rrTV-PHOTO   New HD TV
HOME   rrTV-PHOTO   GALLERIES   MY GALLERY   HELP-FAQ
myHOME PM pmRR MEMBERS 396 ONLINE 43 EVENTS SEARCH REGISTER  START HERE
 
2 pages [ <<    <    ( 1 )     2     NEXT    >> ]2635 viewsPOST REPLY
Ace Hobby . Esprit Model . Thunder Power RC

.
.
UAV Autonomous Unmanned Aerial Vehicles > Autonomous, flybarless, T-Rex 450
 
 
wxyzSenior Heliman - Location: calif*, U know where -
That is a living, breathing flybarless T-Rex 450 in auto hover.

Carbon fiber blades
ADXRS150 gyros
ADXL330 accelerometer
Micromag3 compass
HS-65MG cyclic servos
S3102 tail servo
SIRFIII non differential GPS
XBee ground link
Neural network navigation
No rate gyros
IMU has full control over rates & orientation
VicaCopter firmware

Unfortunately, weather windows R so short between Feb & May, have to use every opportunity to fly & not put on the Vicacopter logo.

02-21-2008 09:04 AM
 
 
ErichF
Key Veteran
Location: Odessa, FL 33556 (Tampa Area)

This is not a flybarless setup, but it's a heck of a lot cleaner. I've personally flown this bird a couple times...it really works well:



See a demo video at:

http://www.adaptiveflight.com/index.html

Team Kyosho Regional Field Representative
02-21-2008 07:41 PM
 
 
wxyz
Senior Heliman
Location: calif*, U know where

VicaCopter is not a commercial product or a day job. It's all after-tax personal income. If U have the money U should buy a commercial solution instead of hacking your own. The flybarless T-Rex 450 on a single IMU is the bleeding edge of what can be done but a long way from being commercial.
02-21-2008 11:16 PM
 
 
wxyz
Senior Heliman
Location: calif*, U know where

Daylight shots of the flybarless beast. She does better at high altitude
than low altitude because of the GPS.


02-22-2008 07:04 PM
 
 
rroback
Key Veteran
Location: Irvine (UCI), Ca

awesome! If you want a bigger test bed, I've got a maxi joker I could test it on

Rhett... There's no power like E Power!
03-18-2008 02:15 AM
 
 
wadiprawita
Heliman
Location: Bandung, Indonesia

Dear wxyz

What processor do you use ?

-doni-
03-25-2008 02:45 PM
HOMEPAGE  
 
 
wxyz
Senior Heliman
Location: calif*, U know where

This is a Gumstix 200Mhz + PIC18F458 40Mhz. Sourcecode & more pictures are on vicacopter.com/
03-26-2008 02:42 AM
 
 
JEEPWORLD2002
Key Veteran
Location: BLUEBELL PA USA

thats very cool

Trex600n Trex500 Mikado LoGo5003d// Hacker, CastleCreations,Ys 50, JR 7703d/8900, Radix
03-26-2008 03:12 AM
 
 
wadiprawita
Heliman
Location: Bandung, Indonesia

Dear wxyz,

Thank's for the reply.

I still have questions :

- How much is the trex's max payload capacity ?
- How long is the flight duration with this payload ?
- Why flybarless ?
- Why use 2 processors ?

Regards

-doni-
03-26-2008 01:46 PM
HOMEPAGE  
 
 
wxyz
Senior Heliman
Location: calif*, U know where

- How much is the trex's max payload capacity ?

The largest payload flown so far is 9oz.
The hand soldered autopilot is another 8oz. The total
amount flown which is not stock is 17oz.

- How long is the flight duration with this payload?

The 17oz payload flew for 7 min using a 3.3Ah battery.

- Why flybarless ?

Removing the flybar reduces vibration, drag, weight, & damage during crashes.

- Why use 2 processors ?

Gumstix has the cheapest single board computers, but they don't have
enough pins to get data in and out. You need a second processor to get
more general purpose I/O.
03-26-2008 07:38 PM
 
 
wadiprawita
Heliman
Location: Bandung, Indonesia

Dear wxyz,

Thank's again for the reply.

