Japanes Ball Drone. - Page 23 - RC Groups
Thread Tools
Oct 15, 2012, 09:22 AM
Registered User
Hello Brandano,

Once again I agree with all of your comments. Ben made a good choice selecting the Ardupilot 2.0 for his spheres, it maybe difficult for him at this point to iron-out the PID settings and going backwards to simple gyros may allow him to resolve fundamental aerodynamic problems without the complexities of a sophisticated autopilot. I strongly suggest that NO "D" term be used

Later Ben can add the IMU add-on board that has a magnetometer, this will give him an external reference point to auto-level his craft. I'm certain the original Japaneses Drone uses a magnetometer too!!!

Look here>> http://store.diydrones.com/ArduIMU_V...arduimu-30.htm

Reflecting back on my comment #2: Stability is defined not only in the ability of the craft to return to level flight as you mentioned but also resist any change in local position or orientation. A large prop creates a great deal of gyroscopic stability that will hold a relative position until the pilot commands a change in flight attitude.

Take a look at my redesign of the stability system for the Austrian Turboplan (English spelling Turboplane). Here Hank Renz is using my design to spin 3 weights mounted on a prop extension, propeller blast drives the weights via air vanes much like a free turbine on a helicopter engine which gave me the original idea.The weights spin in the same rotational direction as the propeller, thus they have an additive gyroscopic force. This is the entire system for stability control of the pitch and roll axis, a HH gyro is used for Yaw control, but it is not really need either. The craft is very stable. My point here is that the spheres will get a form of stability by spinning a single large diameter prop, using a counter rotating prop system completely eliminates this added stability. Sam proved this in the early design stages of his craft too.


RC UFO Turboplan V7 (0 min 20 sec)

RC UFO Vtol Turboplan V6 (0 min 16 sec)
Last edited by corocopter; Oct 15, 2012 at 11:12 AM. Reason: Clarified the spinning weight direction of roation
Sign up now
to remove ads between posts
Oct 15, 2012, 01:47 PM
Registered User

Flying Can UAV

Here is a simple craft, the designer emailed these fact to me.

3 piezo gyros, no HH used. (I'm not so sure I believe that no is HH used, but that was his claim in the email, he said he had problems with HH and did not choose to use that feature, thus it is using only the P term for all stability control)
Digital servos,
Polished carbon sheet -very thin- (mirror finish); sections are cut out out to lighten the craft, then it is rolled into a tube and glued.
Large single propeller of very light weight design
Lipo power.
End designer notes:

After looking at all of his U-tube postings of this device, it looks like it has only 4 control vanes, maybe Brandano or others, can take a look and report back what they see.

The craft is very stable, fairly easy to construct and it looks very structurally strong and rigid. I'd say this would be a good starter design for anyone building a sphere.

Other than "roll on to the ground landings", the sphere shape has no real utility and it severely limits "constricts" the distance between the CG and the control surfaces, thus making the sphere very difficult to control and dampen oscillations.

"The Can UAV" has many fewer parts to build and break, and it looks as stable as the original Japaneses sphere, but it uses very low cost parts, no counter rotating prop is used.


UAV Test Flight - Flying Can Design 2009 (3 min 15 sec)

UAV Test Flight - Flying Can Design 2009 (1 min 26 sec)

UAV Test Flight - Flying Can Design 2009 (2 min 0 sec)
Oct 15, 2012, 02:44 PM
Registered User

Flying Can UAV email details

I found the email response to my questions from the Flying Can UAV designer.

Here is the response he sent me. I responded to this email with a few more question but sadly he did not respond to my second list email questions.

Listed on the U-tube page he noted that 3 piezo gyros are used.

Please note his comment about using a light weight prop to reduce gyroscopic forces. I'm sure his intent here is to reduce yaw motion from motor/propeller mass with power changes, perhaps not to reduce gyroscopic stability as I suggested in my previous email today.



Rai41478 has sent you a message: Date 3-03-11

Re: Flying Can UAV

Hi thanks for the interest.
The Servos are digital.
The gyros are all set to rate, I had a lot of issues with heading hold on the yaw axis.
The entire frame and shell of the craft is constructed of carbon fiber.
The spars and frame are made from carbon square tube with epoxy glue.
The shell is mirror finished carbon fiber sheet, The use of this material is very important to reduce weight and maintain rigidity of the craft.
Brush-less motor, Li-Po battery pack, use the lightest prop you can find to reduce gyroscopic effect. I'm using cheap light plastic 3 bladed props. also balance the craft in the middle of the prop to save yourself headaches.
Have a new version of both craft posting new videos as soon as the weather gets nice.
Oct 15, 2012, 03:29 PM
Registered User
Originally Posted by corocopter
Rai41478 has sent you a message: Date 3-03-11

Re: Flying Can UAV

also balance the craft in the middle of the prop to save yourself headaches.
Please note Rai411478 commented "balance the craft in the middle of the prop to save yourself headaches".

