New User - Great Job! (and advice for those starting out like me, and how to cr

Tookings · 2007-10-01 06:10

Hello -- mainly wanted to say great job the the Parallax team!

I used the BS2 many years ago from some theatrical work, and they were a pleasure to work with... Now I'm just getting around to the Propeller, and I'm entralled!

Also wanted to mention, while at first I was a little lost, and not looking forward to the 450 page manual while hunting around on the forums for info -- for those who want a "quick start" but to still understand (most) everything, check out the Propeller Education Kit (http://www.parallax.com/detail.asp?product_id=32305). You don't have to actually have the Kit -- just a few switches and LEDs. The important part is to go through the Labs available for download. Depending on your previous knowledge you can just skim, or actually try everything if you want, but once you're done reading or doing, you should be able to avoid almost every newbie question I found or was going to ask. (More Kudos to the Parallax team!)

Now off to get this accelerometer to stop trying to crash my helicopter. (Anyone here done any work with autonomous RC heli/plane/car control?)

And once I get the heli to at least hover on it's own...I'll send up a video. It'll probably be a while before it goes anywhere though...

Thanks!

-Rick

Fred Hawkins · 2007-10-01 07:36

Rick,
the Manual becomes manageable when you use Adobe's search function. There is also a version (recent post -- within 2 weeks) with bookmarks.

F

LoopyByteloose · 2007-10-01 08:05

To start, try to think of the Propeller as eight BasicStamps with shared I/O pins on a common bus.

Of course, the Propeller does it all with 32bits rather than 8bits, but that opens new horizons as it greatly saves time in many situation and allows more precise math. Of course, it still communicates with 8bit world in a friendly fashion.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
"Everything in the world is purchased by labour; and our passions are the only causes of labor." -- David·Hume (1711-76)········

···················· Tropically,····· G. Herzog [noparse][[/noparse]·黃鶴 ]·in Taiwan

simonl · 2007-10-01 08:53

Hey, welcome!

I think you'll find there are several of us looking at helicopter stabilisation / autonimous flight. Some have made a start, but many (like me) are still dreaming :-( [noparse][[/noparse]Note to self: pull yer finger out. LOL]

Have fun

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cheers,

Simon

www.norfolkhelicopterclub.co.uk
You'll always have as many take-offs as landings, the trick is to be sure you can take-off again ;-)
BTW: I type as I'm thinking, so please don't take any offense at my writing style

Gadgetman · 2007-10-01 11:30

I've been looking, too...

I've gotten as far as getting an X-UFO to modify, a set of Propeller protoboards, a SparkFun '5 degrees of Freedom' XY-Gyro/XYZ-Accelerator and a few IR sensors.
As soon as I've moved to new apartment, I'll start playing with ADC routines and generally mess about.
(Haven't had a chance to code for the last couple of months)

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Don't visit my new website...

QuattroRS4 · 2007-10-01 12:23

TooKings,
1) Welcome aboard
2) Autonomous UAV's seem to be quite topical

Here are a few links -
http://forums.parallax.com/showthread.php?p=639638
http://forums.parallax.com/attachment.php?attachmentid=46401
www.parallax.com/sx/contest/autopilot/Jens-Autopilot2.pdf
http://forums.parallax.com/showthread.php?p=530691
http://forums.parallax.com/showthread.php?p=645854

If you plan to search the forums - the inbuilt search is not very good - use this link instead
search.parallax.com

Regards,
John Twomey

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
'Necessity is the mother of invention'

Tookings · 2007-10-01 20:07

Thanks QuattroRS4!

Those are some excellent links (and some great links from them...I don't know how I never ran into sparkfun.com). Will always use the other search now, as I never found any of those using the forum search.

I'm working with a coaxial heli -- as a human can learn to fly it without crashing constantly in a few days, versus months for a normal heli....so I'm hoping that will make my job easier. I've got the dreams of wearing the stereo goggles while clicking around on a map in the house -- but for now, I'm just getting the left/right/fwd/back to be auto, while I still RC control power and yaw. (And this accomplishes failsafe too ...)

I'm actually using a an E-flite Blade CX that probably doesn't have a single original part left...

Thanks again all!

-Rick

Post Edited (Tookings) : 10/1/2007 9:57:51 PM GMT

simonl · 2007-10-05 11:26

Hi Tookings,

That heli' will definitely make your life easier - any coaxial heli's a lot more stable that a traditional one.

Keep us updated on progress

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Cheers,

Simon

www.norfolkhelicopterclub.co.uk
You'll always have as many take-offs as landings, the trick is to be sure you can take-off again ;-)
BTW: I type as I'm thinking, so please don't take any offense at my writing style

Tookings · 2007-10-05 14:38

If only it were as easy as taking the integrated distance from the accelerometers, add-subtract from current position, and then have servos seek a 0 distance offset for a hover. (ie. If accelerometers magically had no noise, absolute distance accuracy, and didn't notice tilt...which of course, is not the case.)

Not complaining -- I've never worked with accelerometers before a few weeks ago, and it's been fun learning. (And learning a lot of math I had forgotten.) Early on, I did build a somewhat nifty 2D autoleveling platform when I was figuring out how accelerometers worked in the first place -- essentially seeking 0 as above, but since the tilt indicates so much "movement" it's very effective in leveling. Working on including rate gyro sensors and GPS eventually to come up with something workable for the heli...

Question, before I assume no, just in case I'm really missing something -- is it possible to maintain a "stable" lateral X-Y position with just accelerometers? (Ignore a helicopter for now -- think of just sliding a breadboard around on a table.) Everything I know or have tinkered with shows the accumulated error and extreme sensitivity to tilt makes this pretty impossible.

Something like this simple pseudocode * axis: (updating from the accelerometer at about 100Hz)

repeat
  g = (accel.x-accel.vref) / 455
  trueg = g + abs(trimg)  'adjusts g with a stored resting state, recorded after about 5 seconds of runtime
  mTraveled = trueg * sqr(seconds_passed) * (9.80665)
  mError = mError + mTraveled
  function_to_effect_correctional_movement(mError)

With something like the H48C and servos, I can get a very nice self leveling breadboard on a table that way, but that's about it...just wanted to make sure there is no way to get reliable (or even just remotely reasonable) absolute position data with just the 3-axis on it's own. I'm working with the math using rate gyros to compensate for the tilt, which seems like what is needed if you want/need to avoid any external position reference. (I know external will be needed later and it's still not that "accurate" without, but I'm experimenting with what you can squeeze out of just self-reference first...)

Thanks for any insight...or links/references. (Most of what I have found makes heavy use of external references and isn't going for absolute position.)

-Rick

Post Edited (Tookings) : 10/6/2007 12:01:51 AM GMT

Tookings · 2007-10-05 21:27

Found another interesting autonomous community site to add to the list:

www.recherche.enac.fr/paparazzi/wiki/index.php/Main_Page

New User - Great Job! (and advice for those starting out like me, and how to cr

Comments