Tiltrotor support for plane #2

Hello,

I ask for some help as I have a couple of problems I’ve been trying to solve without success. I’m flying a tilting quadrotor with only the 2 front motors tilting.

The first issue is that it tends to yaw at the beginning of the transition (with 2 front motors tilting) and sometimes it starts spinning around without control. If I tilt the four motors, the problem seems to dissappear, but it is not optimum as it requires more power to cruise. Recently I activated the vectored yaw control for the 2 front motors, and it still yaws at the start of the transition, but it manages to rectify the transition and complete it, but sometimes in an awkward way.

I was thinking about maybe changing the frame type from “X” to “H”.

The other problem, which I’m more concerned, is that after the back-transition sometimes I loose the throttle command. Despite applying full throttle, the vehicle starts to descend at quite a high speed. This happened at least 4 times.

Here is the log of the last time it happened:

Regards

Yaw: It sounds like one of your motors pulls more than the other or rotates faster than the other?

H frame was to offset frame twisting due to torque, so the motor rotation then opposed the twist rather than exacerbating it.

Descending: Maybe the weight overcomes the lift at that point or the props stall, any unusual noise when it happens?

Anyone have some VTOL tips to share? What advice can you give to someone flying his first VTOL?

Hello,

I have checked the motor output and you were right, motor 1 seems to give more power during the transition, but while flying in copter all 4 have similar output. I will try to figure out how to solve it.

And about the uncommanded descending, the log show that the motors were not at maximum and I did not hear any noise. If it was a propeller stall problem the autopilot should raise the motors power when it detects the descent.

My advice to fly an VTOL, the most important is that you make sure the vehicle flies flawlessly as a copter. This way you can always switch to QHover or QStabilize mode if you encounter any problem.

We have done the Nimbus 1800 (tricopter vectored yaw) maiden flying on 6S 18650GA’s (Titan packs).

It was perfect, an anti-climax really, I switched nervously into FBW-A and was amazed by the fluid, easy transition into forward flight, completely without fuss for both transition directions - a testament to the awesome and diligent work Tridge and the others involved have put into this firmware. Thank you!

Had some sudden yaw twitches but suspect that’s due to a compass interference issue we were having. Also the distance traveled after transition back to hover was more than expected but that’s just something to get used to.
There is a yaw oscillation that I need to sort out but that’s just some tuning.

Here a bit of video:

Thanks again guys!

2 Likes

Congratulations :trophy:

1 Like

Thanks Rolf, plane is mostly RTF so I didn’t have to do much. Just held thumbs that everything worked as expected.

Have asked a question here: VTOL RTL behaviour? so as not to clog up this thread.

1 Like

Anybody know if the current code could support a quad motor tailsitter? For example the plane would be pointing up with 4 motors and it would hover like a quadcopter in qstabalize then when switched to fbwa the two front motors would possibly slow down until the drone tips horizontally and the transition completes then be flying like a regular plane but with 4 motors providing forward thrust?

1 Like

You said mostly RTF. Is this from FOXTECH? I’ve been thinking about purchasing a VTOL NIMBUS from them.

1 Like

Foxtech yes, mostly RTF, yes. It’s not something I’d just throw a battery in and fly. Our airspeed sensor didn’t work at all out the box and had to be replaced, the battery tray and fastening system is not good at all. Pitot tube tip was loose (we nearly lost it) and the pitot tube itself is very loose too.

The airspeed sensor was loose inside the nose and the Nimbus nose plug/socket system can’t be used as the plane can’t be booted without the airspeed sensor connected as it’s then not recognized and you can’t plug the battery in with the nose cone and airspeed sensor on. So a rethink had to be done.

There was also noticeable tail wag, visible too in their videos.

Otherwise it’s a good platform, wingspan a bit small for the weight it carries, I’d have preferred at least 2.2m not 1.8m with the 10Ah battery recommended. The fuselage is unfortunately very small, not easy to get even a medium camera in without major surgery.

Oh, and the Pixhawk Cube’s internal compass had to be disabled as there is too much interference around it, kept getting compasses inconsistent errors.

2 Likes

Hi Graham, thanks for the insightful details on your mods. I was just wondering how well the plane flies, your AUW, the battery you use and the endurance you typically get?

1 Like

@Aeroakram, it flies very well so far, there’re a couple issues which I’ll detail below. Battery is a 10Ah 6S, we tried 10.4Ah 18650GA cells but they don’t do well with high currents, weight - not sure, around 3.5KG I think, endurance not tested, max 32 minutes so far.

We’ve had a few good flights with our Nimbus 1800 but there are some behaviours which don’t seem right. Below is just a few observations from the few flight we’ve had.

