Servers by jDrones

Dual-motor tailsitters

( research air) #382

this is on my last post

(Jack Pittar) #383

Thanks Yonika.
Your method of making the motor mount with tilt servo is interesting.
Rather than just repeat your pioneering project, I think my efforts would best be served by going to the next step.
I am presently evaluating the offer of what, for our purposes, might be the ideal fast glider wing. The challenge will be in adding to an existing build.
It will be optimised for efficient and fast forward flight, so the design will have to facilitate extensive streamlining in the following build. The wing has a higher aspect ratio, and is not a symmetrical section, so I will have to add a short tail with just an elevator for longitudinal stability.

( research air) #384

keep me in the loop on your build, sounds great.

(lorbass) #385

Next trial with new props
heavy shaking of the elevons when safty switch ist pressed.Ch1, Ch2 oscillate 1000-2000.
Also after new accel calib.
When moving around by hand, shaking varies, some positions even calm but not on the ground.
Even when armed. Didn’t dare to start with this effect.

Can you help?

(tridge) #386

@lorbass Looks like your Q_A_RAT_PIT_D is much too high. I’m also curious how you have the autopilot mounted. Is it possible it has such a soft mount that its oscillating on its mount at 13Hz?

(lorbass) #387

Thanks for the quick answer.
Is Q_A_RAT_PIT 0.01 to high? Your advice after my second start with crash.[quote=“tridge, post:278, topic:15302”]
I also think you need more D term on pitch, I’d start by increasing it from 0.0036 to 0.01 and it may need more

Pixracer whithin printed housing, on damping balles.
I mounted the GPS on top with the idea to have a damping Mass.
(Your advice with an aluminium plate)
This is beside the D-Param the second difference to the last tryel with no shaking

supplement: Same behavior with ROLL and PICH D set to 0.0036
and GPS disassembled from PC Housing
When the wing is positionned horizontal, and Mode FBWA, then no shaking after safty switch on.

(tridge) #388

ahh, I’d forgotten that! That would have been because the pitch D was not contributing much to the overall pitch control, so I suggested raising it. Now the pitch D is dominating your elevator output, so it is clearly too high.
What I see in your latest log is the following:

  • both your Y gyros show 13Hz oscillation, showing the IMU is oscillating about Y axis (pitch)
  • the servo movement is in time with this oscillation
    something must be coupling the movement of the servos closely to the movement of the IMU. One possibility is through cables, perhaps that GPS cable?
    When the oscillation is happening try putting your hand on various parts such as cables and see whether restraining the part stops the oscillation.

(lorbass) #389

Yes, you are right.
Its a perfect 16 Hz Oscillator.
The results:
When I touch the FC less vibs, when I hold strongly the Wing which is oscilating at the back, no shakes.
GPS on top or not doesn’t change the frequency.
Same reaction with hard mounted FC.

When I hold the wing in the hand at the props vertically, a few shaks and normal, but hard, nervous and hypersensible, reaction of the Elevons.
The few shakes are after moving. The Q System seems to be at the limit of the oscilating system.
But when put carefully on the ground it starts to shake. The back, where at a Nuri the motor ist mounted, +/-10mm.
The whole wing is shaking because of the velcro fastener which hold the wingtips acting as landing gear.

Horizontally in FBWA
I can move, hit or shake the wing with absolutly no shaking elevons. Just smooth corrections according to the direction as expected.

Verry bad, the only one day with good wether since long time and bad wether anounced for the next few days:confused:.

( research air) #390

Hi, could you post a pic of your aircraft,I have had some of the same type of issues on some designs

(lorbass) #391

Hi Richard,
Here the Pics, Prop-Nuts not mounted.

Vertically = shaking
Horizontally = no shaking
Also with Pitch Q_A_RATT_PIT = 0.1 instead of 0.25

( research air) #392

any play in the motor mounts or control linkage in the pitch axis

(lorbass) #393

No play aniwhere.
Here the issue as video:

( research air) #394

the only thing I notice is the foam center section is flexing what happens when you disconnect the alerons, film it again ,don’t zoom in

(Perth UAV - Sam) #395

Wow are your motors even spinning or is that with the motors off?
It sounds like the servos are going spastic.
Yes please show the whole aircraft.

( research air) #396

also how stiff are the aleron hinges

( research air) #397

another question, what’s different from the videoed flight and now?

(lorbass) #398

The motors are off (not yet armed)
Elevons disconnected from the lever, no shaking.
Will make another video as requested.
The differences to the video with the crash are:

  • new Props (same size)
  • GPS from the back to the wing. (Interference with the RX)
  • Beeper from back to the front (camera mount)
  • FW change to the latest Master.

For the 2 first starts (mid April), no such behavior with the same setup.

( research air) #399

Hi, tryed hover with gray/ pixhawk,[quote=“tridge, post:361, topic:15302”]
Once you’ve tried QHOVER in flight,

logs sent google drive,video not so good, shoots out of frame

(lorbass) #400

Here the new shaking Video:

And the Graph of the Servo Outputs.
I can see a higher and a lower Frequency

For more details the Log:

Next try and last hope with hard mounted Pixracer (double-sided Adhesive)
May be less (dampening) is more.

(tridge) #401

I notice your throttle stick was quite high during the QHOVER tests. That is why it climbed so fast.
When in QHOVER or QLOITER the throttle stick maps to climb rate, not to throttle. So anything above 50% means to climb, and anything below 50% means to descend.
To do a takeoff in QHOVER, just start with zero throttle stick and slowly raise it to be just above 50%. Then when you get the desired height, move it back down to 50% to hold that height.