Aeronavics / Droidworx Skyjib X4 Ti-QR "wobble"

MadMonkey

Bane of G10
I've had this issue for a while and I'm getting nowhere fast.

We ordered this rig from Intelligent UAS (first mistake... don't ask) last year and finally got it. Unfortunately, it has an odd tendency to wobble very time I return pitch/roll back to center.

The controller is an A2. Motors are KDE 5215XF 435KV 22P, ESCs are KDE 75A HV. Props are APC 16X5.5MR (switching to new 18X5.5MR once it's sorted). Batteries are Tattu 6S 16,000mah (2X).

Here's some video. I had a friend take it today and didn't realize I had it on low quality settings, but you can still see the wobble.


So I first noticed this on the early test flights. Having the MoVi gimbal installed didn't make much difference, just dampened it slightly (at about 9lbs AUW).

I tried a multitude of gains ranging from scary low to scary high with no effect except for speeding up or slowing down the wobble. Firmware was fine. All the setup was fine.

I finally sent the controller back to DJI, and they sent it back with a new IMU, saying the old one was bad. Great, I reinstalled it, updated everything, set it all up again.

Test flew it.... same problem, no changes whatsoever. Went through the gains process again with no improvement.

Emailed DJI, they told me to send it back again, so I did... they sent it back, saying that the problem was a firmware mismatch (even though that's the first thing I checked when I received it previously). Great, whatever, tested again today... same problem.

Now, I'm starting to think that the issue might actually be with the combination of airframe, motors, ESCs and props that I'm working with. Our Cinestar X8 with U7s and T-motor ESCs is absolutely LOCKED IN with a WKM, no issues like this.. once I got the gains where I wanted them it flies like it's on rails, gimbal or not.

Nothing seems to be working with this though.

What's really frustrating is that we were originally supposed to get a Wookong on this bird, along with U7s and T-Motor ESCs, but Intelligent UAS decided to "help out" by upgrading to these motors and ESCs and the A2 without asking (I mean, good on them for trying to help, but... crap).

My next step of testing is partially completed. I have the A2 installed on a DJI F450 and have already flown a short test hop in the kitchen. I'll fly it outdoors tomorrow and see if the problem is still there.

Any suggestions?
 

MadMonkey

Bane of G10
So I test flew the F450 with the A2 today and had no issues, it was rock solid with no sign of the wobble that the Skyjib has, which leads me to believe that it's a motor, ESC or prop problem. Prop can be mostly ruled out as it happens with both 16" APCs and 18" T-Props.

Sooo that leaves the motors and ESCs, unless it could be a frame problem, which I doubt.

Does anyone have experience with KDE motors and ESCs? Are they known to have slow response or strange tendencies in large multirotors?
 

Bartman

Welcome to MultiRotorForums.com!!
no experience with KDE motors/ESC's but a few questions and I doubt it's a frame problem

Are you absolutely sure that all of the ESC's are set exactly the same way? (can settings on KDE ESC"s be changed?)
That wobble looks more like a low gain problem, high gain problems usually include a screeching of the motors as they over control everything
Are you absolutely sure the IMU positioning dimensions are correct? X,Y, and Z with positive being positive and negatives being negative?
Was the F450 hovering at a similar throttle position? KDE 5215 motors sound like very big motors, maybe too big for the props you're swinging and if the motors are excessively powerful then the flight controller will have difficulty smoothly controlling the helicopter. Try adding a lot of weight under the helicopter and see if control is any smoother.

I had a wobbly helicopter once and I'd swear the wobbles were from the FC systems tending to overcontrol and the ESC's being slow to respond (I think it was a Mikrokopter FC2.1 but with standard ESC's instead of MK BL's which are much faster).

Too much lift is as bad as not enough lift, IMHO, and if it's too out of whack no amount of tuning will get it optimal.
 

MadMonkey

Bane of G10
no experience with KDE motors/ESC's but a few questions and I doubt it's a frame problem

Are you absolutely sure that all of the ESC's are set exactly the same way? (can settings on KDE ESC"s be changed?)
That wobble looks more like a low gain problem, high gain problems usually include a screeching of the motors as they over control everything
Are you absolutely sure the IMU positioning dimensions are correct? X,Y, and Z with positive being positive and negatives being negative?
Was the F450 hovering at a similar throttle position? KDE 5215 motors sound like very big motors, maybe too big for the props you're swinging and if the motors are excessively powerful then the flight controller will have difficulty smoothly controlling the helicopter. Try adding a lot of weight under the helicopter and see if control is any smoother.

