3D printed Cygnet VTOL build (motor-only, no control surfaces, PETG version)

ArduPlane VTOL Plane
1

Hello everyone,

I am building a 3D printed Cygnet VTOL by richardg3 from Cults3D and I would like some advice before I go further with the setup in ArduPilot.

This aircraft is unusual because it appears to be a motor-only VTOL, with no elevons and no other control surfaces.
As I understand it, control is done only by motor thrust, and the motor angles are part of the design to help with roll/yaw authority.

I already contacted the designer, and this is what he told me about his original build:

  • wingspan: 1249 mm
  • battery: 4S
  • motors: DYS 2212
  • propellers: standard 9040
  • firmware used: ArduPilot only
  • no control surfaces
  • motor angles were intentional in the design
  • body and wings were printed in LW-PLA, 2 walls
  • landing legs are a weak point and should be reinforced with a carbon rod and stronger material

He does not remember the target AUW, exact CG, or the exact ArduPilot vehicle/frame setup he used.

My situation is different in one important way:

  • I have already printed about 70% of the aircraft
  • I am printing it in PETG, not LW-PLA

So my main concern is that my version may end up significantly heavier than the original concept.

At this stage I am trying to avoid guessing and would really appreciate advice from people who have built or tuned anything similar.

What I would like to ask:

  1. Which ArduPilot vehicle / frame type / setup is the closest match for this kind of aircraft?
    Is this best treated as a tailsitter, a QuadPlane variant, or something else?

  2. Has anyone here built or tuned a motor-only VTOL / tailsitter with no control surfaces?
    Especially something 3D printed or with angled lift motors.

  3. Does anyone have a parameter file, build log, or example setup for a similar aircraft?
    Even an approximate starting point would help.

  4. What are the most important parameters to get right before the first hover test?
    For example:

    • frame / motor setup
    • output mapping
    • orientation
    • hover throttle estimate
    • gain starting points
    • transition-related settings
    • safety checks

  5. How sensitive is this type of aircraft to extra weight?
    Since mine is PETG instead of LW-PLA, I expect a higher all-up weight.

  6. Would you recommend staying near the original 2212 / 4S / 9040 idea, or should I expect to move to a more powerful setup if the final weight is too high?

  7. If this design is likely to be difficult or unstable in ArduPilot without control surfaces, I would appreciate honest feedback before I invest too much time in final assembly.

Current build status

  • around 70% printed
  • body/wing structure mostly done
  • I will weigh all components individually once printing is complete:
    • fuselage
    • wings
    • landing legs
    • motors
    • ESCs
    • FC
    • GPS
    • wiring
    • hardware
    • battery
    • final AUW

My goal

My immediate goal is not full forward-flight performance yet.
First I want to understand the correct ArduPilot approach and achieve a safe first hover.

If anyone has experience with a similar airframe, parameter file, logs, or even just guidance on what category/setup I should start from, I would be very grateful.

Thank you very much.

Best regards,
Bogdan

20230406_083400
20230406_0834001920Ă—909 214 KB

20230404_095733
20230404_0957331920Ă—909 296 KB

2023-03-04 11_31_19-Swan k1 V8 - DesignSpark Mechanical
2023-03-04 11_31_19-Swan k1 V8 - DesignSpark Mechanical1018Ă—585 67.6 KB

20210925_110928
20210925_1109281920Ă—909 498 KB

20210924_090222
20210924_0902221920Ă—909 396 KB

1 Like
2

This isn’t unusual, it’s part of Arduplane already. It would be a copter style tail sitter. The guidance is here:

Switching from LW-PLA to PETG is going to be an issue. A quick search on google suggests a 50% increase in weight. So that means motor and prop suggestions form the designer are going to be way off. Get a subscription to ecalc.ch and run some of the numbers, just as a normal quad to see what it’s going to take to get it to even hover.

3

Thank you, this is very helpful.

I read through the tailsitter documentation you linked, and my airframe now looks much clearer to me as a CopterMotor tailsitter with no fixed wing control surfaces.

This model has 4 motors total, with all 4 mounted at fixed angles relative to the airframe, and no elevons or other control surfaces. So from the documentation, it seems the correct ArduPilot approach is:

  • Q_TAILSIT_ENABLE = 2
  • copter-style motor layout using Q_FRAME_CLASS / Q_FRAME_TYPE
  • motors providing attitude control even in fixed wing modes, since there are no control surfaces

My main concern now is the PETG weight penalty compared to the original LW-PLA version. The designer used 4S, DYS 2212 motors and 9040 props, but I suspect that may no longer be enough once my final AUW is known.

