Make sure your Xserver (vcxsrv) is running first.
The console opens in a new window.
I think I can make it but I need some time. My main problem is to choose a model to disassemble.
1 Like
@hwurzburg The forward transition slews nav_pitch_cd from zero to the transition_angle at a rate of transition_angle/(transition_time_ms/2).
So it should take half the transition time to reach the transition angle.
The backward transition pitch rate must be set by the FW parameters; the comment in QuadPlane::update_transition is “nothing to do, this is handled in the fw attitude controller”
still cant figure it out…the backward transition from FW to VTOL has a period (1-2s) where the VTOL only motors (3/4 on my plus frame) remain off and the elevons are trying to pull the nose up (~30 deg) as the plane decelerates…then they switch on and it flips nose up the rest of the way …
its not related to transition angle used for forward transition…not related to deceleration factor in Qplane params as I thought it might be…ie take the airspeed diff between FBWA min and current, divided by decel factor…it varies from 1 to 2 seconds during several transistions…cant find a param that would account for it…Q_MOTOR_SPOOLUP is only 0.25sec…stumped…
pictures of the beater…btw Q_ANGLE_MAX doesn’t work well in 12mph winds…as I discovered today …never thought FBWA as a bail-out before for A Qplane
1 Like
This is what transitions look like for the BiWing in SITL:
All 4 motors are enabled in FW modes, and the throttle is set to hover thrust during transition from FW to VTOL mode. The transitions are “done” when the nose has pitched down (up) by Q_TAILSIT_ANGLE from vertical (horizontal). Q_TAILSIT_ANGLE is 35 degrees for this log, so “FW done” occurs at roughly +55 degrees and “VTOL done” occurs at about +35 degrees.
The timeout for transition to FW is TRANSITION_MS (2000 here) and the timeout for transition to VTOL is hardwired at 2000 msec.
great thanks…pushed a PR…you are a reviewer…
Wrong text last time.
The right answer from Ubuntu was:
unable to access the X Display, is $DISPLAY set properly?
Loaded module console.
But by the way. I can fly and even land without tilting with the BiWing in QHover. I needed to set the Q_A_RAT_RLL_P to 0.2 because it shaked in Roll.
The command explorer.exe . is the best command so far in Ubuntu. It gives more overview in the file structure and lets copy files verry easely .
Thank you for your great help.
@lorbass Glad to hear you have your BiWing model flying. It’s great to be able to test with an accurate model.
What physics frame rate are you getting in RealFlight? If it is too low you might have to reduce PID gains. RF8 can display the physics frame rate in the NavGuides widget:
Also, try starting MissionPlanner on the machine running SITL, then connect to UDP port 14550
Now everything works great. Even MP. 
Framerate about 300 fps. Graphic reduced as recommended in Wiki.
Pitch in QHover verry waggly. Need to push full throttle, else it tip over bevore takeoff. Is it possible to calibrate the InterLink? Need always to steer against constant Yaw. So far not found in manual.
300 fps is very good. I find that 200 is adequate, and I’m using the 3D obstacle course field with fairly high graphics quality.
Try running QAUTOTUNE; I haven’t tried to optimize the tuning for qhover.
I just ran a pitch autotune and these are the resulting PID gains:
Q_A_RAT_PIT_D 0.049041
Q_A_RAT_PIT_FF 0.100000
Q_A_RAT_PIT_FLTD 10.000000
Q_A_RAT_PIT_FLTE 0.000000
Q_A_RAT_PIT_FLTT 0.000000
Q_A_RAT_PIT_I 0.475337
Q_A_RAT_PIT_IMAX 1.000000
Q_A_RAT_PIT_P 0.475337
To calibrate your controller go to Simulation:Select controller in RF8
Nearly everything works. I could even download a log.
After a certain time, the left stick doesn’t react anymore. When swiched the mode, it came down.
This was already the case days bevore.
No one hurt.
Here the log: https://drive.google.com/open?id=1Utt4RBQaWk5lZaa96KDNUaGAs3xmLIL7
That sounds like it could be a network problem. Does the mavproxy (Ubuntu) console have any messages when this happens? (perhaps “link down”)
In Ubuntu console nothing to see. Continuously indicating the height xx line per line.
The right stick reacts, to aircraft symbol in MP move. Will check again.
Viewing the log, I think Pitch Tuning is good. Only at the lift of it tilts by ~90 °
Edit:
Strange, I calbrated the Tx and modified the modes 2 and 4 from 23 (QAcro) to 5 (FBWA) and it lifts of as a moon rocket, slow, vertical.
In FBWA no luck, Transition good, but then it rolls left and sink.
Will test more to show logs.
@lorbass Maybe we’re using different versions of the BiWing RFX.
I just re-exported the model I’m using and pushed it up to:
SITL_Models/RealFlight/markw/Tailsitters/CopterTailsitters/BiWing/BiWing_JWL065_EA.RFX
cd to your SITL_models repo in Ubuntu and type
git pull
That should update it to my latest version. Then re-import it in RF8 (I think you’ll have to delete the old version first, or overwrite everything). And see if the new version behaves better.
I haven’t been able to figure out what’s wrong with the prop visuals. They are swapped left to right on both wings. This is a demonstration of why I say they’re swapped:
I spin up only the bottom left motor, and the bottom right prop (visual) spins. But the vehicle flips over to the right, indicating that it is really the left prop generating thrust, not the right.
If I swap the right and left visual frames for the engine, both props disappear.
Now I’m a little bit confused. Your Wing ist tilting, an mine hovers perfect.
https://drive.google.com/open?id=10bmD18YoTd4lMRbtcW3jkdbO3MOJiZgw
Therefore I didn’t change my files in SITL_Models.
The date of all files I have in SITL_Models are 11.12.19. In RF8 I imported BiWing_JWL065_EA.RFX and BiWin9swapLR_JWL065backTorque witch is selected as Aircraft.
The quality of the Vid is not the best, I had it to record with a Handcam. The capture of the RF8 screen was not possible due to the decreased Frame Rate to 45 wich produced a wrong fligth attitude. And recording with RF8 produces a file which can’t be read by other programs and strange the mots are lateral as they are while designing in 3DS Max.
How do you record RF8 Screens?
1 Like
I can’t believe it is a simulation, the way it fly is so realistic !!
Yes, I’m satisfied with your design, QHover makes fun.
But when I switch to FBWA, after a good transition it crashes immediately with left roll and high speed, also verry realistic, sorry.
By the way, did you receive your RF9?
When I switch to FBWA, it rolls left after a good transition and crashes. When I put the left stick full to the left, it rolls nearly not and I even could switch back to QHover and avoid a crash.
May be, the elevons are also crossed and the more it rolls, the more SITL steer to left?
@lorbass The video I posted yesterday is a demonstration of what happens in manual mode:
I manually throttle up only one motor, and the bottom right prop (visual) spins. I therefore expect the model to flip over to the left, But the vehicle flips over to the right, indicating that it is really the left prop generating thrust, not the right.
If you look at the video closely, you will see that only the bottom right prop is spinning.
This model flies well (for me) in all VTOL and FW modes, and it is the one that you should be using also. You would definitely benefit from updating to BiWing_JWL065_EA.RFX