I had a wobbly helicopter once and I'd swear the wobbles were from the FC systems tending to overcontrol and the ESC's being slow to respond (I think it was a Mikrokopter FC2.1 but with standard ESC's instead of MK BL's which are much faster).

Too much lift is as bad as not enough lift, IMHO, and if it's too out of whack no amount of tuning will get it optimal.

Too much lift was actually my very first thought when I noticed the problem... when idling at medium speed it was barely under takeoff power with the original props. I quickly dropped it to low... the first few landings were terrifying because the aircraft acted like it wanted to fall over, and it nearly did once because it was so lightly touching the ground (the little bitty landing gear didn't help matters, I zip tied longer tubes onto them as a temporary stability fix).

As for your questions...

1. Wobble appears to be in no way, shape or form related to the gains. I've tried a TON of different gain combinations, usually using a knob on the transmitter so I can get instant feedback. The only effect they had was to make the wobble slower or faster, it was still there regardless even near max and minimum gains on both basic and attitude and any combination thereof (I played with them a lot).

2. IMU is set up exactly as it's supposed to be (to the best of my knowledge). Ever since having some issues with the very same F450 I'm using as a testbed, I've been paranoid about everything being perfect... this was the third time the A2 had been installed (twice by me), and I've measured and re-measured everything to be sure. On the F450 with the Naza V2 I actually played around with incorrect measurements to get an understanding of how the aircraft would react. I even mounted the GPS backwards (that was fun! :D).

3. I didn't test either aircraft in manual mode, I no longer use it except in my 3D/FPV multis, so I don't know what the actual hover throttle setting is. I just set both the Skyjib and F450 to mid stick for the main testing, though I've flown each at varied throttle settings and through many maneuvers and both exhibited the same behavior (Skyjib wobbled, F450 was perfect).

4. Those motors are VERY big. Half throttle supposedly puts out almost 7lbs of thrust with the recommended 18" prop... 100% is 17.53lbs... 8 motors at 17.53lbs of thrust each is over 140lbs of total thrust. I mean... sure, that's a nice excess if we wanted to carry a fully set up Epic but holy crap. I think if 1UAS had stuck with the original U7 specifications, which are the same as what we have on the Cinestar X8, we would have been fine!

5. I also thought of adding weight. Our fully-built gimbal (MoVi M10 + GH4 and accessories) weighed about 9-9.5lbs at the time I tested it. We attached it and it had the exact same wobble, but very slightly dampened.

6. ESCs being slow to respond is another thought I had when the supposedly perfect A2 came back to me and had the same problem. It does seem like that.

I'm not sure what to do. My first thought is to tell my camera guy to just sell the entire rig and I'll start over from scratch and build a new rig myself, but we'd end up losing a ton of money selling this thing. Second thought is to sell off the motors and ESCs, which work perfectly, just not in this application. Then go to either T-Motors or whatever.

Ugh.
 

Can you post some pictures of your set up?
There's not much point trying to tune the multirotor with out the correct weight there, so you need to add weight to it while you tube. Strap some extra batteries or a dumbbell to it. Make sure its the same weight as with the Movi.
Id also set it up and get it flying in manual mode first before trying any other modes.
Find out what your throttle hover % is. That's a good start.
 

Bartman

Welcome to MultiRotorForums.com!!
@MadMonkey

have you seen any of my videos where I list everything for a build and use all of the weights to determine motor hover % and battery usage?

you might want to back calculate your rig as accurately as possible and see if those motors even have a useable sweet spot that you can build from. you might find by looking at hover lift per motor that the motors are way out of the power band where they would be good for the weight they're flying at.
 

MadMonkey

Bane of G10
Can you post some pictures of your set up?
There's not much point trying to tune the multirotor with out the correct weight there, so you need to add weight to it while you tube. Strap some extra batteries or a dumbbell to it. Make sure its the same weight as with the Movi.
Id also set it up and get it flying in manual mode first before trying any other modes.
Find out what your throttle hover % is. That's a good start.

As I said earlier, I already tested with the built MoVi and had the same results, just slightly dampened (with various combinations of gains).

With the way this thing has been acting I don't feel comfortable at all flying it in manual mode.

@MadMonkey

have you seen any of my videos where I list everything for a build and use all of the weights to determine motor hover % and battery usage?

you might want to back calculate your rig as accurately as possible and see if those motors even have a useable sweet spot that you can build from. you might find by looking at hover lift per motor that the motors are way out of the power band where they would be good for the weight they're flying at.

I haven't seen those, got a link? That definitely sounds plausible. Something about the whole setup is bugging me but I can't figure out what.
 

Top