I am facing a few of issues with my drone -
i) Motor heating - all 4 motors are getting quite hot
ii) Possible yaw imbalance
Now the fc (pixhawk 2.4.8) and its sensors are probably cheaper-side only but I think it has enough ability to atleast fly a normal quadrotor properly.
Possible causes which might be causing this according to me:
ii) Summer Heat
iii) worn motor since they are clones of DJI 920KV
iv) Low voltage (3s)
fixes I am thinking of:
i) Autotune (i still haven’t run it because of heating)
ii) Replacing motors with EMAX 2213 935kv and props from 9450 to 8045/1045 depending on cell count
iii) changing to 4s
Now I have no idea which one of these might fix it enough so that I can comfortably a fly a full-charged pack without worrying about everything or having to land to check motor temps in between. the PIDs (at defaults) are likely not causing motor heating as earlier same PIDs there was no heating (maybe my entire analysis is flawed here). I fly in indoor lobby which is approx 2m wide, 5-6m long and 2-2.5m high. Earlier I had tried some manual tuning but I couldn’t get to better performance. log file
Perhaps it may never fly worry-free because of the quality of the parts.
What changes/steps should be taken to fix these issues and if any motor replacement, is the EMAX 935KV a good choice? any other option in this price range (approx 1200-1400 INR per motor)?
notch filters are tuned and you had helped me with these values and they seemed to be effective enough.
I think even if the PID tuning steps reduces motor heating and improves stability I’ll always be in the fear that these motors would fail anytime probably because of poor windings and bearings. I would rather replace the motors once and be confident that they are reliable rather than having motors which ‘just work’.
I think the title indicates your problem.
What is the size of your drone, somewhere I read propsize 9.7 inch. So this is not a small drone for such a small tunnel. The propwash of your drone causes lots of turbulences by the short distance to the walls and ground and your moreless untuned drone reacts on it with high frequency change of motor inputs.
So before investing more money in this built try to analyze the true root cause.
Follow the advices given before in your other threads.
I have flown outside open area as well and heating persists some time ago.
Propsize is 9.4 inches on f450 (also not the best choice)
the default pids were of the 3dr iris IIRC and it was also a similar weight and similar motor-prop platform so PIDs should not be that far off.
Moreover, in same conditions, heating was not there earlier and such a tremendous increase cannot be because of PIDs since they have always been the same.
@RudrakshTek , I think you have the same motors. What is your experience with them? Vibrations, thrust, etc? thanks
He also ordered parts from the Hardware To Avoid store…
That log shows it’s a bit underpowered/overweight and the Vertical Acceleration Controller gains are too high. Hover thrust looks to ~0.35. Take some weight out of it (smaller battery?) refactor the hover thrust based parameters, reset the notch and take it outside and run Auto Tune.
And I’m not sure if it’s an artifact of the Gyro sample rate difference but your external IMU looks like crap.
One example:
I dont think its possible to take weight off without removing something entirely, like gps, or gimbal.
Maybe I can go to 4s.
Also, according to datasheet, the current motors give 500gm on 3s, and 980gms on 4s. the emax ones say 860gm on 4s and 670gm on 3s. the weight of the quad is ~960gm w/o lipo and around 1160-1170gms w/ lipo. then the twrs should be:
Motor
3S
4S
DJI CLONE 920KV
1.78
3.5
EMAX 935KV
2.3
2.9
theres a decent chance that the clone motor's ratings are inflated. I also suspect damaged bearings since they stop quite quickly after I let give them a spin and they may be showing gritty noises (I am not fully sure) and the bearing area is also the hottest part of the motor.
That statement was in context with the the Hover thrust change from a weight change. And there is no way to know if anything is wrong with it anyway as logging is disabled.
A few months ago, I made a drone with these parts:
A2212
F330 frame (not the extremely low-quality junk, it seems)
Emax Bullet 20/25A ESCs which I flashed 48 Khz Bluejay on
2.4.8
Currently using 8038 props
AUW about 764g
And it flies nice indoors and outdoors.
I did run AutoTune initially, but I think the winds were more than acceptable, so I did a quiktune indoors with a bit of manual tinkering to get it tighter. So I am running with the stabilize rates suggested by autotune and the PIDs after quiktune and manual tuning. All that remains is a bit of tighter yaw tuning and feed-forward. A lot of manual tuning was for the position controller as I use optical flow.
Point is, I have a worse quality motor than you and don’t have much problem with heat. I haven’t done aggressive maneuvres yet but they only get very slightly warm after a basic flight outdoors. Hover thrust is about 0.4.
So I would say it might be that your tune is off after the bit of changes you made. Didn’t check your log though.
Your drone’s each motor generates around 600gm of thrust and thus the TWR is quite good and this reduces motor load. Your setup gives more thrust than my motor does. but then how is your hover thrust so high?
In my case to increase thrust only choice is to go to 4s, which gives quite high twr. but then, is there some chance of improvement with it?
The props! 3.8" pitch. Might not be 600gm with this combination; I don’t know how much. I think I got 0.27 - 0.29 with 9047. Also after a few crashes, there is a bit of a yaw imbalance problem but still flies fine because I haven’t flown truly aggressive manuevers where you go full throttle with some attitude changes. In addition to maybe some motors needing more thrust due to the low quality making it age faster?
Also, I checked your logs. The gains are a bit high. Also any yaw problems?
What ESCs are you using? No BDShot?
Depends on today’s conditions and the exact location in this country. In fact, with the onset of monsoon, my location is back to normal summer temps.
Damn, that’s hard. Never experienced that much. Peak here in Kerala now is about 30 deg C with pre-monsoon.
not too much, little bit like sometimes it yaws little bit randomly then corrects itself, probably some PID issue.
ESCs do not support BDshot unfortunately, even though are blheli (apparently the old 8bit type).
I also noticed that either temp change has made the current shunt calibration inaccurate, but some earlier logs it was around 15-16a and now it seems to be 25-30a approx so I dont why it increased. Can such temperature cause this?
If I go outdoors during afternoon, I might record 60c on the IMUs (45-48 ambient)
Which model? You can probably flash Bluejay on the BLHeli_S.
In your current log, it is as much.
Yeah maybe PID. Could also be interfence. You can probably do magfit, even if indoors. Just make sure to yaw around a few times and do some normal flying around. You need atleast 30% coverage though and this might make it fly better than it currently does.
Direct sunlight sure makes it go that high even here if the autopilot is kept still without much air circulation like when I did IMU tempcal, speaking of which, it might bew a good idea you also ensure that you tempcal covers this range.
temp cal is around 54c on external, 40 or 45 on internal. But, if in any case the Imu crosses this range, will ardupilot extrapolate the curve or just clip the compensation at the max recorded?
Is it a good idea indoors? The buildings have big iron pillars which might affect the readings and lack of gps fix. and even then, I dont think I can even yaw 20degree comfortably in that lobby.
