Thread Tools
Mar 02, 2010, 06:48 PM
Registered User
Quote:
Originally Posted by Buzz_Lightyear
My apologies. I didn't know that this was not the final product.
No reason to apologize. I am sorry for sounding irritated.

Chris
Sign up now
to remove ads between posts
Mar 07, 2010, 03:35 PM
Registered User
Quote:
Originally Posted by d_wheel
Wow! Ask what time it is and people try to tell you how to build a clock .

Yes, I know the bank angle has to change, speed makes a difference, any wind will make things change, yatta, yatta, yatta. That will all be taken care of by the live downlink and ability to change the camera angle in real time. I now have a place to start and will fine tune it as I go.

Thanks again to everyone.

Later;

D.W.
HI
we have been doing lost of hold patterns. At this time it is not possible to hold any angle of bank in a hold at the set home point. Never tried one at a way point but guess it is the same. our R are about 190 M speed is about 35 MPH bank angle averages about 5 degrees.
You are missing one thing and that is adding in up elevator.
If you start adding up elevator your circle going to get tighter and the Gs will start to build. So it is possible to fly a 50 foot circle at 30 degree bank. If you were flying the x15 and had it in a bank of 30 degrees your circle would be around 20 miles. Speed will determine the G force on your model. Could reach 50 G or more. Only problem is Atto cannot do that yet. I do not think it needs to do that.
My suggestion is to set an R to 190 at the front door of the house and set way points around the house. Each one another 50 feet higher than the last.
Sim
Last edited by sim559; Mar 07, 2010 at 04:45 PM.
Apr 05, 2010, 04:38 PM
Registered User
ooops
A little miss information on holding bank angles in a hold circle.
will explain later
Sim
Apr 07, 2010, 12:24 AM
Registered User
dmgoedde's Avatar
Thread OP
With assumptions of level flight (pitch and altitude being held) in a banking turn, and assumption that alpha is not increasing drastically... other words the aircraft is riding the air without slippage and instead is cutting through the turn as if on rails, the G force (in Z direction of aircraft body) is independent of flight speed. It only depends on the roll angle. What is baked into these assumptions is that the against-gravity lift vector is constant and equal to the aircraft weight (this comes from assumption that altitude is being held without loss or gain) but because of the non-zero roll angle, the body Z-axis G force is higher than 1G (in order, again, to have the Earth-centered force = 1G despite the wings not level). According to these assumptions, 1 extra G would be generated at a 60 deg roll angle for a felt 2 Gs. As the roll angle approaches 90, the Z-axis G force quickly goes to 5...10...20 Gs. Pylon racers don't pull 20+ Gs because of their flight speed, they generate this much accel force in the Z axis due to knife edge turns at 80+ roll angle.

As sim points out, speed (like mach 6+ in the X-15) only causes the turn radius to become huge for a given roll angle, not generate lots of G force per se.

Back in the stupid federal 55 mph speed limit days they said "speed kills". Idiots... acceleration (m/s^2) kills.. connective tissue doesn't like 10+ Gs or jerk either (m/s^3).


This is why I added G-force corrections to the gyros in my IMU method. Even +/- 1g causes a gyro bias shift of up to +/- 0.2 deg/second. Relying on a feedback filter to track this bias is shoddy at best. Feed-forward is essentially instantaneous. AttoPilot IMU is proven to stay converged at continual 60 deg bank angle, where the Z-axis g-force is approx 2G's. If you are not accustomed to flying around in RC mode at 60+ deg bank, or your previous autopilots didn't tolerate 60 deg roll angles, I can tell you that it looks uncomfortable to watch in person. I suppose I feel the need to post another video on this...
Apr 07, 2010, 01:43 AM
Chris Anderson
Ah, just like the good old days. It's great to have you back, Dean!

(now get back to work! )
Jul 14, 2010, 02:01 PM
Gaftopher
Gary Mortimer's Avatar
Where did this thread go?? So many people have used it for research its a shame not to ping it back again!!! The version 2.0 AttoPilot is out now. Full of collaborative UAS flight goodness and correct flying techniques. The RTL thermopile version is only $500 more than you might ever need for photo mapping.
Jul 17, 2010, 06:37 PM
Registered User
dmgoedde's Avatar
Thread OP
Quote:
Originally Posted by Gary Mortimer
Where did this thread go?? So many people have used it for research its a shame not to ping it back again!!! The version 2.0 AttoPilot is out now. Full of collaborative UAS flight goodness and correct flying techniques. The RTL thermopile version is only $500 more than you might ever need for photo mapping.
Gary is referring to some masters work completed based on AttoPilot, with a few in process right now.

