Full VTOL F-35 Foamy

[NickRehm]

Hi all,

Here's my most recent quarantine build while I'm back home with nothing else to do: A full VTOL capable foamy F-35

Specs:
Wingspan: 36"
Length: 44"
Weight: ~900 grams? (haven't weighed since I painted it)

Flight controller: teensy 4.0 and mpu6050, custom software
Motors: lumenier rx2204 2300kv
ESC: lumenier 35a 4-in-1
Battery: 1500mah 4s 40c
Yaw tilt servo is just an sg-90 along with the elevons. Flap tilt servos are standard size spektrum. Flight time is about 5 minutes assuming half of that is in hover. Airframe is a 10mm poplar skeleton with 9mm EPP foam. Some 3mm carbon tube in key areas for additional support

Two rear motors are directly mounted to the flaps "airhogs" style. These flaps actuate down for hover along with the front lift fan, and are fixed in forward flight:

Pre-electronics and paint:

Undercarriage:

Custom flight controller setup:
View attachment

And with paint!

CG is a little forward toward the front lift fan, so that motor is working a little harder in hover--I'm looking to upgrade that motor to something a little bigger. My 'intermediate' flight mode just consists of actuating those flaps to about 45 degrees so it can pick up some forward speed. Forward flight is just elevon control and the front fan cuts off (shown in the first video which is why it pitches down going into forward flight, I have since updated the code to fade the front motor out and am waiting on a flight video of that).
One more video of early indoor hover testing:

Any other VTOLs here? Happy to answer any questions, or nerd out about other cool VTOL platforms

Cheers

 

["Corpse"]

Hi all,

Here's my most recent quarantine build while I'm back home with nothing else to do: A full VTOL capable foamy F-35

Specs:
Wingspan: 36"
Length: 44"
Weight: ~900 grams? (haven't weighed since I painted it)

Flight controller: teensy 4.0 and mpu6050, custom software
Motors: lumenier rx2204 2300kv
ESC: lumenier 35a 4-in-1
Battery: 1500mah 4s 40c
Yaw tilt servo is just an sg-90 along with the elevons. Flap tilt servos are standard size spektrum. Flight time is about 5 minutes assuming half of that is in hover. Airframe is a 10mm poplar skeleton with 9mm EPP foam. Some 3mm carbon tube in key areas for additional support

Two rear motors are directly mounted to the flaps "airhogs" style. These flaps actuate down for hover along with the front lift fan, and are fixed in forward flight:
View attachment 168688

Pre-electronics and paint:
View attachment 168689

Undercarriage:
View attachment 168692

Custom flight controller setup:
View attachment 168690

And with paint!
View attachment 168691

CG is a little forward toward the front lift fan, so that motor is working a little harder in hover--I'm looking to upgrade that motor to something a little bigger. My 'intermediate' flight mode just consists of actuating those flaps to about 45 degrees so it can pick up some forward speed. Forward flight is just elevon control and the front fan cuts off (shown in the first video which is why it pitches down going into forward flight, I have since updated the code to fade the front motor out and am waiting on a flight video of that).
One more video of early indoor hover testing:

Any other VTOLs here? Happy to answer any questions, or nerd out about other cool VTOL platforms

Cheers

That's the coolest thing I've seen all week! How did you code your own software for the FC?

 

[NickRehm]

That's the coolest thing I've seen all week! How did you code your own software for the FC?

Very carefully
In all seriousness, the flight control code was developed from scratch for some of my research because I was sick of searching through the spaghetti mess of code that is ardupilot/betaflight/etc. Too many features I'll never need to use that get in the way of the core functionality of the flight controller. I'm using cascaded PID (rate on inner loop and attitude on outer loop) for stabilization. Here's a quad I built running this same flight controller (minus the VTOL mixes) with the addition of an onboard flight computer running ROS for some autonomous capabilities:

This is the generic framework inside the FC:

 

[JasonK]

1 - looks really good
2 - is there a good reference for the PID loop? that you used (looking to do something similar, but only using 2 motors for everything).

 

[NickRehm]

1 - looks really good
2 - is there a good reference for the PID loop? that you used (looking to do something similar, but only using 2 motors for everything).

