HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Aug 25, 2009, 02:55 PM
Registered User
CrashingDutchman's Avatar
The Netherlands, Utrecht, Amersfoort
Joined Oct 2004
2,193 Posts
Discussion
Beta Detail - dmgoedde part deux

Someone closed the first thread here: http://www.rcgroups.com/forums/showthread.php?t=734456

Don't know the reason, but let's continue here.
CrashingDutchman is offline Find More Posts by CrashingDutchman
Reply With Quote
Sign up now
to remove ads between posts
Old Aug 25, 2009, 03:02 PM
Registered User
brakar's Avatar
Asker, Oslo, Norway
Joined Feb 2009
422 Posts
AttoPilot v1.8 is in production by now, but Atto IMU is in beta. Do you want to dedicate this thread to one of them, or both?

Btw, the old thread seemes to be back, ...
brakar is offline Find More Posts by brakar
Last edited by brakar; Aug 25, 2009 at 03:11 PM.
Reply With Quote
Old Aug 26, 2009, 01:46 AM
Registered User
CrashingDutchman's Avatar
The Netherlands, Utrecht, Amersfoort
Joined Oct 2004
2,193 Posts
The old thread was (still) closed when I checked this morning.

It doesn't have to be my thread in which we continue or I would have to change the thread title. Maybe it is better to give both the new IMU version AttoPilot and the current AttoPilot 1.8 their own thread?
CrashingDutchman is offline Find More Posts by CrashingDutchman
Reply With Quote
Old Aug 26, 2009, 02:16 AM
Registered User
Joined Aug 2007
1,512 Posts
Imu

Some statements:

Unit-to-unit repeatable performance of uncorrected gyro drift on the order of 0.02 deg/second (and better), attained through rigorous R&D (lots of $$ and time).

Hermetically sealed IMU under clean and dry conditions. Full mil-spec temp range cal'd and capable. Any claim by others of this performance range (especially on the low temp end) is dubious unless they can assure condensation cannot reach the analog traces, which unless hermetic sealing is used, is probably a load of BS.

Each IMU individually temperature calibrated on no less than 9 parameters, plus 6 angular corrections, and 3 scaling factors.

IMU methods are fresh, and developed in house from scratch (i.e. first principles), not on any open-source methods or papers. Focus on complete handling of all important physics for fixed wing aircraft, with additional focus on efficiency and speed of code at high math reoslution (32 bit processing, yes integer math for speed, but precision in the IMU model vectors to 6 sig figures). Attitude in angles used to the 2nd decimal, along with automatic trimming and all the other things that made Atto v1.8 so accurate in both attitude control and navigation/altitude control.

The IMU has no attitude limits, including inverted flight (tested) and high vertical climb (tested). GPS reception quality is the only issue at high attitudes, and that has been handled as well.

High wind is no issue at all, as the GPS ground track heading is NOT used in any IMU corrections. Reliance on airspeed is minimal, but adds an extra level of precision, especially while in continual loiter. I am explaining this partially muddled on purpose, however in another thread I went into fine detail.

NOTE: My goal is to deliver very high flight precision with this initial 6 DOF IMU without resorting to INS solutions, especially those that employ magnetometers and ability of dead reckoning, in hopes that I can get a 9A012 ruling on this for export. Whatever ruling is obtained does NOT affect USA and Canada customers, as these sales are not exports.

Airmcn_3 has seen many hours of Atto IMU flight under torture test conditions, and can testify to the robustness and accuracy. In fact, we tried to confuse the IMU in very small 3D foamy plane (20 ounce flying weight, 3:1 thrust/weight ratio), and could not do so over many weeks of trying. This was July heat in Arizona, with air temps of 115 F (46 C) and ground temp pushing 140 (60 C).

Personally, I have always been well pleased with how the adaptive attitude and navigation algoriths work in the real world with the thermopile v1.8. Customers of the v1.8 can attest to it as well. The IMU version is not simply using an IMU to replace pitch and roll attitude measurement of thermopiles, but rather the complete 3D nature of the attitude equations is used to greatly assist navigation in ways that thermopiles cannot. Focus has been on extending the precision of attitude and navigation, which will be demonstrated here with data and videos. I see plenty of short videos from 'another' autopilot company, but never any plots or data.
dmgoedde is offline Find More Posts by dmgoedde
Reply With Quote
Old Aug 26, 2009, 02:23 PM
Registered User
Near Austin
Joined Dec 2001
5,406 Posts
Believe it!
CenTexFlyer is offline Find More Posts by CenTexFlyer
Reply With Quote
Old Aug 27, 2009, 03:01 AM
Registered User
brakar's Avatar
Asker, Oslo, Norway
Joined Feb 2009
422 Posts
It is not believing that is the problem - it is the awailability part, in particurlar outside of the US...
brakar is offline Find More Posts by brakar
Reply With Quote
Old Aug 27, 2009, 03:52 AM
Nemo in a blender
Connexxion's Avatar
The Netherlands-Amsterdam
Joined Mar 2007
940 Posts
Second that Brakar!

I'm still waiting for a report from someone who finally managed to get one outside the US.

I believe Gary Mortimer was in the process of getting one exported.
Connexxion is offline Find More Posts by Connexxion
Reply With Quote
Old Aug 27, 2009, 11:14 AM
Registered User
Moore, Oklahoma
Joined Oct 2004
4,322 Posts
Quote:
Originally Posted by Connexxion
Second that Brakar!

I'm still waiting for a report from someone who finally managed to get one outside the US.

I believe Gary Mortimer was in the process of getting one exported.

Nobody outside the USA has a full working unit. There is one unit outside the USA that is a RTL not V1.8.....

