I got an issue with HD camera and bigger props and possibly tuning. I have a 2.5 kg copter with Sunnysky 3508-700 kV motors on 4S and 15 x 5.5 props. The camera records almost perfect video during climb and hover, virtually without jello (if you look closely, from time to time you may notice something, but it is virtually invisible). However, once I have it in descent, especially in ACRO mode, the video becomes with very noticeable jello, to the point that such footage is unusable. I.e. if hover power is 210 W, and I descend at 120 W, with 70 km/h and -12 m/s, the HD video is very bad. Horizontal flight at 70, even 90 km/h is very good at 300.. 500 W.
I once swapped props for 13 x 5, and the quality improved considerably (needed retuning), but the problem is that the 15 inch prop provide 20 % more hover time.
I am not sure if it is a problem of tuning the copter or something else and how to solve it.
Below is a screen shot of the descent in acro mode. The oscillations in pitch are around 0.1 degrees and in roll around 0.2 degrees with a frequency of 3 Hz.
I tried to make a double suspension (see photo), with the idea of putting a small battery (220 grams) as a stabilizing mass, i.e. the first antivibration mount would have a short camera boom (80 grams) and a battery (220 grams) as stabilizer, and then the camera itself (90 grams) with its own antivibration mount. I managed to do just a test flight without battery, but the results were disappointing. Overall, the jello increased. On hover and slow flight it is noticeable, and during fast descent it was obvious even in goggles, i.e. overall, it became considerable worse (there is one point, that the boom now is much longer to clear the prop view, so if one would discount the boom effect, I would say that the vibrations in the HD footage could be similar to what was before). I doubt that the battery would improve significantly the situation and I will do a test tomorrow.
I would appreciate any suggestions, i.e. should I try tune the copter better for flight, should I somehow tweek some parameters to improve the video vibrations, is there a better design for the mounting of the camera?
The best approach is to attach the camera to a large, isolated mass. If you’re strapping a battery under that frame. think of isolating the battery on a tray, and attaching the camera to the battery isolation. It’s worked wonders for me in the past, and it’s usually pretty simple to design.
Yes, now I see my mistake. My original thinking was to use a big mass, but the camera antivibration mount is not able to handle heavy weight. So I made a secondary antivibration mount, but left the original antivibration mount. The battery in the photo is not the main battery, by the way. The main battery is 1.2 kg.
Probably I will eliminate the smaller antivibration mount altogether, and make a fixed boom on which the antivibration (the bigger one) will be installed
Well, I removed the small original vibration mount, added 150 grams stabilizing weight, did a test flight, it is exactly the same as was… Jello on descent, good image on climb, some jello occasionally in horizontal flight. Zero improvement.
I think I solved the problem, I made a cross out of 5 mm carbon fiber tubes 20 cm x 20 cm with 10 grams of iron weights glued at the end. Total extra weight was 50 grams. That gave me a video with zero visible jello.
Not sure if in the long run I may have gymbal overload problems (I increased rotational inertia by about 2).
Canteliver like that is going to have a lot of resonances at different frequencies and having a secondary vibration isolation is asking for trouble. I think you should deal with frame vibrations though fast descents are always going to be problematic as the copter is flying through its own downwash.
It seems to work very well, zero jello, even in descent. Still need to test it in the field. The only small mistake is that the forward counterweight sometimes enters the upper field of view, I will amke the boom slightly shorter with a bigger mass.
The prop downwash is not the problem, because the descent is done in acro mode with high inclination forward (20 and more degrees), fast speed 60..120 km/h and low power setting (about 50 % of the hover power).
I am kinf of mad at myself because I spent a whole day building that double monster…when I finished it, I immediately suspected that it would not work..,
Personally I would suspect rigidity of that boom, one way to observe that would be to rotate the FPV camera 180° and looking for boom movement. You could try adding 50g to camera mount without the long arms.
Those rubber dampers are useless for gimbals. Basically they are too stiff and springy so the vibrations go right through it. look at the dampers that come with DJI gimbals they are almost like jelly. I think phantom 2 and phantom 3 dampers should fit.
I disagree that short arms would be beneficial. The whole idea to increase the rotational inertia, and that one is proportional to mass and square to distance from center. So doubling the arm length would result in 4 times the rotration inertia. And rotational inertia is the thing which would stabilize higher frequency vibrations which the gymbal could not. In any case, so far the arms worked very well. The only downside is obviously the air resistance increase.
It could be, because the small antivibration rubber which comes with the camera is extremely soft, compared to the bigger rubber shock absorbers… Anyway, the whole big mount plus 8 hours of work went to the rubbish bin…
