Separate names with a comma.
Discussion in 'Radian' started by Brad Meier, Sep 19, 2012.
Here is a video showing the Judder dissapearing when i turn off the radians..
Heres another video to give you an idea of the Judder were dealing with.
I'm just going from my observation that when I try to tighten the belt to what feels as tight as it can go without breaking anything, then it appears that there is still 2-3mm deflection at the mid-point between the large and small pulleys.
So it's purely experiential data. Let me know if your experiments produce less deflection than what I have observed.
Hope this helps.
You want to tell me that Radian with CS gimbal and CS8 combo is a crap??? it wont work???
@PaNt: Not at all. What we're discussing is that you have to fine tune the copter/gimbal system to get it to work correctly and when you do, you will get excellent results. Have a look at some of Jeff Scholl's videos and I think you'll see what I mean. There are also other folks who are using the Radian sensors on C8 gimbals and getting excellent results.
The problem is that the fine tuning for one copter is not transferrable to another because we are dealing with a different mechanical situation each time: different motors, different props, different masses in motion, different moment arms, differently adjusted. It's a complex series of simultaneous equations to be solved.
What was it in my message that you construe as me wanting to tell you that the Radian/CS8/gimbal is crap?
When i say crap i mean it doesnt work as it should..1
CAn i see a video from an amateur without post stab to see what you can do..>!
I am trying to tune mine and its not so easy..! i have balance everything and i have tighten the belts..! Also i have put the tireups on the dampers..!
Now i have to play with the gains..
I agree that the tuning is not easy, but then the problem is not easy.
I'll be interested, as I'm sure you are, to see what videos are posted.
Try increasing your gains. I had the same type footage as you running gains sub 100. Make sure everything is balanced and crank those gains up to 140 or so. With my heavy-ish 7d and lens, my footage is much better. When adjusting the radians it seems that increasing the gains until just before oscillation is the spot, where oscillation is the uncontrollable and increasingly unstable movement.
So.. you can hang it and increase the gains until tapping or moving the controls creates this uncontrollable oscillation then back off the gain a few and try again. My gains are between 135 and 145 and I only have minimal rotational jitter
Brad: what did you end up with the pan gain set to? That's the one that seems to work best down below 100 -- I get oscillations (with a Sony CX760) even down to about 60%.
Andy, I was running 70 at the beginning of this thread and never had a problem but I noticed recently in one of my videos that it seemed to drift around in that axis very slightly. After upping the other two gains I decided that id try the pan as well and I havent had an issue with oscillations.
Im sure your balanced and the belt is tight. Maybe youve got too much weight mounted out on the 3 booms of the gimbal? Ive tried to mount most of the weight as close to the point of rotation as I can. Maybe just having the weight of the 7d there helps mine that much more...
Yeah, I was wondering about the location of the various masses on the gimbal but the only slightly heavy one (beyond the Radians and the servos), is the QC1200 battery mounted on the right hand (viewed from the rear) side boom -- and I put it there to balance the mass of the servo and pulley.
Apart from that, there's really only an HDMI convert on the down-post and that's close to the pivot point.
The real mass is centered around the camera plate.
I think i have gotten to the bottom of my "Judder" issue..
After Many hours of testing, i have come to realise that it stems from prop cavitation!!
I have changed the props, and only get this Judder now when i fly aggressively, changing direction, and causing the props to cavitate!!
Also, im very surprised at the gain percentage.. Im approaching 200!! on both tilt and roll, with only a slight "Bounce" but again, only when i fly like an animal. All the rest of the time, its beautifully smooth compared to my previous videos.
I have also re installed the Cable tie mod.
Hmm great news.... Can you tell what props you had??? And why you had cavitations??
Also whats your gain on Pan??
THe props we were using were the Xoar wooden 14x4.75.
Before we were using these, we had the APC slowfly 14x4.75 props.
The only difference between these props is Matterial and profile.
The Cavitation comes with too much weight,m on too smaller prop. The APC still cavitates when i fly harder, but not all the time.
The wooden props arent as wide a Chord as the APC, so arent as efficent.
pan gain is only around 30%, as we have a fair weight outboard on the gimbal legs.
Im now up at 235% on the tilt, and 220 on Roll which is as far as i can go before it shows signs of being too high.
I think seriously, that if we want to eliminate the cavitation. we need to change our power setup so we are in the middle of our carrying capability, rather than trying to tune toward the limit!
This i am going to look into next i think.
Ill try to put a vid together to show the difference between the shot yesturday, and the shot today!
One min.. The cavitation was with the xoar right???
Also what motors you have and whats you flying weight??
Hi all, I have almost the same problems as Pete and I´m sure as many others. Even in dead calm weather there is that little but annoying flex on the roll. I discovered that there is a little play in the servo gearset. I swooped the tilt and roll servos (I have a two axis setup) as it seems that there was a little les play in the tilt servo. That helped but not perfect. The teeth on the belt also seems to contribute to a unwanted movement. So I turned it inside out. That helped me also a little. But it’s still not perfect. I think the radians is all good. Here is my theory. If there is a little paly in the servo gearset and the belt maybe also in the hole gimbal construction the radian trying to stabilize and if nothing happens immediately because of the play, it asking for more and more and more to compensate and when the servo finally deliver the movement the radian asked for it´s to much. Then radian trying to fix the over compensated movement by doing the opposite and then we have the problem... Everything is tighten up the camera is perfectly stabilized APC props also balanced. My gain on the roll is 130 and its seems like a good choice.
Is it possible to get a servo with no play in the gearset? Or am I just unlucky with mine? There is a clear difference between the two..
What do you guys think?
I think your theory is sound -- certainly if there were play in the servo/pulley mechanical system, it would produce that kind of "hunting" where the sensor is "chasing after" the mechanical system and causing an overcorrection.
So the pivotal (no pun intended) question is whether there really is sufficient mechanical play in the system to cause the problem? If you push your finger against the belt at the midpoint between the small pulley and the large pulley, how much lateral deflection do you get in millimeters, please? I would also be interested to see, if you hold the small pulley tightly, is there any looseness in the roll axis?
I know Tabb has gone to some lengths to get servos with zero (or minimal) backlash, so that's why I'm starting with the rest of the mechanical system -- I'm not in denial that it couldn't be the servo, just working the hypothesis that it's less likely to be the servo than the rest of the waggly bits.
Thanks for your immediate reply.
The deflection of belt is not more than 2-3 mm (did not know you Americans was into the metric system ) All is tighten up and if I try to hold the small pulley there is no waggly parts..
It turned out that the tilt has less play in the gearset than the roll servo - so I swooped them and turned the belt inside out. It helped but its still not perfect.
I´m very interested in the servos Teed looks in to. I hope he will win and spread the word..
Some of us Americans were born and raised in England and lived in France and Canada for a while.
OK, that's eliminated that.
Interesting -- so there is a mechanical difference in the FF servos, you're saying?
Sorry - I meant Tabb
There is a play in both of my servos, but the tilt is differently "better" than the roll servo. That´s why I swooped them. If Tabb finds a servo with no play i´m sure it will solve my problem. I guess that the freefly servos will be the only ones fitting in to the radians?