks with more users testing as soon as the Release candidate will be out.