some one working with quadrotor helicopter ?? we are planning to build the same using a processor 51 series. but really speaking i am new to this programming and control.. if someone can help me it would be a real help.. i need the help even for the algorithm development.. if yo can give me code i would like.
we have brought a 6 degrees of freedom inertial measurement system from sparkfun dampening which has 3 axis accelerometer and 3 axis groscope.. but dont know how to use it for the stabilization..
our first phase is to make the helicopter stable in air.. the movement and control for motion will come in second phase only.. if someone can help i would be really thankfull..
thank you xlove
You better get some high MIPs on that '51 micro to implement the multiple axis controls.
Remember it will take time to make the measurement, then you need to normalize the data, run it through a PID or Lead-Lag/Lag-Lead filter, then scale the results for the output control mechanism (most likely PWM).
Depending upon the loop-closure bandwidth expected (which I suspect would be in the 18-25Hz range), you can get a single 8051 to perform the loop-closure.
You might have to use multiple 8051's and synchronize them... IF you need to stay in the 8051 series.
I recommend that you do some simple calculations on measurement transfer, scaling/normalizing, loop-closure algorithm, scaling, and output timing pre-analysis.
And that is just to run the bare basics of axis control. Then you'll need to coordinate the system.
Seems like an 8051 would be working pretty hard on that project. (Just my initial guess).
--Cpt. Vince Foster 2nd Cannon Place Fort Marcy Park, VA
I don't mean to scare you off from the 8051 doing this project. I've been able to get a single 8051 to get that kind of bandwidth on 6 channels, but I had to be very careful on when the data was converting, coming in, and the timing on the output.
can you send me your code asap to xlove.fransesci@hotmail.com
can you send me your code asap
I just knew this was coming. don't you get it? this assignment requires a lot of work and understand even if you HAVE the code !!!!!
I meant, of course:
"...and understanding even..."
Then you really need to start with something much simpler!
Control Systems is a whole major field in its own right;
Programming in general (and embedded programming in particular) is another whole field.
How long do you have for this project? How many are on the team? Isn't it supposed to be your own work?
Take a look at http://www.mikrokopter.de, it is an opensource quadcopter.
has been mentioned in many posts.
if you use a SILabs 8051F120 you get 100 times the speed of a traditional '51
Erik
... is not much help if you don't know how to use it!
;-)
is it really 100 times faster or was it a figure of speach? that's quite an improvement!
I meant "speech" of course!
It is comparing the very first, original 8051 to the very latest, top-of-the-range technology.
The SiLabs parts (among others?) both run at much higher clock rates and execute many more instructions in each clock cycle - so you have a multiplicative gain, and I wouldn't actually be surprised if it really is 100 times.
No doubt Erik can give us the real numbers...
Quite an improvement, but by no means unusual.
Just look at PC speeds. The first I used was a genuine IBM running at 4.77MHz. A few years later I spent £1750 on a top of the range 12MHz. Now look at what you can get for peanuts.
Way beyond 100 times faster.
can you send me your code asap to...
Sorry dude, I cannot send you such code: that is both highly proprietary and classified.
Yes, an actual 8051 went into a military application: the 8xC51FB mil version cost over $200 USD. I have one at home--we lost the paperwork for it so it couldn't be used in the 'widget' so the company gave it to me. Now all I have to do is figure out what I'm going to do with a mil-spec 8051
some of this will be true for other SILabs chips, the below is solely based on the f12x/f13x
YES, it runs 100 times faster a) 100 MHz clock b) instruction cycle = 1 clock as opposed to the traditional instruction cycle being 12 clocks.
a caveat the flash (which appear as a regular 8bit flash) is 32bit wide with prefetch (appears as a cache), so when the code "runs linear" you do, indeed, have a 100 times (really ~90 times, see below) speed improvement. However when the code 'unlinearizes' (jmp) you get a penalty of a few cycles.
The reason you have ~90, not 100 times the "linear speed" is that not all instructions take the same number of instruction cycles as the traditional did.
All in all, my rough imprecise, experience based, number is that in actual applications it is about 50 times faster than a 12 MHz steam driven '51.
The chip has some very elaborate cache configuration logic and, if properly applied the perfomance can be tuned to approach 85 times faster than a 12 MHz steam driven '51, this number is VERY application dependent.