250g Osprey

[CampRobber]

I imported the wikipedia drawing, scaled it to make the rotors 6", and approximated it with round-ish numbers:

Known weights so far:

Motor FT1806/2380	20 x 2 = 40
ESC LittleBee 20A	7 x 2 = 14
Servo MG90	13 x 2 = 26
Prop 0603	2 x 2 = 4
FC F411Wing	7
Optic Flow / Lidar	3
Carbon Fiber Tube 6mm x 1mm x 350mm	8
Nacelle	3 x 2 = 6
Fuselage	13
Battery 3S850	75

=196g

Stretch goal is 249 grams including wings, fpv and the 3S850, but I'm not sure if I'll get there.

The goal for now is to just get it hovering indoors in badly-designed-bicopter mode, then see how many grams are left for wings.

Known unknowns in no particular order:

• Wing size. I don't think an exact scale Osprey would fly. My reasoning is, without variable pitch props it's not going to be able to both hover reasonably efficiently and also reach the speeds required for a tiny wing to work.
• Airfoil. My thinking is that it would work best with some airfoil that has tons of lift and no pitching moment, ever, but maybe shitty drag. I'm not sure if that exists.
• Props. Going to try the 0603s at least, but sort of wondering if there's something quieter and more efficient and not much heavier than 2.3g.
• Servo performance. Some bicopters I've read about use servo feedback. Or at least are using higher performance servos. Are cheap servos with no feedback going to be a problem?
• Nacelle angular range. I'm planning to link the servo shaft to the spar tube with laser cut gears so I can set the ratio. The MG90 supposedly has 180 degrees of motion. I want as much torque as possible.
• Tail. If the wing is bigger this will have to be bigger and/or further aft. One option is using another carbon spar tube as a tailboom.
• FC software. Does anything actually support this type of vehicle out of the box? I know you can do 2+2 mixes in inav but no idea how transition would work.

 

[CampRobber]

 

[JasonK]

how heavy are the servos are yo using here? I have been working on something similar myself (link should be in my sig).

edit: I apparently can't read well the first time... see my own answer there.

 

[CampRobber]

Lessons learned so far:

• There are two kinds of nearly identical cheap metal gear micro servo: MG90 and S52. S52s melt on 6v.
• Inav has airplane and multirotor output and control modes, and it doesn't seem to be the case that you can choose the 2motor+2servo flying wing output config and still have it controlled like a multicopter. But messing with the target config and rebuilding inav fw isn't a huge deal.
• Confirmed 37 watts for 250 grams of thrust. So if it ever hovers, it'll hover for 10-15 minutes!
• 43 grams for wing + elevator + fpv seems... very optimistic.
• Sam Sheperd built a plane with wings that are just entirely slats. There's a video on youtube. It has tons of drag but incredible lift even at low speeds and high angles of attack. This thing is never going to be able to fly super fast due to the props, and its huge disc area means thrust at low speeds is relatively "cheap". Some I'm wondering if that might be a good wing. I'd like to find some way to simulate it, though.
• I'm going to mess with pitch stability a bit just holding it in my hands, but I think it might be worthwhile to build a one axis gimbal to get that dialed in.

 

[CampRobber]

• Dropping the battery 1/2" improved stability to the point of barely flyable.
• I've been able to hover it for 10 seconds at a time in angle mode. A *good* miniquad pilot could probably keep it in the air.
• In hindsight, swiveling the servos was a bad idea. The gyroscopic forces of the rotors cancel when you swing them together for pitch, but they still have substantial rotational inertia. With upward facing props, it goes the wrong way.
• The laser cut gears I made have some backlash. Even cheap servos have some backlash. If there's one thing I remember from my balancing robot days, backlash = death.
• When doing low-performance takeoffs, landing gear, or even friction constant matters a lot. Ground friction below the CG is destabilizing. Furthermore, actual landing gear would be neat because with the tilt rotors it can taxi perfectly.
• I am going back to the drawing board to see if I can come up with some V-280-style nacelles with spring-pull linkages.

