Thread Tools
Mar 17, 2015, 06:16 PM
Registered User
Ran D. St. Clair's Avatar
Quote:
Originally Posted by JulietKiloMike
Here's a video of a motor run-up, if any of you guys are interested. The motors seem to spool up at the same time and run at similar power levels...I wonder what's causing the leftwards yaw! https://www.youtube.com/watch?v=4EVP...ature=youtu.be
I can hear a beat frequency of 1 to 2 hz, which is good, it means the motors are running very close to the same RPM.

You have something very slightly crooked in the airframe, but it is being magnified by the general yaw instability. Once you add the vertical stabalizer area you should be fine.
Sign up now
to remove ads between posts
Mar 17, 2015, 07:31 PM
Registered User
Quote:
Originally Posted by Ran D. St. Clair
I can hear a beat frequency of 1 to 2 hz, which is good, it means the motors are running very close to the same RPM.

You have something very slightly crooked in the airframe, but it is being magnified by the general yaw instability. Once you add the vertical stabalizer area you should be fine.
Hey! Who needs a spectrograph when RnD's in the house. He pretty much nailed it. Lots of noise and some harmonics, but no clear distinction between the two motors which seem to be running around 6500 rpm.
Interesting to notice the slope difference between spin up and spin down.
Mar 17, 2015, 07:38 PM
Winging it >
leadfeather's Avatar
Here is a quickie test you should be able to do to determine whether you do or you don't have asymmetric thrust.

One guy holds the plane pointing straight up. Fast forward flight mode selected. The other guys carefully throttles up slowly. The guy holding the plane should easily be able to tell if the thrust is equal. Be careful (of course) if you decide to do this.
Mar 17, 2015, 08:01 PM
JulietKiloMike's Avatar
Thread OP
Quote:
Originally Posted by PlumbBob
Hey! Who needs a spectrograph when RnD's in the house. He pretty much nailed it. Lots of noise and some harmonics, but no clear distinction between the two motors which seem to be running around 6500 rpm.
Interesting to notice the slope difference between spin up and spin down.
Cool! Thanks for the analysis guys I'm glad to hear that you guys agree that the motors seem to be running in sync...it's always great to have a second set of eyes (or, in this case, ears). Extra vertical stab area, here we come!
Mar 17, 2015, 08:05 PM
JulietKiloMike's Avatar
Thread OP
Quote:
Originally Posted by leadfeather
Here is a quickie test you should be able to do to determine whether you do or you don't have asymmetric thrust.

One guy holds the plane pointing straight up. Fast forward flight mode selected. The other guys carefully throttles up slowly. The guy holding the plane should easily be able to tell if the thrust is equal. Be careful (of course) if you decide to do this.
We tried doing that with the aircraft on the ground during the run-up today, but neither my classmate nor I were able to feel a discernible pull in either direction. It feel a little sketchy running the engines up while holding the model in the air, but we might do that if we have to. Right now it doesn't seem that there's a huge issue with asymmetric thrust, so for now I think we'll stick with adjustments to the tail and the addition of some yaw stabilization through KK2.1 programming. If that doesn't seem to help, we'll give the asymmetric thrust idea another look.

It also might be worth noting that we re-measured the distance of each nacelle from the fuselage centerline, etc...all seems to be well in that department, too
Mar 23, 2015, 09:00 PM
JulietKiloMike's Avatar
Thread OP
Here's a video of the sketchy wing-tilt mechanism we currently have. It's a bit wobbly, and I'm thinking that we might have to switch to something with a little less give.

Right now the wing-tilt mechanism consists of an HS-645 MG high-torque servo running on a servo stretcher to bring it through a full 180 degree arc, with some carbon tube and foam bits added on to prevent flexing.

Foamboard Trirotor Tiltwing VTOL Wing Tilt Test (0 min 20 sec)
Mar 24, 2015, 03:01 PM
Registered User
Ran D. St. Clair's Avatar
Quote:
Originally Posted by JulietKiloMike
Here's a video of the sketchy wing-tilt mechanism we currently have. It's a bit wobbly, and I'm thinking that we might have to switch to something with a little less give.

Right now the wing-tilt mechanism consists of an HS-645 MG high-torque servo running on a servo stretcher to bring it through a full 180 degree arc, with some carbon tube and foam bits added on to prevent flexing.

https://youtu.be/RUo6jf8mEUw
180 degree servo travel is a bit much, but it has advantagees if you can make it work. In theory it means that there is no torque feedback to the servo when the wing is in the vertical or horizontal position. Even if that theory doesn't work out perfectly it at least mean that you have a large mechanical advantage near the ends of the travel.

I suggest you use that mechanical advantage to your advantage. Specifically, when the wing is in the flat or FFF position the leading edge of the wing should be hard against the fuselage supports from below. This implies there will be little or no wobble in any direction.

You can do the same thing when the wing is in the vertical position but it is probably not necessary. The wing will be lifting the fuselage and the slop in the tilt bearings will mostly dissapear. Also, with the motors under and ahead of the wing tilt axis, they will be pushing the front of the wing up, which will compress the wing tilt servo linkage and take out all the slop in that direction.

Slop in general is not a good thing, and you should remove it where possible, but it is probably not a problem for your design. It looks to me like the main source of slop is the wing tilt bearings. It may be difficult to do much about that at this point without a major rebuld. Next time take care to make sure the shaft is a snug (but not draggy) fit in the bearings, and that the bearings are as far apart left to right as possible.

