I'm also getting really twitchy video and I can't figure out why. Would appreciate *ANY* advice. CS8, 3-axis gimbal with Radians, Canon T2i with 24mm lens, balanced props, balanced cam on gimbal Started with all red o-rings but panning was jumpy even with low gain. On the advice of Quadrocopter, added blue o-rings for an even mix. That seemed to fix the pan jumpiness but video got really twitchy so I thought maybe it was too stiff. Softened it to 2 blues on the outside and the rest red. Have dialed gain and slew down (pan 30 gain, 40 slew / tilt 60 gain, 40 slew / roll 80 gain, 50 slew). Still super twitchy and totally unusable video. I have a short sample uploaded here: http://dancoplan.com/client/CS8shaky.mov (less than 6MB so should load relatively quickly) The heavy compression to keep the file size small adds to the twitchiness, but believe me, the original video is not much different. Dan
By no means have I got this figured out but the gains seem low. Check out what others are using over on the spreadsheet at https://docs.google.com/spreadsheet/ccc?key=0AvZl8S9wZihWdHBhV3F2ZXZLQm5zM2hfci1tN1pPdlE#gid=0
In order to truly figure out whats going on and causing the bad video we first need to find out if its the copter or the Radians. Do a test flight without the Radians active and post that video. If its totally smooth without the Radians then we know the issues are down to gimbal (balancing, servo or belt slop, etc.) and Radian adjustments. I also agree that your gains are probably low but with low gains you would be getting softer twitches. Harder, more frequent movements would most likely be gimbal vibrations. Brad
I would check the cg of the camera to the center hub. I was noticing the same thing with a 5d. My issue was that the camera cg was forward of the copters center axis. This will cause the inertia of the camera to put more force on the front dampers, forcing the front to act as a pivot point. As it pushes on the front it acts as a pivot point and forces the rear dampeners to move th rear dampeners As the gimbal rotates it will transition that force to the next dampner and will decay off the others and thus cause a spring force back. If the cg is center on the hub as much as possible, you are distributing the force evenly across all dampeners reducing the springiness of the system, because all dampeners will get the same force no matter where the gimbal is rotating. The load will be equally distributed across all dampeners reducing the fulcrum effect from one dampener to the other. This is accomplished by rotating the tilt arms up to bring the cg closer to the center. You can test this on the ground by putting the copter arms between two tables and running the gimbal.
I've been going through a 3-step process, following the videos online, of balancing all 3 axes of the camera on the gimbal. Either my method isn't very good or I'm having other problems. I need to do what Brad said and test with the Radians off and then on. I'm using APC 14x4.7 props and a pair of 5000mAh batts in parallel. Dan
If you done all that then probably is something else. One thing to consider, is that if you are close to an object or zoomed in. It won't matter how good your gimbal stabilization is because you cannot negate the physics of the vertical and lateral movements of the aircraft. If the air is turbulent and the aircraft is bumping up and down left and right it will project this in your footage. Even the 4 million dollar gimbals that Raytheon makes cant negate this. But it is a good idea to make sure your gimbal is solid as possible.
if you are using 5cell lipo with 14x4,7 props ,that does not work well,the machine wil look like a crow in the air!!!
Really? Why? What then do you recommend? My motors are MT3515's which require minimum 5S batts. Smaller props? The copter seems to fly really well but of course it's impossible to see vibrations just by looking at it unless it's all over the place. My lens, by the way, is a 24mm (36mm equivalent on the T2i) so it's reasonably wide. Dan
hello i fly 5cell cinestar 8 with 360 gimble and 13x6,5 carbon and it is supersmooth i just tryed to change props to apc 14x4,7 and the cinestar starts to jump and humm and shake like H i fly with the sony 730 so maybe your ship it heavyer then mine i will try the 14x7 carbon props and see, i think it will be mutch more stable with 14x7 then 14x4,7
Kjetil - What's the theory behind the octo flying more smoothly with 13x6.5 props? I got the APC 14x4.7 because that is what was recommended.