First was inability to hold position in wind, this was just after RTL transition, I had to switch to stabilize and carefully nurse the plane closer to a safe area to land. Plane seemed unable to pitch forward to hold its position in AUTO/RTL, but was able to do so in Stabilize, desired pitch (NavPitch) did not request a higher pitch angle. Why? Also, would some tilt motor assistance be possible & work in this situation? (log: https://drive.google.com/open?id=1ysovtcK9OcdtuirmuX2-1eJrqi0am5Bn )

Then we kept having a huge altitude loss just after takeoff transition, somewhat heart stopping from the observers viewpoint as the plane appears to be heading straight for a collision with a planet, from 31m to 10m! in one takeoff.
Have changed Q_TRAN_PIT_MAX from default 3 deg to 7 deg and this also seems to have slowed the mad increase in speed (>24m/s, cruise is 13m/s) which was apparent with the altitude loss. Will try 10 deg next.

Still keen to have the plane descend diagonally down from ALT_HOLD_RTL just to save battery, rather than straight down vertically.

Learning more every day with this impressive system.

2 Likes

Thanks @Graham_Dyer. FYI, For Kris’s tiltrotor conversion of Nimbus 1800, he uses SONY 18650 VTC6 6S5P 15000mAh Li-ion battery, which apparently gives him up to 90 minutes of flight time (https://www.youtube.com/watch?v=arQNWQrE06w). Perhaps, this setup would be able to handle the current draw reasonably okay.

I’m trying to design a quad tiltrotor with vectored yaw that can also use the tilt mechanism as pitch and roll. The plan is to have the wings tilt with the motors and be able to control them all independently. It’s redundant to have ailerons/elevators on that type of aircraft and it would simplify things a great deal to only have 8 moving parts. From what I’ve seen, it looks like arduplane supports control of vectored yaw already but not pitch or roll. Anyone have any Idea how hard that would be to add?

1 Like

Its would seem I am not the only one having issues with the tilt rotor config. I hope that somebody can assist me. My vtol is similar to the Nimbus config except that it is a flying wing. Aircraft fly OK in copter modes, apart from some yaw oscillation. The issue I have is with transition. It starts to perform the transition fairly good and right at the point where it recons it has achieved minimum speed and completes the transition to forward flight the aircraft starts to yaw violently into a spiral dive to the ground. This happened twice now with the last attempt to recovery in qloiter resulting in my blades hitting the air frame (motor angles too high…) Anyway according to the log it looks as if my two front motor outputs are the same so that would not explain it. My only theory is that it is the lack of an airspeed sensor and that the airspeed it too low at point of full transition, thus stalling. My last log can be found here: https://www.dropbox.com/s/5b7ei1s6xwzsaef/00000109.BIN?dl=0

What happens to the elevons when in FBW-A and you roll then plane side to side? Do they oppose the direction of roll?
Your plane reaches 14m/s GPS speed which should be enough to not stall, but it rolled violently just after that, twice. You also briefly switched to Manual which didn’t help (best not to have manual as a mode).
Double check the elevon direction for both roll and pitch.

Thanks Graham for taking a look. I have double checked and my elevons are indeed correct. (for what it is worth I have three conventional wings that all fly well) I have noticed that some trim was required for level deflection but nothing that could induce that kind of roll. I however note that the airspeed seem to drop after transition speed is reached and the main inducer of the roll is initiation of yaw/spinning. Here is the log for the first attempt: https://www.dropbox.com/s/p1juqkhx6lxrwcg/00000105.BIN?dl=0
ps only reason I chucked manual in was that in worst case scenario I should be able to fly manual just with two front props stuck tilted down and direct control…that however did not pan out the way I imagined :slight_smile:

The GPS speed dropped due to the change from horizontal speed to vertical speed, unfortunately there’s no airspeed sensor so we can’t see what that did. I remember a comment a while back in a thread somewhere that airspeed is almost mandatory for a VTOL as it’s not so easy for the FC to always estimate airspeed correctly. Also the motors stop dead in Manual mode so best to just have Stabilize and FBW-A, then If something goes wrong just go back to Stabilize.

If the plane yawed uncontrollably then usually, either the vertical stabs aren’t enough or the motors are uneven. If it rolls violently then often the elevons are the wrong way round accelerating the roll.

Which channels are your front two motors on? And a photo or 2 of your plane?

I also seem to remember reading about the airspeed sensor somewhere, so perhaps a next step. I also agree that with a wing config (little vertical stabs) and with my motor spacing being very wide the air frame config is challenging for this setup. By the way my motors are on 6 and 7.
I included a picture here: https://www.dropbox.com/s/kt2xvot7wnirudm/rvvtol.jpg?dl=0
Perhaps I should just get a nimbus or equivalent. I like the tri setup with only having to haul one additional motor as parasitic drag during the fixed wing flight profile (which is 99% of flying time) as opposed to the quad setups with a 5th motor for forward flight. Thanks again

If you look at the tailsitter development page a guy named lorbass had a similar issue I believe. I think what fixed his was that he hadn’t added the wing end tips. his post is close to the end of that blog.