According to the Mission Planner logs, the outputs of motors 3 and 4 are double that of motors 1 and 2. Consequently, motors 3 and 4 quickly overheat.
I'm trying to get my new multi dialed-in but after only 5 minutes of hover attempts [at about 45% throttle] both of my CW-rotating motors (3 & 4) get so hot you can't touch them, while the CCW motors remain cool. All calibrations have been re re re re re-run. ESCs, radio, accelerometers, compass, and everything appears to be balanced — fore-aft CG, props, etc. — yaw seems stable, I even disassembled the motors, balanced the bells, lubricated the shafts and bearings with EPIC oil for electrics (note that these motors are new and have less than 15 minutes run time on them).
To make matters worse, This is only the third real attempt to get a stable hover but I can't quite get there from here. I still require manual trim input on the Tx of about 5% right roll and 5% up pitch for it to behave as well as it does in the video. I'd love for it to be rock steady but I'm not sure know how to get there and with two overheating motors there's no way I'm going to attempt to fly outdoors.
I have an email in to HK tech support but no reply yet.
If anyone is interested the video is here and sample log files are here:
2014-03-08_2015-17-29_2062807.zip
For safety, the skids are covered in foam pipe insulation and a prop protector stick was added to each arm (the sticks are from 2 carbon fiber arrows -- $3.50 each at Walmart).
And yes, the props are right-side up =)
Specs
===================================
Frame: Tarot Iron Man 650
Arducopter: 3.1.2
Props: APC 12x4.5
Motors: NTM Prop Drive 28-30S 800KV
ESCs: Turnigy Plush 40A
Battery: Turnigy 40C 5A 4S
Flight controller: APM 2.6 w/uBlox GPS
Receiver: FrSky D8R-II PLUS (modified antennas)
Transmitter: ER9X with FrSky ACCST 8CH module
Weight as shown: 2.4kg
Thanks!
My current PID settings:
Replies
I think I'm making progress. Adjust the PIDs upwards is bringing motors outputs together. Went from a P of .08 to .2+ and that helped enormously. New problem: I'm holding the bird while it's at 60% throttle if if I move it too aggressively the power suddenly drops to idle and won't recover.
My Tx is a Turnigy 9X that I converted to ER9X via the SmartyParts kit. Pretty sweet. Yeah, pretty sure it's straight airplane mode. I don't recall seeing any mixes in the sticks.
Ha! There is another guy on these forums who had the exact same issue last year. He didn't post a solution so I emailed him. No reply yet. Also found several people through Google -- all without posting solutions.
hi,
i had the same problem, see my log : http://www.diydrones.com/forum/topics/need-help-on-a-thrust-issue-l...
my thrin and out were really different, and i finnally link my issue to yours by displaying the motors in log and see that 1 and 2 were spinning slower so :
i first thought that it was a radio cal issue, then a code issue, then motors issue, then props issue, then esc sync issue.
i did compasmot, accel calib, radio calib , live calib, check vibration, check magfield, check hdop, downgrade from 3.1.2 to 2.9, a dozen of times.
so from the beginning of my test i upgrade motors, esc and finally props to quality one (from rc timer to tigermotors).
i only have the motor spinning at the same level before i do auto trim flight, after that it s always different.
so i think it s a bad cg (despite the fact that it s visually leveled).
In Order to keep enough power i finally upgrade from 17" prop to 18", after autotrim i still have 1 and 2 spinning slower but i now have enough power to conteract a fast descent.
anyway i stopped my build after a dozen of crash and rebuild cause i can't tune it (maybe too low kv and too large props).
Wow. That sucks. And your outputs look nearly identical to mine. I've done everything you have including balancing the CG. What frame are you using? If it was in your post I missed it.
I think R.D. Starwalt is on to something. Another important note regarding his post is that the front direction is determined by the apm board so if the board is rotated 90 degrees to what you would call the front of the quad based on looks, then that will have the same effect. This would normally cause obvious control issues but you can probably get it to appear to respond normally to inputs by reversing channels in your transmitter.
Just checked -- the wiring is correct. Actually wish it weren't.
Hmmm... Both the APM and GPS receiver point forward and my Tx is a typical mode 2 setup. The copter responds the way you'd expect -- forward stick equals forward pitch, left equals left roll, and yaw rotation is correct. Aside from a few terrifying moments of playing chicken with the walls of my garage it's actually pretty stable. I'm going to trace the APM outputs now.
it s an homemade frame with aluminium square tube, on this picture it s the old motors and props.
i attached a log of my crash today if you want to have a look.
2014-03-11 13-29.log
Pretty! I like the clean lines. Looks like a serious thrust monster. I'll take a look at your log. Did you have any issues with the frame flexing?
not so much, i blocked each end of arm and put 5 kilos on the center and it moves of 2.5 mm.
-
1
-
2
-
3
-
4
-
5
of 6 Next