Thanks! This is a good place to start for basic basic PID implementation: http://robotsforroboticists.com/pid-control/
It really depends on what you're trying to do and what type of performance you want. For example, basic rate setpoint is the easiest to get running because you can just use the raw gyro data as your feedback. Angle setpoint gets trickier with attitude estimation requirements and sensor noise issues you must deal with either through tailored filtering or cascaded PID

 

[JasonK]

perfect thanks... I have an EE w/ Computer Engineering work with code sometimes more then I want. but it has been a while since I did any control system loops, just need a refresher to get started.

 

[NickRehm]

In all honestly, the most time consuming thing to setup is your data input/output and the associated troubleshooting (IMU, radio, and actuator outputs). All in all, less than 500 lines of code will get a quad flying

 

[JasonK]

In all honestly, the most time consuming thing to setup is your data input/output and the associated troubleshooting (IMU, radio, and actuator outputs). All in all, less than 500 lines of code will get a quad flying

yup, something in the 'this is about how hard it should be' in the back of my head was in that ball park also. This is my meanderings so far: https://forum.flitetest.com/index.p...80-valor-bell-xv-15-or-the-v-22-osprey.63247/.

 

[JasonK]

ohh... one more question, were did you get 80Hz for your low pass filter and what is the cutoff shape look like? or was this a FFT and just dropping any terms above the 80Hz? was this related to self vibration from the motors?

 

[NickRehm]

yup, something in the 'this is about how hard it should be' in the back of my head was in that ball park also. This is my meanderings so far: https://forum.flitetest.com/index.p...80-valor-bell-xv-15-or-the-v-22-osprey.63247/.

Oh awesome, you should have no difficulty getting a bicopter like that up and running. Once you have the framework to fly a quadcopter, the additional code for a transition vehicle like that is less than 15 lines

ohh... one more question, were did you get 80Hz for your low pass filter and what is the cutoff shape look like? or was this a FFT and just dropping any terms above the 80Hz? was this related to self vibration from the motors?

The IMU filtering is just a first order low-pass with cutoff frequency targeted at 80Hz (pretty tame roll-off so realistically up to 150 Hz still comes through). With the dynamics of the vehicles we're flying, anything above that is more or less just garbage from motor or aero vibrations. And cutting any lower starts to impart significant time delay with respect to the loop rate. Start with cutoff at 200Hz and work your way down until control response starts to suffer--that was pretty much my approach

 

[NickRehm]

Just got out to fly her again. Upgraded the front lift fan to something a little larger so it doesn't have to work as hard in hover. Also modified my transition code for smoother transition into forward flight. The front fan remains active and progressively fades out over ~2 seconds when entering forward flight. I'm pretty happy with the results

 

["Corpse"]

Just got out to fly her again. Upgraded the front lift fan to something a little larger so it doesn't have to work as hard in hover. Also modified my transition code for smoother transition into forward flight. The front fan remains active and progressively fades out over ~2 seconds when entering forward flight. I'm pretty happy with the results

That looked much smoother! Nice work man!

 

[JasonK]

transition looked much smoother this time.

 

[CrazyGuy]

This is very cool! I'd love to build one, would you share the plans and code?

 

[NickRehm]

Went out last night for another flight. Didn't have any pavement to takeoff from (grass gets caught up in the props) so just took off from the truck bed! VTOL is awesome

 

[NickRehm]

Just curious, is there any interest in plans? I would like to release the code, parts list, part files (tilt mechanism, motor mounts, etc.), and plans at some point when I feel the design is ready. Here is the next, slightly smaller version at roughly 22" span:

 

[CrazyGuy]

Just curious, is there any interest in plans? I would like to release the code, parts list, part files (tilt mechanism, motor mounts, etc.), and plans at some point when I feel the design is ready. Here is the next, slightly smaller version at roughly 22" span:

View attachment 170271

You already know my answer

 

[JasonK]

Just curious, is there any interest in plans? I would like to release the code, parts list, part files (tilt mechanism, motor mounts, etc.), and plans at some point when I feel the design is ready. Here is the next, slightly smaller version at roughly 22" span:

We only keep questioning you on how you did it. (so yes)

 

[NickRehm]

Got started on the v2 build today. Everything went together very nicely and I'm gonna add paint before installing electronics over from version 1 in the next day or so. Airframe is made from dollar tree foamboard and hot glue; weight is 132 grams pre-paint and pre-electronics

Gonna get some flights in, make some tweaks to the design, then I think plans will be ready

 

[NickRehm]

Got a little excited and did paint + decals this evening: