Special thanks to Paul Dowling and Jason Short for their help with PID's tunning! I would have never taken off without their advice.
Both stabilize and simple mode work flawlessly.
The beast is flying with four 4 cells batteries, 400 gram each. wingspan is 120 cm. Propellers are 14 X 7 3 blades.
Total flight weight: HEAVY.
Comments
I can see the tricky take offs with higher RATE_I values, but it corrects faster against wind!
Careful with rate_i it will also make your take off tricky and tend to dip on one side. try to keep it as low as possible. Also high value can cause your quad to oscillate. Its the balance of PI that we want it equates to the balance of sluggish and oscillating quad. Its tricky to get and I myself dont have the privilege to use the tuning as I seldom have windless conditions here so I do it trial and error.
With a high I value, I think that the correction happens faster and more aggressively. It is a multiplier as I understand it.
Read the flight modes in the manual, they have expanded that chapter a bit, i learned that RATE I takes care of wind.
''Rate_I is used to compensate for outside forces that would make your copter not maintain the desired rate. A high I term will ramp quickly to hold the desired rate, and will ramp down quickly to avoid overshoot. A low I term will take too long to have an effect and may cause undesirable overshoot. It's better to have an I-term of 0 than a poorly chosen I term.''
I guess if your I is too high then it overshoots?
I think the I-Max is the maximum you allow the I value to scale to. So I suppose if you got a gust of wind, the I value would start to increase to overcome the wind, but it will only increase to the maximum value. I have absolutely no clue what in-flight effect this has. I haven't really played with that value yet. I left it at the default from APM 1.55 (old revision) and haven't been brave enough to change it. When I finally get out and fly 2.0.45, I plan on using the default value that comes with that revision. I will try to scale back my stabilize P values and see what happens! Thanks!
having my stabilize roll and pitch at 2.2 makes me think that lowering those values would get rid of the wobbling. My values were about 3.3 and the model was wobbling; lowering those worked out.
I have a question for you: by lowering your 'max' to 1000, what do you change?
Looks really stable! Wow. I have a very heavy Octo setup. I am wondering if the general rule of thumb for larger multirotors is lower PI values. I was flying with mine at:
Stabilization PI
p 3.84
i .024
max 1000
Rate PI
p .105
i 0
max 1500
and I get a large wobble when descending. Any advice?
Ohh I meant scary landing on them during bad landings, I have bad crashes where one arm always hits the ground.
Not to sure what its the same, actual those are my exact values until v41 for my frame but I changed my mount now even leans forward, its not the same as in my blog about the first vid with gopro. I just balance by hand and seems to do the trick of course not perfect and not for pro use but enought to satisfy my noobness for now :D
My AUW is 1.5kg, 10x5 blades (now but 10x45 (jdrones composite) blades when I had the same values as yours)
Jeff:
I'm not landing on the batteries of course, I have this metal rods as legs instead. they are shouting to be replaced by some kind of carbon fiber rod.
So you have similar values? This quad is flying at 4 kilos and a half... how can the values be similar? I'm confused.
I have seen your video of the gopro on the nose and a battery on the tail. Is your balance ok?
I thought so, scary landing gear :)
Interesting PI values, we have basically similar results but mine is a fraction of the size and weight but I have a gopro mounted in front almost aligned to the props and a battery sticking at the backside. Now I found my logic is pretty close on how thing work (still clueless).
Hope to see more vids, with that smoothness its looking to be a good AP platform, just concerned about the weight penalty.
Good job thanks for sharing
-
1
-
2
of 2 Next