Dan he is flying with a 5s battery, higher voltage, faster rpm's, therefore you can use a smaller prop.
And smaller props = better stability and less vibration? I heard that using larger props makes a motor more efficient but I'm fairly ignorant (though learning) when it comes to this stuff.
That's pretty much accurate Lager props grab more air at lower rpms. Trade off is frequency vibration. Lower frequencies can propagate further and are not attenuated as easily. If the motor is spinning faster than the vibration frequency is higher an will dampen easier, naturally through attenuation in the structure. If you have smaller props revolution has to increase to produce the same lift vector that the larger props at lower RPM creates. I am at higher altitude so my 14 apcs have to spin faster by design so I don't have issues here. Although proper balancing of the props make a huge difference in reducing the vibe.
i see you got the answer from gary just try them, you will see and feel the big different,if you fly 5cell lipo
Finally got around to doing a quick test with the Radians OFF and then on in FIXED mode. Would very much appreciate if any of you have a moment to watch my short comparison video (just over 1 min.): The video on the left is with the Radians OFF. The camera is a T2i (by nature not a vibration-friendly camera) with a 24mm lens. Please ignore the Dutch angle in the videos. I probably should have done a simple hover test as well, but I didn't. I flew the octo around here and there and then did some small but aggressive side-to-side and front-to-back maneuvers. From what I can see, the video on the left is pretty damn smooth (all things considered) and tells me that the physical nature of the copter is sound, whereas the video on the right exhibits twitchiness, bumpiness, and the gimbal not keeping up with the movement of the copter. That indicates it's an issue with the Radians. So then what do I do about it? I've had the same problem adjusting the gains and slew rates from high to low. These were replaced from a batch that had a service recall so I would assume they're the latest and greatest. To be sure, I downloaded the Freefly Bootloader and 1.03 firmware update to confirm, or update if necessary, but that's another issue... Ran the Bootloader fine and loaded the firmware, but when I click 'Connect to Target' (which turns green), I get a message that says, "No response." I have power to the Radians. I used the latest version of the Radian software to program the Radians. Many thanks for any help with this! Dan
Have you followed all the balancing procedures for the gimbal? The "dutch angle" as you call it implies to me that maybe the camera is out of balance. Personally, I have found it frustrating to balance my cameras, as the gimbal's level of adjustment (in particular the height of the camera on the plate) has prevented me from getting a 100% clean balance on the tilt. The roll is much easier, but the critical thing there is not only balancing, but aligning the optical center of the lens with the hinge point of the roll axis. With the Radians, it all starts with them being balanced. If they're not, all bets are off. I have the adjustable tilt bars on their way to me, and I hope they'll allow me to achieve a perfect balance, at least with my 5D Mark III (whose CoG is a bit too high for the standard tilt bars). My Sony CX760V is the opposite problem, as its CoG is a bit too low, and I have to find a way to elevate it. Sigh.
Dan, Your video on the left looks pretty stable. One obvious issue is that your camera is tilted so far over in the video. Can you post a photo of your gimbal, specifically where the Radians are mounted? With the Radians connected, the roll should not be far from level... and if needed can be adjusted within the Radian software. If your Roll Radian is not mounted parallel to the camera mount plate then that could be causing issues with stabilization. Also, as you said the camera seems to lag with respect to the copter movement. That would be a gain issue where your gains aren't set high enough. It seems to be in both the Roll and Tilt axis. Brad
Steve you are gonna love the adjustable tilt bars. Only other item that can speed up the balancing dance is the adjustable camera rail. If you have a camera where the center point for balance needs to have the camera screw right over the top of the tilt bar or you want to quickly be able to remove the camera to change lens, battery, memory cards the rail is great. Doesn't take much up and down to find the COG for the camera using the adjustable bars.
Steve, why would you need to align the center of the lens with the pivot point of the roll axis? And how would you do that with cameras like the 5d that are asymmetrical in weight distribution? If we are balancing our roll axis then typically isn't it done by adjusting the camera and/or gimbal bar left or right to achieve balance, rather than adding weight to one side or the other? Maybe I've missed something....