 

[JasonK]

• In hindsight, swiveling the servos was a bad idea. The gyroscopic forces of the rotors cancel when you swing them together for pitch, but they still have substantial rotational inertia. With upward facing props, it goes the wrong way.

the Bi-Copter done by David Windestal had issues. After doing some research into my own version of something this, I found out about gyroscopic progression, which, depending on which way your props are spinning (in/out) you get positive or negative roll coupling with yaw. (tilting the prop forward/backwards causes roll from the gyroscopic progression)

It doesn't seem that Tom Stanton vtols had significantly noticeable issues here, so I suspect the level of coupling is effected by your 'wing span' as his are much wider then David's or yours.
https://www.youtube.com/playlist?list=PLj3Bh6Krv9CXC8L8c3bPgonU7mtuRoIIa

 

[CampRobber]

you get positive or negative roll coupling with yaw.

Yes. I'm expecting to do some kind of yaw/roll mix later, but even with that issue, yaw and roll seem more stable than pitch.

Is yours wired up yet?

 

[JasonK]

nope... sitting in parts on my work bench and plans in my head.

 

[JasonK]

ahh... yes a bi-copter is going to have a 'much less stable' direction, in the format you have, it will be pitch, if you do it forward/back (like the Chinook), then roll will be the less stable direction.

 

[CampRobber]

 

[PsyBorg]

First thing to do get all that added weight off the wing tip like the servo and gear box. Do that internal at the fuselage. Second is to counter balance those motor pods so they pivot with no changes way out on that arm. That alone will make the process MUCH easier so the fc does not have to fight those two forces. Tuning would become much easier not to mention having the weight on centerline and low would help stability even more.

 

[CampRobber]

First thing to do get all that added weight off the wing tip like the servo and gear box. Do that internal at the fuselage. Second is to counter balance those motor pods so they pivot with no changes way out on that arm. That alone will make the process MUCH easier so the fc does not have to fight those two forces. Tuning would become much easier not to mention having the weight on centerline and low would help stability even more.

Why do you think the yaw/roll moments are so important? It's already stable in those axes and it's not intended to be highly maneuverable.

 

[PsyBorg]

Why do you think the yaw/roll moments are so important? It's already stable in those axes and it's not intended to be highly maneuverable.

Stacking compensation on pitch axis. As those motors tilt the torque used has nothing to push against so the fuse will tilt slightly. Equal and opposite reactions and all that. As the gyros work to keep the motor tilted where you want it to be an oscillation on pitch axis builds because there is nothing but drag from air to push back. This is where side by side bi copters get wonky. You wont see or feel the pitch changes as the motors rock back and forth but the gyros and accelerometers do. Its just the opposite for in line bi copters like in a Chinook. They suffer less because they use a cyclic but that reaction is still there in the super sensitive gyros and accelerometers.

Putting the servos and gear box in the fuse will allow more mass on pitch axis to have to move and thus fight this at least a little.

I was also thinking maybe a little deadband on pitch could help / cure this tendency but have no way to test it. Also maybe float the battery tray to absorb the changes like they do in high rise building with the stone slabs on oil to counter vortex shedding. I would have to make a model to observe changes to figure how to compensate.

 

[CampRobber]

In order to appease the old-timers here I've decided to replace the FC with something else, and it hovers great!

 

[Pieliker96]

View attachment 187963

In order to appease the old-timers here I've decided to replace the FC with something else, and it hovers great!

It's so cute! The lil lego test pilot is the cherry on top.

 

[CampRobber]

 

[JasonK]

angle or rate mode?

 

[CampRobber]

angle or rate mode?

Rate. Angle might work better but right now it hovers at ~+12 degrees pitch or something so I'd have to figure out how to add a bias. (I'd been hoping to get away with less servo throw by doing that.)

 

[JasonK]

iNav has a setting that lets you set the 'zero' pitch, I actually used that before trying my hover by balancing it at the motors, then seeing what iNav said the angle was, then set the correction factor to make that zero.

if your hovering at that much pitch, your CG isn't right under the spar, but I suspect you know that.