Other (lots of) questions :

- How much is the total weight (heli + autopilot + payload) ?
- Any serious penalty on the flight performance with this payload ?
- What are the combinations of motor - pinion - main blade ?
- What battery brand and type (the 3.3AH) ? Have You ever compared to regular 2.2AH (such as 2170mAH flightpower 25C) ?
- How You can solve the interference from motor to the magnetometer ?
- About the flybarless vibration, does it worth the modification or actually You can live with standard flybar version ? Does this also effect duration ?
- The Autopilot is started after certain safe altitude is reached or started from ground ?
- Why don't you use higher update rate GPS such as uBlox LEA 4 ?
- From the photo it seems you do not use COTS gyro for rudder ?
- Any tips how to minimize crash damage during test ?
- Do you use neural net to supervise the low level PID algorithm ?
- You are Jack Crossfire from RCGroups right ? I love your blog.

Regards

-doni-
03-26-2008 11:35 PM
HOMEPAGE  
 
 
wxyz
Senior Heliman
Location: calif*, U know where

Autonomous sky painting

Finally figured out how to stack timelapse photos. This is what can be done with 1Hz GPS, LED's, and successive exposures. Battery life severly limits what can be drawn. The results look better on Google Earth.





- How much is the total weight (heli + autopilot + payload) ?

Have never weighed it.

- Any serious penalty on the flight performance with this payload ?

Have never flown a stock T-Rex. The stock Corona was more stable. The
wide electronics table creates a lot of turbulance.

- What are the combinations of motor - pinion - main blade ?

Align 430XL motor
Plastic 315mm or Carbon Fiber 325mm blades fly the same.
12 tooth pinion

- What battery brand and type (the 3.3AH) ?

Currently using MaxAmps 3.3Ah with over 200 cycles. Those fly for 11
minutes on the T-Rex. On the Corona, they flew for 13 minutes with 8oz
autopilot. On the Corona, the fresh 3.3Ah flew for 20 minutes with 8oz
autopilot.

- Have You ever compared to regular 2.2AH (such as 2170mAH flightpower 25C) ?

The 2.1Ah kit battery flew 9 minutes when it was new. It was down to 5min before it puffed.

- How You can solve the interference from motor to the magnetometer?

The magnetometer calibration eliminates the effect of the NiFeB
magnets. Modern motors don't leak much electromagnetism. They would be
inefficient if they did.

- About the flybarless vibration, does it worth the modification or
actually You can live with standard flybar version ? Does this also
effect duration ?

The autopilot flew once with a flybar. Flight time increases 10%
without it. The gyros cost too much to not use them for cyclic. It's
an opportunity cost issue.

- The Autopilot is started after certain safe altitude is reached or
started from ground ?

There is no auto takeoff or landing. It would mean adding a boundary sensor.

- Why don't you use higher update rate GPS such as uBlox LEA 4 ?

Tested the uBlox NEO-4S & the EB-85 last year. The uBlox had too much
noise at 4Hz. The EB-85 couldn't detect slow velocities. The EM406 is
slow but very accurate. Can't afford a Novatel OEM-V1.

- From the photo it seems you do not use COTS gyro for rudder ?

The IMU provides all stabilization.

- Any tips how to minimize crash damage during test ?

Don't use a flybar. Use training gear to test rate damping. Leave the
tail struts on. Fly over tall grass. Stay on the engine cutoff.

- Do you use neural net to supervise the low level PID algorithm ?

The neural network is only for predicting velocity.
03-27-2008 07:27 PM
 
 
FrankC
Heliman
Location: Ocala, Florida

This is an interesting project wxyz. Would you be willing to estimate your investment so far? I have a T-Rex 450 and would be curious as to how much it would cost to duplicate what you have. I'd like to be able to do some stable autonomous hovering for AP work.
03-28-2008 06:51 PM
 
 
Eco8gator
Key Veteran
Location: Palm Beach Gardens, FL

Very impressive work!

Carlo
03-28-2008 07:08 PM
 
 
wadiprawita
Heliman
Location: Bandung, Indonesia

Dear wxyz,

Thank's again fot Your reply. Would You mind telling me how you calibrate the magnetometer (any reference will be very helpful).

