Shop OBEX P1 Docs P2 Docs Learn Events
microcontroller choices — Parallax Forums

microcontroller choices

copacetic353copacetic353 Posts: 52
edited 2008-11-07 03:47 in Propeller 1
so i wasnt quite sure which forum i should have posted my topic in, but this one was at the top of the list on parallax.com [noparse]:)[/noparse]

anyway, i am going to be endeavoring into some robotics projects. my first project will involve basic motion and obstacle avoidance and evolve from there into distance detection, and hopefully gps chips and 3D "attitude" control, like on a helicopter for example.

basically my question is, what chips/MCU's would be best suited for these purposes? even if they are not from parallax i would really appreciate the info here.
and if discussion does lead me into buying a parallax product, even better, they have lots of support and manuals which from what i hear are very helpful.

thank you for your time and knowledge!

Comments

  • TJHJTJHJ Posts: 243
    edited 2008-11-07 02:48
    From my experience with the world of micro Controllers, it comes down to just put a prop on it.

    I can tell you how many projects I have done using other chips, while sitting there thinking this would have taken me 5 minutes on a prop, 2 weeks into the project.

    The propeller is one of the most powerful and versatile chips out there, sure you may say Im bias because this is a propeller forum, but I cant even think of anything that compares to what this chip can do. I mean really 8 processors?? In one Chip.... Im not joking. Yeah its that bad ***.

    As for your application specifics, I am sure you find that as your robotics become more powerful and complex, the code that runs them gets much more into the number crunching territory. Something the propeller excels at, If you are going to try and do a autopilot on a rc helicopter, depending on the type you chose, (Counter rotating/ aerobatic) It will need to be very very fast, in the case of a helicopter there is very little time to sit around processing numbers for your next move, because that correction is already passed being use full. So you need speed and lots of it, lots of muti-tasking capabilities.
    I actually have been playing with a autoPilot for a helicopter, named PropCopter, it is a E-Flite Blade CP-PRO. I use a magnetic compass, ping sensor, GPS, and a 3 axis accelerometer. I use 3 processors to control its flight controls, One for Forward/backwards,, one for left/right, and one for Throttle and Rudder( There not that radical so I put them together.) I then dedicate another processor to gathering data from the "high priority sensors" , accelerometer, and compass(Used to control yaw), and ping. Then I leave another to get the gps data because its slow. Dedicate one to running the float math operations. And there is still 2 processors left.... Im thinking digitized video, but it still needs work on its flying. The counters on the various cogs(eg processors) feed the pulses out to the servos, so that doesnt even use a processor. To top it all off its running at well over 2,000 updates a second.

    Sure you could do this on another micro controller system, but it will take a lot of very complex coding, and I doubt you would be able to do it all on one chip, so you end up with a very complex system. That gets even harder to program, ask anyone who has worked with pic's and what they feel about interrupts, most will tell you that they make them want to take a large hammer to it, no need for interrupts when you have 8 processors.

    So in short, there is nothing out there that I know of that is as flexible, powerful and easy to learn as the propeller. As has become my motto, just put a propeller on it well get to go home tonight.

    Best of luck out there,
    TJ

    EDIT: I forgot the obex, a large collection of pre made objects to run code that simplify your life infinity, For all of the sensors parallax sells, there is a pre made object that let just use the data instead of spending forever trying to get it to interface correctly.

    Parallax has this great habit of taking something really hard and making it really easy. I once tried to build an H-Bridge from scratch. A very simple idea that is quite difficult to actually make work correctly. Sadly a ton of money in parts later, a massive headache, I just bought the one parallax makes, its much better than what I was building, and it lets me put 2 motors on the same pin, works flawlessly and was cheaper than what I was trying to build in the end, per unit cost. So now days if parallax makes a sensor for it, I just use that, because it work from the moment you turn it on and it does what you want it to do. They just know how to do it right. if you ignore everything else I said here, take this home, if parallax makes it use it, you will be a lot better off in the end.

    Post Edited (TJHJ) : 11/7/2008 2:59:54 AM GMT
  • Mike GreenMike Green Posts: 23,101
    edited 2008-11-07 02:48
    Generally for complex robotics stuff, the Propeller would be an excellent choice. It's fast, has the ability to interface to all sorts of sensors. It has a reasonable amount of memory (32K). It can be programmed in several languages like Spin and its assembly language (for free), C (inexpensive), and Forth (inexpensive).

    For learning basic robotics, the Stamps are better. They're easier to use although they are much slower and have much less memory. There's a tremendous amount of educational material, particularly on robotics use, and a lot of sample code and demonstrations of obstacle avoidance, distance detection, etc., particularly based on Parallax's BoeBot (Board of Education Bot). If you do go with a Stamp, pick one of the BS2p series. The BS2px particularly is one of the faster Stamps, has more memory than many of them.

    There's really nothing else out there with the level of support and educational material than the Stamps

    You might look at the Propeller BoeBot described in the Completed Projects forum here. This uses a BoeBot chassis with a PING))) ultrasonic sensor for distance sensing and obstacle avoidance. There's a Basic interpreter (BoeBotBasic)·available from the Propeller Object Exchange and a couple of sample programs for the Stamp BoeBot have been translated to work with it
  • copacetic353copacetic353 Posts: 52
    edited 2008-11-07 03:15
    I have previous robotics experience with the basic stamps, I am looking I venture into more complex things. I am very interested in making an autonomous counter rotating quad blade helicopter, that platform is very stable and has lots of potential. I have a decent coding background, computer science is my major in college. It is all of the electrical stuff that I have trouble understanding. Basically, I can code almost whatever I want but have no clue how to put my robots together or what parts to use.

    Any further info is greatly appreciated!

    Thank you!
  • Mike GreenMike Green Posts: 23,101
    edited 2008-11-07 03:47
    Browse the Propeller forum and the Completed Projects forum. There are others who have been working on quad blade helicopters using Propellers.
Sign In or Register to comment.