From my observation, this craft looks like the motor is placed in the center of the tube and is a pusher.

#1. Pushers are more stable in most circumstances.

#2. Having the CG on the center-line of the prop reduced asymmetrical forces transmitted to the airframe associated with gyroscopic forces created by the spinning mass if the prop and motor rotor. This is a very good design feature! for the Flying Can UAV, it could work for the spheres, but would require fundamental design changes, but maybe worth consideration.

Oct 15, 2012, 04:12 PM
Registered User
Please reference the attached PDF document, Page 2, Figure 2 and Figure 3 . CG location of Drones.

Notice the location of the CG in line with the propeller center-line. This is very important for mitigation of various gyroscopic forces associated with both power application>>>> i.e. torque induced yaw motion,<<<<and >>>>pitch and roll coupling to the propeller forces<<<<. These craft have fairly large diameter props compared to our little models thus these forces are of concern.

The Flying Can UAV uses this concept very well.

Oct 16, 2012, 12:21 AM
Registered User
Hello Kelly and Brandano,

Thank you guys for all the wonderful info!
I personally prefer single propeller though
First, to answer Kelly question, i was using a single prop (Turnigy 790Kv motor from Hobby King) on UFRO version13-"A". (this is actually version #15, i just called it 13-A because i was just re-using version 13 again)

The UFRO version 13 (and other earlier version), had the AEO-RC CR28M Contra-Rotating Motor w/ 2-Blades @1050kv and also with APM.
I like the contra-rotating prop but the con is the weight penalty due to the additional ESC and battery Y-harness.

I have even tried Himax contra -rotating motor (tried all models except the two last big one - CR 5025 and CR6320)
, but then quit after my right hand (between index finger and thumb) got sliced open by CR3516 during propeller testing last year (phobia ? lol)

I will definitely revisit APM 2.0 again. I do want to use Magnetometer and other features offered by APM ( that why i using it on the first place )
I was just using the firmware in the "Mission Planner" software - "Arduplane". I thought i had a good understanding on how to change some of the parameters in the existing software and expect "magic" happen on my sphere .

There are some parameters in the software loop that create some strange outputs. I wasn't aware these until i switched to gyros (again, please forgive my limited RC experince).
For example, I turned off all PID on all axis (i never use I or D before), and other parameters like T to P mix, rudder mix and etc, but two of the bottom vanes was still responding to YAW motion.
I was in desperation to see my sphere hover and decided to keep it simple by switching to gyros.

As for all my other models (flying and non-flying one), i will post them after i'm done working on version 16. The attached picture here is the latest version 16 that i am working on for the past two days using the same Turnigy 790Kv motor. The bottom half of the sphere frame where the edge is reinforced with carbon fiber stripe. The fuselage is made of Pringles can. You can see version 13-A next to it.

I read about the Flying Carbon Fiber Can and Kelly's Austrian Turboplan, i really love it!
I am so envy that both of them can fly so well~
Again, thank you both for your advise Will keep you guys posted.

Last edited by Ben Loh; Oct 16, 2012 at 12:50 AM.
Oct 16, 2012, 12:30 AM
Registered User
UFRO version 16 in progress.
Last edited by Ben Loh; Oct 16, 2012 at 12:36 AM.
Oct 16, 2012, 12:35 AM
Registered User
Version 16 (Work In progress)
Oct 16, 2012, 12:44 AM
Registered User
Version 13-A (close up pics)
Oct 16, 2012, 01:01 AM
Registered User
Version 8
20" diameter
Weighs about 700 grams
CG at the center of the sphere.
Contra -rotating prop : 11x 4.7 & 11x4.7 R
Oct 16, 2012, 01:03 AM
Registered User
Version X (10)
24" diameter
Carbon fiber stripe on the edge of the frame.
Last edited by Ben Loh; Oct 16, 2012 at 01:10 AM.
Oct 16, 2012, 01:20 PM
Registered User

I'm impressed! Excellent work......!

Have you been able to get spoilers to deploy and work to help dampen oscillations?

Important Concept

Dr. Steve Morris from MBL has posted many valuable photos and comments on RCG over the years. He is the father of the Honeywell Kestrel shown yesterday in the PDF I posted, and also the creator of the V-Bat series of VTOL ships.

Dr. Morris makes a comment on one of his RCG posting about VTOL control systems. I don't recall the exact phrase of his comments but I do recall the intent of his comment.