By "correct flying techniques" Gary is referring to:
1) The newly added pitch-airspeed/throttle-altitude method, as well as 0-100% mixability of this method with the older method (pitch-altitude/throttle-airspeed). Also, user can set altitude bands over which to mix from one % of mix to another %.

2) Navigation auto-corrects for wind magnitude and direction.... drum roll... without a potentially laggy/noisy/incorrect wind model required. I am not anti wind model, and can see where it adds value for things like knowing when to abort a mission early if the UAV is X distance from home with Y mAh battery remaining and Z head wind vector (predictive method to estimate mAh to get home, and ability to effectively abort early or later given the conditions), HOWEVER I developed a method that yields high-end results for navigation despite wind under any and all circumstances. It works so long as the UAV has enough thrust to maintain positive ground speed against the wind. Also, AttoPilot now is 100% self-tuning on navigation without need for user intervention. Navigation uses an LQR-inspired method.

Navigation has been torture tested on a variety of airframe types. Some airframes have over 1000 autonomous miles in about 200 flights with the IMU system.

I re-vamped the control board to be about 1/2 size of the V1.8 but retains full size header pins, and to support 3 products: IMU, thermopile, and PPK (a $250 complete turn rate wing leveller autopilot for motor gliders using MEMS yaw gyro on board the controller). "K" stands for killer, and I'll let people guess what the "PP" is.

New pressure sensors. Upgrade to 16 bit ADC onboard. Servos ground and power isolated from Rx and control board via ferrite chips. Pan and Tilt compensation enabled. Got around need for a case altogether: back of the control board is potted in electronics grade epoxy to protect against routine handling (10 grams with potting, about 7g without). Additional protection circuitry for the processor. Trimmed telemetry strings to use less modem bandwidth, SWARMING supported and vehicle ID, improved loiter methods, Simplified LOG file, LOG file name built using year month and date so you don't get 20MB single log files from a week of flying, added override commands via GCS for pitch/roll/airspeed (one Masters student is using this for his custom flight control laws to achieve a pretty cool project), greatly simplified SET file as well. Several SET parameters are now auto-calc'd and thus reduces task of the end user in tuning.

*NEW*
The V3.0 IMU system comes with a tiny 0.55 ounce keychain video camera that provides high quality video + sound to onboard micro-SD. The camera is available for purchase as well, as in the case when the V2.0 thermopile system is purchased for the new reduced price of $700. Purpose of this is for stabilization tuning maximization, but that is only half the story. Additionally, we are in process of generating a series of short videos resulting from a DOE style matrix of flights where a well tuned airframe is purposefully de-tuned in the 9 tuning parameters at the +/- 25% and +/- 50% levels; we SHOW YOU symptoms from FPV video for each of the 9 parameters being tuned to weak and too aggressive. Then you compare your video to ours and achieve rapid success in tuning.

*Documentation*
Below is a paragraph size mea culpa for a few things. However one of my greatest regrets about 2009 and 2008 is the AttoPilot documentation. Some people had rapid success with their AttoPilot V1.8, others had difficulties due to subtle issues that existed. Hardware marginalities have been addressed in the new design, but what would have really helped last 2 years is an exhaustive troubleshooting matrix. A list of symptoms for problems along with likeliehood-ranked possible root causes for each. The upcoming documentation contains this, and is rich with flowcharts, diagrams, pictures, etc. The documentation is now much more fitting for a system as sophisticated as the AttoPilot. I ended up spending far more time helping people individually rather than sinking that time into assembling the known issues into a document.