Servos will always have a slight amount of backlash in the gears, and linkages will have slop as well. Once again, I don't think it is a huge problem for your design, but for future reference, the longer you make the lever arms at both ends of the linkage the less the linkage slop will translate. The servo slop is what it is, but if there is natural pressure or tension on the linkage it will make it so it doesn't matter. Where you get into trouble is if the pressure on the linkage is balanced or oscillates from positive to negative. In that case the surface, or wing in your case, will bounce back and forth between the limits, and if that goes on for long it can build into an oscillation that will destroy all kinds of things, including your entire aircraft. This sort of thing is most agrivated by airspeed, so make sure the wing is solidly locked down in the FFF position.
Mar 24, 2015, 04:23 PM
JulietKiloMike's Avatar
Thread OP
Thanks Ran. D! I'm going to CAD some hardware for the wing pushrod pivot and will add some foam to take out the slop in the vertical flight position. Hopefully with the addition of the pushrod hinge things will tighten things up a bit, including in the FFF position. Updates shall be forthcoming!
Apr 13, 2015, 12:25 AM
JulietKiloMike's Avatar
Thread OP
Well, due to some problems with Hobbyking we won't have the stuff we need to get the KK2.1 working next week, so we'll have to focus on more fixed-wing testing. I'll post a video of the engine run-up when I have a chance, but for now I'm thinking about doing some throttle mixing to get differential thrust working (we really, really need some yaw control). Do you guys have any recommended settings for this kind of thing?
Apr 13, 2015, 11:48 AM
Registered User
Ran D. St. Clair's Avatar
Quote:
Originally Posted by JulietKiloMike
Well, due to some problems with Hobbyking we won't have the stuff we need to get the KK2.1 working next week, so we'll have to focus on more fixed-wing testing. I'll post a video of the engine run-up when I have a chance, but for now I'm thinking about doing some throttle mixing to get differential thrust working (we really, really need some yaw control). Do you guys have any recommended settings for this kind of thing?
Plus and minus 10% is a good place to start. If you use only rudder (stick) as an input this should be fine. On my various aircraft that use differential throttle for rudder, I also often use P feedback in yaw as well. This is great for smoothing out oscillations like your possible dutch roll problem. It can also cause oscillations if the P feedback is too high. I would start with a value of 30.
Apr 13, 2015, 11:55 AM
JulietKiloMike's Avatar
Thread OP
Thanks Ran D! We don't have the KK2 up and running yet, so this is just a differential mix on the DX6i The DX6i only has two mixes, so right now I have it set to vary the left motor's speed based on rudder by 10%, ad you recommended. We'll see if this works as a differential thrust proof of concept, and will hopefully have the chance to implement yaw damping on the KK2 if we don't total the aircraft before then...
Apr 13, 2015, 12:04 PM
JulietKiloMike's Avatar
Thread OP
Video of the wing tilt test:
DTF Tritotor Tilwting VTOL Motor Run-Up and Wing Tilt Test (4 min 0 sec)


The wing tilting assembly seems to hold up to a high-speed tilt at full throttle, which is good. There's a bit of slop in the pushrod, but we've now added a small wood wedge in the point where things were flexing, so hopefully that sorts things out
Apr 13, 2015, 02:47 PM
Registered User
Ran D. St. Clair's Avatar
You will find that you want the wing tilted forward by 5 to 10 degrees for a level fuselage hover. This is both because of any flex in the mechanism and also because the camber of the wing will redirect the thrust slightly.

When actually hovering it doesn't make much difference because the fuselage will just hover a bit nose down, but it does mean that you will tend to go backwards initially on takeoff, and it will also make it harder to make headway into even a slight wind. Tilt wings are bad in that respect anyway, so bias the wing to tilt forward a bit not backwards.

Also, when hovering avoid going backwards at more than 1 or 2 MPH. If you go backwards faster the negative lift on the horizontal stabalizer can overcome the control power of your back motor and it will go into a divergent situation resulting in a back flop loop. If you don't have altitude to recover it is a guaranteed crash.

Finally, in forward flight position, have your servo drive the wing up against the mechanical stops under the wing leading edge. It is not good to stall the servo like that so arrange the linkage so it is at its point of greatest mechanical advantage. That way the holding torque on the servo is minimal.

Bear in mind, when hovering you will be at about half throttle, not full throttle. When in forward flight you will also be at a modest throttle setting most of the time, and the propeller won't be pulling forward all that hard because it is pitch limited.
Apr 13, 2015, 05:03 PM
JulietKiloMike's Avatar
Thread OP
Thanks for the tips Ran D! I think you've saved us a few crashes with that one post

Right now the wing-tilt servo seems to be about as good as we can get it without a structural re-design, but we might move it forward a bit in the next revision so that the wing is tilted forward a little more during hover and horizontal flight. It doesn't quite jam the wing against the mechanical stops during horizontal flight, but there isn't any major wing deflection up to 7/8 throttle or so (at full throttle, the wing deflects by about 10 degrees, now that we've stiffened the pushrod assembly). We're going to try a test flight at Rancho San Antonio today...I'll let you know how things go! (feel free to join us, too! We'll be there around 4:30ish )
Apr 14, 2015, 01:44 PM
JulietKiloMike's Avatar
Thread OP
Video of yesterday's test flight:
DTF Trirotor Tiltwing Fixed-Wing Test Flight and Soft Crash (3 min 26 sec)


We had to do some significant tweaking with a large amount of right aileron and some down elevator to get the aircraft trimmed straight-and-level. The differential thrust attempt didn't work very evenly, (there was a lot of altitude loss on the left-hand-turns), but that was mostly because only the speed of the left motor was varied due to mixing limitations on the DX6i. Hopefully we'll have things working more evenly once the KK2.1 is up and running.


Quick Reply
Message:

Thread Tools