Basically if you place a control vane too close the the CG you can expect a reversal of this control vane out-put. I.e. if the CG on your spheres is raised up too high towards the Center of Pressure (Cp) of your upper (canard) vane you can expect that control vane to create a force opposite of your intent.

Keep in mind the CG placement for best hover is NOT the same for CG placement for high speed cruise (low alpha) flight. Thus the need for spoilers or an adjustable CG to accommodate these two extremes.

Also recall the original design criteria of the Japanese drone is "search and rescue", thus it it must be able to extricate itself if it crashes, this is why the canard is used, plus reverse thrust from the prop, this allowed the crashed sphere to back-out/rotate from a hole via remote control. How many modelers do you know require this function?

The canard vanes are not required for general hover and modest horizontal flight attitudes and they create a great deal of problems listed below.

1. They are too close to the CG. (I.e control reversal)
2. They are off-set rotationally from the lower control vanes, creating an odd relationship of thrust forces that complicate stabilization.
3. They add a great deal of weight since you also need the servos to control the canard vanes.
4. They are hard to program and adjust due to the relationship of distance from the prop compared to the lower vanes, i.e. (control surface volume verse air velocity = asymmetrical control force) is different on the canard vanes as compared to the lower vanes.

The preferred arrangement is the "Flying (carbon) Can UAV" which is a form of a ducted fan that both increases thrust (as much as 20 %) and control vane effectiveness, it is lighter, i.e. few parts to construct and break and is extremely stable in both hover and high speed horizontal flight, or you can replace the can with a cage as Paco did, just add a skirt; see photos below and look here for credits and details

Make sure you place the CG in-line with the propeller!

Look here>>>> https://www.rcgroups.com/forums/show...1640196&page=6

Last edited by corocopter; Oct 16, 2012 at 02:47 PM. Reason: Spelling error
Oct 17, 2012, 09:55 AM
Registered User

Simple VTOL design:

Here is another excellent example.

Please note:

Mid-Engine (motor) placement.
Pusher propeller.
4 vanes for control.
Great hover and horizontal (low alpha) flight characteristics.
Nearly a sphere in shape, just add the rest of the cage if desired.
Simple design, many fewer parts to construct.
Lighter in weight.
Easy to program and balance.
Easy to repair.....!

Older KK board = piezo gyros

Newer KK2 board would be better.


Kk Single copter (1 min 27 sec)
Last edited by corocopter; Oct 17, 2012 at 01:37 PM.
Nov 02, 2012, 03:15 AM
Registered User

update from singapore

hi guys. just to share, we have gotten rid of the APM control board and all the v tail mixes and have moved on to use the KK2 board V1.2 from Hobby king.

advantages are:
no need for any additional mixes (save cost and weight)
simple on screen LCD programming (no need to lug laptop around)
8 outputs which are completely customizable to allow for custom mixing
has an inbuilt single copter mode setting
has linkable pitch and roll PI setting function (since pitch and roll are symmetrical)

Initial results (after tonnes of PI permutations) show rather good stability. we can smack it side to side, spin it by hand around the z axis and fly it into a wall and it will stick to it until an opposite command is input.

Version 3 has an AUW 550 grams, 42 cm diameter and 1047 SF prop. CG approximately center of the sphere. it flew pretty well without the spoilers in hovering flight but was erratic in higher speed forward flight. with the spoilers on, it feels to fly better when in higher speed forward flight and still has good hovering capability.

BUT it did not perform so well in mild wind outdoors due to its large frontal area. it can stabilize and hover but kept drifting with the wind. it was pretty intense trying to fight the wind : /

videos to come

thanks to Sam and Kelly for their help!
Nov 02, 2012, 01:51 PM
Registered User
Hello Tim,

Congratulations on your test flights and thank you for the valuable feedback.

Can you please share your KK2 PI settings with us?

Looking forward to your vid postings

Last edited by corocopter; Nov 02, 2012 at 04:24 PM.

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Mini-HowTo RC Controlled AR.Drone w/o WiFi! nosaari Multirotor Drone Talk 117 Dec 07, 2014 05:04 AM
Discussion New AR Drone (i-drone) light mod bungle68 Multirotor Drone Talk 3 Apr 13, 2011 06:30 PM
Discussion AR Parrot Drone For Sale Matt Gunn Atlanta Area RC 4 Mar 16, 2011 08:41 AM
Found Parrot AR. DRONE iceman77 Aircraft - Electric - Helis (FS/W) 3 Mar 15, 2011 02:51 PM
Discussion Piloting Drone From iDevice BUT takeoff/landing from a computer chope1 Multirotor Drone Talk 0 Mar 07, 2011 11:12 AM