pic: Chantilly Academy's 4 Wheel Drive 4 Wheel Steering!!!!

[psquared]

right now it is joysticks, and we are still in the testing phase to get it running exactly as we want, if we are able to get everything running perfectly before friday, which may be a stretch, then we might be able to put together a nice looking steering wheel, gas pedal and reverse shifter, but all of that would take a lot of time, and we are tight on time right now, so as of now, i would say we will only be using joysticks, unless a miracle happens

[Conor Ryan]

how fast does it take to do a 360? Nice engineering btw.

[Mr. Ivey]

Cant wat to see this at VCU. I will have to see how all this works. Great job!
ivey

[CRAZYMADI]

hardly five to ten secs to turn 360`. We just needed to do the final test and w will be ready to go. Our arm is still under work and will be ready befreo the DC scrimmage. Its also light weight and can go upto 12 ft.

[Andy Baker]

Sweet design. This is one that we all wanna see up close. Great job!

Andy B.

[Lil' Lavery]

Originally Posted by CRAZYMADI

hardly five to ten secs to turn 360`. We just needed to do the final test and w will be ready to go. Our arm is still under work and will be ready befreo the DC scrimmage. Its also light weight and can go upto 12 ft.

Oh god, 360s....(anyone on 116 will understand, and soon enough the rest of you might too...)

Question about the joystick drive though. Are you just having the X axis serve as a throttle and the Y as steering, or are you using some other more ingenious way?

[psquared]

the joystick values are from 0-255 in the x direction we convert the x-value into a degree measure from -45 to 45 degrees, then after some math, we have the new angles that the wheels need to be, then on the same joystick in the 0-255 y direction, the speed will be controlled, the speed is controlled based on the 0-255 value and on the angle the wheels are currently oriented, the arrays to hold the values were quite large, and to do it on the fly would be too slow, so we had to do some creative handling of the values to keep memory low

so in short:

-each (0-255)x value corresponds to a relative angle with 127 being straight ahead
-each (0-255)y value creates the velocity for the robot by combining aspects of the angle and its value

what i would like to do on sunday monday and tuesday, not saturday and tomorrow b/c we have to get ready for the DC scrimmage is make a steering wheel out of potentiometers, a gas pedal with a potentiometer that would give a value 0-255, and a shifter for forward and reverse, the steering wheel would work exactly the same, the pedal would be halved to give us a 0-127 value, and the shifter would say if in reverse add nothing to the 0-127, otherwise if in forward add 127, however this is largely time consuming and for right now we want to make sure that the basics are perfect,

hope that answers your question

also, i believe on another thread with a pic there is a description of the math involved

[EricH]

Looks good, but the back right wheel doesn't seem to be steering in the picture. I hope that is fixed soon. Also, could you go at an angle by lining all the wheels up the same way?

--Eric

[psquared]

don't worry all the wheels turn, it might just be the angle of the pictures, as per the all going the same way, its not going to happen, that was in our original design, to be able to shift between same direction steering and the way it currently is, but due to some complications we had to throw that way out, and the only way it can do it now, is how it is looking in the picture

our original design could --> had to scratch that for a different design --> this design can't b/c of the way the wheels are set up, but current design is still good

[CRAZYMADI]

Actually we;ve used two trig function at the same time to program the sterring mechanism. It is also posted on the forum.
The tyres needed to be turned with two different angles to keep the center point allighned.
For example if the left tyres are turning at 45` the right tyres needs to be turning at 26` angle to keep the center or origin of the circle same. This whole system needs preety neat programming, and i appreciate the efforts of our programmers to come up with these functions.
Not mention ing the fact that left and right tyres needs to be turn on different speeds, to prevent the robot from tipping over to one side. Once again, too heck of programming and trig. Thanks to Our all AP Calculus team members who really helped biulding this thing.

[psquared]

to those of you expecting to see this in action at DC tomorrow:

-we tested the steering code and it works perfectly when propped up on blocks

however...(yes there is a however)-
today was the first time programming team had a chance to test out how the robot would perform with perfect steering code on the carpet, well there is a small design problem that is keeping the wheels too loose. if you look in the picture where it says LF, on the right hand side, that is part of the thing that controls how our wheels steer, it slides up and down the shaft in the middle to create the angle for the wheels, well the length of that middle section in the z-direction for the photo, (if left and right is x, up and down y, and forward and back within the picture is z), is too short, by being too short this creates a situation similar to a see saw. if you had a 2 fulcrum see saw with both fulcrums exactly in the middle, you would get a lot of movement on both sides, if you moved the fulcrums towards opposites sides, you would eventually get to no movement when you have both fulcrums at opposite ends,

well right now our see saw for the steering is about 33% and 66% the length of the see saw, and that is creating too much non-needed movement for the wheels, after the DC competition tomorrow we are going to fix that by extending the middle plate outwards to maximize the distance between our "fulcrums", this should solve our steering problem as of right now, also one more small design problem we noticed, if you look at the black circle on the left hand side middle, that piece attaches to another circular piece which is controlled by a motor to move the wheel component up and down the shaft to steer it, this is not a sprocket gear, but something else that i don't know the name to (haha, i am a programming not a design team member ), well that isn't working b/c in the simplest turns it is not a sprocket gear and does not have the advantages of being one, it slips too much

however to all of you that really seemed enthused with the design, don't worry, we have figured out all the problems with the steering, and although they won't be able to be corrected by tomorrow for the DC scrimmage, by richmand vcu and atlanta, which means by tuesday afternoon, they will be ready, and i hope (and am sure) will be very impressive, wish us luck in fixing this problem

talk to you next time...
~Programming Team 612