Explanation for myself in 2009:
I haven't slacked off or stopped working, or disappeared. 2009 was a huge learning experience for me, and towards the end of 2009 it took every material and mental resource I had to not give up, but I made it through. I'm sorry for a particular rant I made last year on RCGroups. I tend to be very idealistic and sensitive, and sometime can't mentally reconcile injustice, so I blew up. The US feds are not making life easy for people in my line of work. In 2009 I thought I was going insane and lose everything despite a few hundred serious inquiries for sales overseas... these sales would have made all this risk worthwhile. A very very dark time for me indeed. Then despite making lawyers $25k richer, export licenses became a giant headache and long waiting. Now I have backers and other people working this for me, so worries has subsided and progress resumed about 7 months ago. Am I still bitter about some things? Have I aged 10 years in experience? Yes, and yes.
Last edited by dmgoedde; Jul 17, 2010 at 10:33 PM.
Jul 18, 2010, 01:03 AM
prefer lift over drag anyday
flyzguy's Avatar
Very exciting stuff and very nice to see my e-mail inbox with a thread update for this one! Professional documentation will be a great asset no doubt.
Jul 18, 2010, 01:09 AM
Registered User
mboland's Avatar

Woohoo he's still with us


Thanks for the post Dean. Nice of you to keep us up to speed on what's been happening

Good to hear the red tape has not strangled the life out of you.
I was beggining g to worry it had.

We are busy gearing up for the next Outback UAV Challenge, in which I am sure there will be more than a few Attopilots in operation.

Keep up the great work and don't let the bureaucrats get to you.


Mike
Jul 18, 2010, 03:29 AM
Registered Crasher
Bilox's Avatar
Hey Dean,

Great to hear from you buddy!

PPK ........... That's gold..... Pure gold!

Cheers

Scott
Jul 18, 2010, 04:41 AM
Registered User
dmgoedde's Avatar
Thread OP
Quote:
Originally Posted by biloxthecat
Hey Dean,

Great to hear from you buddy!

PPK ........... That's gold..... Pure gold!

Cheers

Scott
Thanks for the encouragement guys (I assume you are dudes!)
Jul 18, 2010, 10:06 AM
Old Fart
d_wheel's Avatar
Quote:
Originally Posted by dmgoedde
Also, AttoPilot now is 100% self-tuning on navigation without need for user intervention. Navigation uses an LQR-inspired method.
Great to see you posting again Dean. Don't let the red tape get you down. I worked in the medical equipment field for many years and must say that the government is the one that has caused most of the high priced health care by requiring so much BS. I totally understand how you feel.

I have been flying ATTO V1.8 since it first came out and just got it updated to 1.9. I have enjoyed its outstanding performance very much. It is so much beyond previous AP's (love the PPK reference!) in its price range that I do not have words to describe it! My friends are always blown over when they see what it will do.

Did you include self-tuning in the 1.9 version?

Later;

D.W.
Jul 18, 2010, 12:40 PM
Registered User
Quote:
Originally Posted by d_wheel
Great to see you posting again Dean. Don't let the red tape get you down. I worked in the medical equipment field for many years and must say that the government is the one that has caused most of the high priced health care by requiring so much BS. I totally understand how you feel.

I have been flying ATTO V1.8 since it first came out and just got it updated to 1.9. I have enjoyed its outstanding performance very much. It is so much beyond previous AP's (love the PPK reference!) in its price range that I do not have words to describe it! My friends are always blown over when they see what it will do.

Did you include self-tuning in the 1.9 version?

Later;

D.W.

V1.9 is no longer for sale. It will only be V2.0 and above from here on out. We will continue to support V1.8-V1.9 owners but will not be selling them any longer.

Look at the plus side! You will be getting a more advanced unit with a slightly lower cost.

Chris
Jul 18, 2010, 03:06 PM
Old Fart
d_wheel's Avatar
Quote:
Originally Posted by airmcn_3
V1.9 is no longer for sale. It will only be V2.0 and above from here on out. We will continue to support V1.8-V1.9 owners but will not be selling them any longer.

Look at the plus side! You will be getting a more advanced unit with a slightly lower cost.

Chris
Hello Chris. I understand, but wanted to know if my V1.9 has the self-tuning that Dean was describing.

Later;

D.W.
Jul 18, 2010, 03:56 PM
Registered User
Quote:
Originally Posted by d_wheel
Hello Chris. I understand, but wanted to know if my V1.9 has the self-tuning that Dean was describing.

Later;

D.W.
Not to the extent of V2.0. Heading prop band, SET line $36 is the only AUTO tuning parameter in V1.9.

Chris


Quick Reply
Message:

Thread Tools