Regards

-doni-
03-29-2008 12:52 PM
HOMEPAGE  
 
 
wxyz
Senior Heliman
Location: calif*, U know where

The autopilot itself using 1Hz code phase GPS is only $600 of parts. A more reliable 20Hz carrier phase GPS would bring you to $900. The cost of crashes, batteries, airframes is $1,000,000,000,000,000,000,000,000.000. It is the calibration & work you do yourself, not the stuff you buy, that is the main problem.

From the VicaCopter manual which is a work in progress:



**** Magnetometer calibration

The IMU must be calibrated and giving the proper roll & pitch before the
magnetometer is calibrated.

In settings.conf, set X_MAG, Y_MAG, Z_MAG to 0, 1, 2. Set DUMP_MAG to
1. Set DECLINATION to 0.

Run the autopilot on the airframe with the magnetometer & everything
mounted in its final position. Rotate the airframe on the X, Y, & Z
axes and note which axis corresponds to which column of mag output.
Change the X_MAG, Y_MAG, & Z_MAG indices to match the dumped columns.
Now the X, Y, & Z axes will be dumped in order & it's time to calibrate
the ranges.

Rotate on the X, Y, & Z axes and note which direction North is from the
printout. Unless you're on the equator, north points into the ground or
into the sky.

Point each axis of the magnetometer towards & away from North and copy
the max & min values from each column to the MAG_MIN and MAG_MAX values
in settings.conf. Next, set DUMP_MAG to 0 and DUMP_THETA to 1.

With the autopilot running, rotate the airframe on the yaw axis. The
printed heading should be North: 0, East: 90, South -179/179, West: -90.
If the heading increases when it should decrease, reverse COMPASS_SIGN
in settings.conf. If the heading is always offset by a certain amount,
change COMPASS_OFFSET in settings.conf to correct the offset.

Finally, you need to set DECLINATION to the magnetic declination for
your area. If magnetic north is 14' East of true north, set DECLINATION
to 14.

With the autopilot running, rotate on the yaw & pitch axes to verify the
compass reading is mostly independant of the attitude.
03-30-2008 08:50 PM
 
 
wadiprawita
Heliman
Location: Bandung, Indonesia

Dear wxyz,

Thank's for the reply. Is there any reference about magnetometer calibration ?

I'm still confused, the declination are 3 for each axes (or maybe in the form of rotation matrix) or only 1 after heading calculation ?

Regards

-doni-
04-02-2008 12:04 AM
HOMEPAGE  
 
 
wxyz
Senior Heliman
Location: calif*, U know where

This is the latest video of autopilot in action. Got decent results from a high resolution camera onboard. The swashplate mixing was upgraded to a neural network. The idea is to decouple the controls & overcome nonlinearities in the swashplate. Results with 1Hz GPS are much less stable than 20Hz GPS.



Ning.com has full resolution photos.

http://diydrones.com/photo/photo/li...e=JackCrossfire



Declination is the difference between magnetic North & true North, so it only affects the heading result.
04-02-2008 12:38 AM
 
 
ErichF
Key Veteran
Location: Odessa, FL 33556 (Tampa Area)

Looks like it holds position very well, but at the expense of very erratic movements. Let up on the PID loops a bit and allow it to float around a little bit more. I think then you will have a much less jittery hover.

Good progress, nonetheless.
04-02-2008 12:54 AM
 
 
Perry Kavros
Senior Heliman
Location: San Jose, CA

Each post you make it gets more and more impressive each time, keep up the good work.

Perry
04-02-2008 04:22 AM
HOMEPAGE  
 
 
2 pages [ <<    <    ( 1 )     2     NEXT    >> ]2635 viewsPOST REPLY
Real Raptors . Mikado Modellhubschrauber . GrandRC

.
.
UAV Autonomous Unmanned Aerial Vehicles > Autonomous, flybarless, T-Rex 450
  UPDATE SCREEN   PRINT TOPIC Advertisers 

Subscribe to This Topic

Sunday, July 20 - 6:50 am - Copyright © 2000 - 2008 runryder.com | email | link to rr | runryder needs cookie