I do agree with you all, the export crap is just that. The facts are it is not due to lack on Dean's end. Every other day they come back with something new they want to know....


Chris
airmcn_3 is offline Find More Posts by airmcn_3
Reply With Quote
Old Aug 27, 2009, 01:51 PM
Gaftopher
Gary Mortimer's Avatar
Nottingham Road South Africa/Bedford UK
Joined Feb 2007
3,938 Posts
Yeah I have an RTL coming my way, it will serve my AP needs very well stabilized flight and a circle hold around subjects.

Dean has flown an earlier airframe of mine in the USA with his gear and I will integrate the RTL into a newer one here.

The setup file can then be used in the 1.8 when finally it all gets sorted that side.

Boy its been a wait, but the trouble is Deans method is so far ahead of the others that they have moved the bar up, to a certain extent.

I know Dean did'nt close the other thread, so I wonder why it was closed?
Gary Mortimer is offline Find More Posts by Gary Mortimer
Last edited by Gary Mortimer; Aug 28, 2009 at 03:46 AM.
Reply With Quote
Old Aug 27, 2009, 05:30 PM
David1
bmw330i's Avatar
USA
Joined Mar 2007
1,478 Posts
Quote:
Originally Posted by dmgoedde
Some statements:
I see plenty of short videos from 'another' autopilot company, but never any plots or data.
Are you one of them? I don't see any plots or data in your comments. I just ask because like you point out real data means more than words.
bmw330i is offline Find More Posts by bmw330i
Reply With Quote
Old Aug 27, 2009, 06:22 PM
Registered User
Moore, Oklahoma
Joined Oct 2004
4,322 Posts
Quote:
Originally Posted by bmw330i
Are you one of them? I don't see any plots or data in your comments. I just ask because like you point out real data means more than words.

Oh come on, you know darn well there are more plots in the first thread then one would need.

What do you want to see? IMU plots?

Chris
airmcn_3 is offline Find More Posts by airmcn_3
Reply With Quote
Old Aug 27, 2009, 08:26 PM
Chris Anderson
San Francisco/Bay Area
Joined Mar 2007
939 Posts
Quote:
Originally Posted by bmw330i
Are you one of them? I don't see any plots or data in your comments. I just ask because like you point out real data means more than words.
That's outrageous! NOBODY had posted more data here than Dean. Why does every thread here turn into a pissing contest?

It reminds me of that joke that competition is so fierce in academia because the stakes are so low ;-)

I blame the regulatory environment for making starting a company in this space so risky that they end up turning on each other.
zlite is offline Find More Posts by zlite
Reply With Quote
Old Aug 27, 2009, 09:56 PM
Registered User
Near Austin
Joined Dec 2001
5,406 Posts
Quote:
Originally Posted by zlite
I blame the regulatory environment for making starting a company in this space so risky that they end up turning on each other.
The "regulatory environment" has really stratified things for those that "can" and those that "can't" by their will. Dean is one - I consider myself one.... blocked at every turn producing a good product that could have tons of uses both domestically and internationally. Just trying to make an honest living.
CenTexFlyer is offline Find More Posts by CenTexFlyer
Reply With Quote
Old Aug 28, 2009, 02:40 AM
FREEdom!!
Joined Feb 2006
294 Posts
Quote:
Originally Posted by airmcn_3
Oh come on, you know darn well there are more plots in the first thread then one would need.

What do you want to see? IMU plots?

Chris
I think that since this thread was supposed to be about atto IMU (as I understand it), then yes, bmw330i was asking for IMU data. I have seen videos, but I don't remember any performance plots for the IMU version. Have any been posted?
Buzz_Lightyear is offline Find More Posts by Buzz_Lightyear
Reply With Quote
Old Aug 28, 2009, 03:27 AM
Registered User
Joined Aug 2007
1,512 Posts
Quote:
Originally Posted by bmw330i
Are you one of them? I don't see any plots or data in your comments. I just ask because like you point out real data means more than words.
I am NOT picking on Paparazzi, so you can calm down. I am a little ticked off that M Hanson of U-know-who company wrote up outright lies that I copy stuff and don't invent new, when he has in fact inverted the situation (maybe he wishes I was his boss instead of Dave). I see short videos of their flagship system in a giant easy to fly 8' telemaster with nasty airspeed and pitch oscillations, no data plots ever posted, etc. It is either really hilarious or sad, depending if you happen to feel for the guys over there.

If you look back at my other HUGE thread (the biggest on the UAV section of RCGroups) there are dozens of data anlyses and plots, flight tracks, etc. You must have never looked at my thread. Examples of how the vector navigation works, some links to videos on vimeo.com of my IMU being torture tested for high accuracy repeatability on the bench (not just some graphic showing orientation, rather it is twisted around quickly in every possible complex 3D motion, then plunked back down on the table and being well within 1 degree of perfect convergance). I even shared stuff on how quantitative testing on thermopiles was done about 2 years ago, real time bias correction of pitch and roll, some AWESOME plots on barometric altitude work that clearly show the landing halfway up my stairs as I ascended and descended them. Just the typical perfectionist stuff I like to show.
dmgoedde is offline Find More Posts by dmgoedde
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Jet Day at the Bay, Part Deux Dr. Jet Electric Ducted Fan Jet Talk 21 Oct 07, 2002 07:50 PM
Jet Day at the Bay, Part Deux gregg f Electric Flight Events 1 Oct 01, 2002 11:15 AM
equalizing cells in a pack - part deux freq Power Systems 2 May 02, 2002 03:27 PM
frustrated with concept -part deux bpu6999 Electric Heli Talk 6 Aug 07, 2001 12:31 PM
S & E M Design Comp - Part Deux Dereck Sport Planes 18 May 24, 2001 06:57 PM