Copter Tailsitters

It could be. I haven’t done any calibration after switching the setting from PWM to DSHOT.

I guess H wing Not suitable for Pixhawk 1.I tried to download ready firmware. I saw that it is not supported. Which board are you using?
Will my project turn into x profiles! Help me😅

I have reviewed most of your projects. What is needed for me is vtol dash landing. I don’t want vtol flight in the air. Wouldn’t that be enough for these conditions?

Thank you. I manually modified the PID based on the information you provided and the oscillation is improved.
The throttle control in Qhover and QLoiter is not so responsive, but I will try to manage it too.

1 Like

Test flight again after PID adjustment (but still have throttle response adjustment).
It was a successful 3min flight, but I could not transition from horizontal mode to vertical mode.

I made the transition from FBWA to QSTAB in the air near the takeoff point. However, the plane never turned its nose to vertical (no vertical climb) and just dropped to the ground.

I guess there are two problems

  1. Insufficient throttle opening when switching from FBWA to QSTAB (QSTAB>FBWA: 40-50%, FBWA>QSTAB: 25-30%), so the attitude change was not possible. (In this case, I should have opened the throttle just before the QSTAB transition?)

  2. Pitch change response during transitions was poor. The action setting for the vertical mode transition was inappropriate?

Do you have any good ideas on how to improve this?

[log]

I agree that there probably was insufficient throttle demand during the transition from FBWA to QSTAB.
@iampete Is this specific to forced QASSIST?

2 Likes

Yeah, I agree. Needs more throttle.

If you update to latest master there is a new param Q_TAILSIT_THR_VT that lets you give the throttle you want to use for FW to VT transition.

The other thing is to increase your throttle mix params, that will allow the vehicle to increase the throttle automatically if there is a angle error. Q_A_THR_MIX_MAN and Q_A_THR_MIX_MAX, the max value is 2.

For the swan K1 we were able to use a throttle mix of 2 and a very low Q_TAILSIT_THR_VT that means it uses the minimum possible throttle and don’t gain as much height.

3 Likes

hello Mark. I want to ask you something. The Pixhawk 2.4.8 im FS-IA6B does not see the PPM signals of my receiver. How can I fix the problem, do you know?

The maximum value of Q_A_THR_MIX_MAN / MAX is 0.9, can it be increased to 2 without problems?

Yeah anything up to 2 works. The ranges are only guidelines. Maybe don’t jump all the way to 2 in one go tho, small steps.

ok, it is understood

Hello,

This is my first post. Many thanks to all those who have made this possible.

I am building a blackfly like tailsitter. I was able to get it to fly in QHOVER mode and transition successfully to FBWA. I am having oscillations along the pitch axis during fixed wing mode.

I’ve tried going into AUTOTUNE, but the values do not seem to be saved.

I appreciate any help getting me in the right direction.

Log file:
https://drive.google.com/file/d/1XV9k-1vUiN516uF-zIrzbvsXx4CSxDAK

Flight video
https://drive.google.com/file/d/1NPMmpPMecq1mGrcqfSDa826kFcKlZfNX

many thanks for the video, It seems to me you are very close to get a very nice aircraft.
Where is the CG ? I thinks your aircraft is just unstable because of rear CG.

Thanks, I am hoping it’s just the CG, but I think it’s more related to PIDs. I am not really that versed in understanding how to increase/decrease the PIDs properly.

I am using this diagram to calculate the CG:

Initially, it was to the back, but I brought to the 25%-28% range. I did not see much difference.

Another thing I have been playing with is changing the angle of attack (by raising the pixhawk from the rear). I started with 4 degrees and been lowering… Currently, it is 0.5 degrees. I have not noticed much change.

One thing I forgot to mention, the oscillation starts at around 40% throttle. It is stable at lower throttle. And, I dont notice much increase in speed after 50% (max speed was around 16-17 m/s).

instability at higher speed is typical of rear CG
Could you calculate and compare CG location with the help of this calculator
https://www.ecalc.ch/cgcalc.php
static margin should be at least 5%
front wing is the main wing and rear wing is the stabilizer.
I like so much the flying flea from henry Mignet and I read his famous “bouquin” but I dont know if the flea CG location is applicable to your black fly. Also the flying flea was badly famous because of uncontrollable dive specially when cg was backward.

Maybe an other idea, are front wing control surfaces active for pitch control ? If yes there should be canard style.

Will look at the CG in more detail.

The ailerons are on the front wings and the elevators are on the rear wings. Do you recommend changes to the configuration (easier than cutting new wings)?

you can set control surfaces on front and back wing as elevon.
At least it will improve hover control authority but will not probably improve fbwa stability
edit: Looking at your log I see you have servo6/8=77/78 so elevon, I suppose on front wing, do they deflect downward to pitch up ?
did you enabled q_assist ? I do not understand why motors output are oscillating in plane mode.

You are right, I played with multiple configurations. This run had the elevons on the back wing. I will make all four controls surfaces elevons. I will try to check the deflection on pitch up.

I have the following values for Q_ASSIST:
Q_ASSIST_SPEED, 10
Q_ASSIST_ANGLE, 30
Q_ASSIST_ALT, 0
Q_ASSIST_DELAY, 0.5

I also have:
Q_OPTIONS, 384

The motors are oscillating a lot and I can even here the change in tone. Looking at the log, you can see power is oscillating and so is servo output (if I am reading the log correctly).

yes, all output are oscillating,
If I remember there is no differential thrust for pitch in plane mode except when qassist is enable (so no oscillations for this axis). I can see PIQP_P oscillating the whole flight, I need to check an old log because it seems to me that should not happen in plane mode. To make things more simple I suggest to disable qassist_speed, it will disable the assistance for angle too.
what feature is enable by qoption=384 ?