AVI Hardware: Orange Cube+, Here4 GPS, SF20 range finder for AGL
Empty Weight: 40lbs
Max Payload: 160lbs
The goal is to get to stable flight and have the vehicle go to GPS waypoint stations with no payload and with full payload.
Currently I have conducted a tethered hover flight, collecting data on vibe, without any filters, plan to incorporate the filters on future flights.
Some questions I am hoping to get advice on:
Manual Tune or Auto Tune is a better path?
Recommended starting points for parameters for a vehicle this size? I am starting with the parameters on the “Preparing for Tuning Page”. Fair to assume custom tunes for empty and payload flights? Payload is under main rotor mass a fixed (does not swing).
Any insight or feedback on the log file from the tethered hover would be appreciated! Note filtering was not used to collect the raw data.
I got ahead of myself on the previous post, so I deleted it. I am trying to determine if your collective setup is correct. For some reason, your collective (CTUN.ThO) is decreasing as you takeoff. I see that you must have your collective stick set to spring to center. But why do you leave it centered during your spool up. Is the aircraft just taking off as the rotor speed increases.
This would be fine in altitude hold as the autopilot is handling your collective and keeping it down until you signal takeoff by raising the collective stick above the center position. With Stabilize though, you need to keep your collective low during spool up and then raise it when you are ready for takeoff. It is a bad habit to just let it lift off when the rotor speed reaches the point at which the rotor can lift the vehicle. There are decisions in the code that rely on the collective being low for spool up. the signal to the autopilot that the aircraft has taken off is when the collective is above 50% (hover collective- zero thrust collective). It also relies on the collective position to determine when it is landed. This is when the collective is below H_COL_LAND_MIN. It is very important that it signal to the autopilot that the aircraft is landed.
Sorry I am being a stickler on this but not handling the collective properly can get you into a unusual state and potentially cause a crash.
Hey Bill, thank you very much for the insight! I am using the Herelink Transmitter that has spring centered sticks, also the rotor blade has a cambered airfoil and a washout of a few degrees, so we made out best guess as to there “zero lift” was located at “mid-stick”, but had to manually adjust collective down to have zero thrust, then brought stick back up to take off.
Is it going to be an issue with this style of controller and stabilize mode?
Or do I just need to find “zero thrust” point and make sure that is the neutral “mid stick” point on the transmitter?
I understand the system needs to know when it has taken off and landed so this is very helpful, thank you!
Hey Ferruccio, thank you for the feedback!
Yes, the tail rotor had a lot of authority for this test. The vehicle was physically tethered to the ground with ~6” of slack so I don’t know if that played a factor here? Also for a lot of the test we were “light on the skids” and up and down, sometimes taught on the teathers.
The center point for the tail rotor was set with some simple hand calcs and level gauge, was a best guess but “seemed” to be fairly neutral in the test, did not need a lot of trimming.
What from the chart causes you to ask? The target yaw being much higher than the actual? This implies more Feed Forward is required or a different trim level? I have ATC_ACC_Y_MAX turned down pretty low (18,000).
Flight controller is mounded inside a ~1lb box that is on 6x soft rubber isolators. Given the mounting it is possible the box is vibrating displacement is too much/too soft and needs to be stiffened up. Is that what you are seeing on GyrY/X?
Where do you think the yaw imbalance is coming from?
No. It won’t be an issue but your pilot will have to develop good habits for using the collective when in stabilize mode or any mode that doesn’t do altitude hold
You should set up the stick in stabilize mode so that full down is near but below the H_COL_LAND_MIN so you guarantee that it will declare landed if you are landing in stabilize mode. You use the IM_STB_COL_X parameters to set this up. Typically with my smaller Heli’s I can set up the collective so that min collective is the lower limit of the collective in stabilize mode. Thus I don’t have to adjust the IM_STB_COL_X parameter. I am able to get sufficient descent speed and I can leave the collective down without bad vibrations. I would suggest trying to do the set up that way but if you can’t then we can discuss other ways to handle this.
It would be best to have the collective stick all the way down during spin up (not mid-stick), this is below H_COL_LAND_MIN, vehicle knows it’s on the ground. This neg collective isn’t too much to cause issues on the ground (resonance, etc.). Once at speed, collective is brought up, passes the H_COL_LAND_MIN, positive lift, takes off. IM_STB_COL_X is used to set the collective curve vs stick position to make this happen.
It looks like I can open up the controller and remove the mid-point spring. This feels like a good idea? If the thumb gets off the stick it would be bad to “jump” to mid-point?
It is ok to leave it sprung loaded. Once it is tuned and you have it flying well in loiter or altitude hold then the sprung loaded collective is fine. You should probably set the RC option that requires zero throttle stick. That way it will fail to arm unless the throttle stick is full down. Then in stabilize and other modes you have to pull down the throttle stick for it to arm. Then in stabilize with the sprung loaded throttle stick, center stick should be about the collective required for hover
the fact that most of the flight was performed in HIGE condition (probably at the lightest weight config also?) and still required all of the integrative to keep the yaw axis steady. It is true that you could increase ATC_RAT_YAW_IMAX to give a little more headroom to yaw axis PID controller but I suggest to consider your MTOW condition and environmental effects that will further deteriorate tail rotor authority. Also looking at the mere servo output vs servo travel limit I think you need more authority.
Any particular reason for this arrangement choice? I’m just wondering if you already tried to hard mount the cube and what the gyro plot was in that case. CubeOrange Plus already has internal damping and usually is not great to have double damping setup. The increased mass suspended over rubber mounts can easily lead to excessive oscillations, like in the gyro plot above.
Ok good to know, just so I understand correctly; you see see the yaw was pegged to keep the vehicle stable? I was in a pretty narrow band of pitch travel on the tail, I will open that up to allow for more tail authority?
Correct this was no payload and I will need more from the tail with more payload, will keep an eye on that!
I started with a gas engine, so started extra isolated. I have the box hard mounted now and will have the dynamic filter set up off the main rotor RPM. Will keep taking unfiltered logs and review the acceleration data.
Hard mounted the avionics box, the vibe seems okay and the angular noise seems to be lower.
The yaw imbalance seems to be better, the I still saturates but seems to coincide with when the vehicle is up against the tethers?
Planning to do some manual tuning to get it more under better control before we cut it free from the straps.
Test ended with a nasty tail rotor mode coupling with the trail structure which ripped the tail rotor off, so that was exciting. Needs more stiffness, that is in work now. Hope to test again in ~2 days.
Any other feedback is helpful!
If you could please not share the youtube link out with the world just yet I would appreciate it
Yeah, that is desirable to have a clearer picture. Also, once you get to a stable hover it’s good to have a batch sampling from the flight and set notch filter afterwards.
I suggest to install an RPM sensor and then tie the notch filter to it.