I have not finished printing yet, but once the build is complete I will weigh all major components and estimate final AUW before locking in the final motor / prop / battery setup. I also agree with your point that I should treat this as a thrust problem first and check whether it has enough total thrust to hover safely before worrying about forward flight.

One thing I am still unsure about is the motor geometry:
this is a 4-motor layout, but all motors are mounted with fixed tilt angles as part of the design.

In your opinion:

  1. Would you still start with a normal Quad X tailsitter motor layout in ArduPilot for something like this?

  2. Do the fixed motor angles create any special setup concern in ArduPilot, or would you first treat it like a normal copter-style tailsitter and only adjust from there if needed?

  3. Would you recommend focusing only on achieving stable QSTABILIZE / QHOVER hover first, and ignoring forward transition until hover is fully sorted?

If you have seen any similar 4-motor motor-only tailsitter setups, parameter files, or logs, I would be very interested.

Thanks again.

4

Follow the guidance in the wiki. Start it as a Quad X in Arduplane. As long as the motor angles are fixed, equal, and don’t move, you should be okay (assuming they aren’t too crazy). I haven’t seen the design so I can’t say for sure.

Get it hovering first. Always start with Q-stabilized, then move to Q-hover. Get that tuned first before you try a forward transition. All too often people rush to transition to fixed-wing mode and things end in a mess. Consider that when you transition back to Q-mode the drone will be moving quickly and there is likely to be instability. If the Q-mode isn’t tuned then you’re risking a crash if it isn’t able to handle the situation yet. 3D printed drones aren’t known for their durability in a crash so you probably don’t want to take chances.

5

Thank you, that makes sense.

So I will treat it as a Quad X copter-style tailsitter in ArduPlane, and I will focus only on getting QSTABILIZE and then QHOVER working well before attempting any forward transition.

The motor angles are fixed and symmetrical, so hopefully that gives me a reasonable starting point.

My next step will be finishing the build, checking final AUW, and then setting up the FC and motor order carefully.

Thanks again.

6

Using a tool like ecalc.ch will probably save you some money in the long run.

7

Thank you, that makes sense.

I think my next step will be to estimate the final AUW and run the numbers in eCalc before I commit to the final motor / prop / battery setup.

At this point I mainly want to answer one question first:
whether this PETG version is realistically viable at all, or whether the weight penalty pushes it into an impractical motor/prop range for this airframe.

If the calculations show that the required setup becomes too large or too aggressive for the printed motor mounts and structure, I may reprint the airframe in LW-PLA instead of forcing the PETG version.

So for now I will treat it as a Quad X copter-style tailsitter in ArduPlane, but I will use eCalc first to check the power system before finishing the setup choices.

Thanks again.

8

If I was to make a bet, I’d say the PETG is going to be to heavy to be practical. You may get it in the air, but it will be maxed out.

9

Thanks again, I think your warning is probably correct.

I now have some rough weight numbers from the PETG printed parts, and they already look heavier than I hoped:

  • center section: 923 g
  • landing legs / motor mount structure: 349 g
  • both wings together: about 350 g

So the printed airframe alone is already around 1622 g, before motors, ESCs, FC, wiring and battery.

The more I add up the numbers, the more it looks like the PETG version may be too heavy to be a practical flying version, even if it can technically be made to fly.

At the same time, I kept looking for similar examples and found a factory-built tailsitter with a very similar general layout and fixed angled motors:

https://www.avito.ru/moskva/sport_i_otdyh/penolet_vtol_bespilotnik_heq_black_knight_gp7_7950270372

I also found another 4-motor ArduPilot tailsitter model example here:

So this gives me more confidence that the overall concept itself is valid. The real problem now seems to be not the concept, but my PETG implementation and its final weight.

At this point I think the PETG version may end up being more useful as a prototype for geometry, layout and setup testing, while a true flying version should probably be rebuilt in LW-PLA or another lighter material.

I still want to run the numbers in eCalc, but right now I suspect your advice about PETG being too heavy is likely correct.

Thanks again — your comments have been very helpful.

1 Like