Saturday, August 10, 2013

What do we do next? date10th August.

The most important thing to take from the mess is that we need to work on what we have as inventory, and make a list of everything. Who has what should be put down on paper.
Apart from that, the next useful thing that is close to completion is getting data from the IMU. Now, there is a question, how do I know if the IMU is working or not. Now, I did make all the right connections. It was double checked by Nitin. Have to try to get useful data from the IMU. The second thing that can be done is learning as much as possible about Kalman filtering. Both these things will be done by tomorrow morning.
We checked the IMU, apparently not a single chip on that board seems to work. They just don't work, which brings me to another unhappy conclusion, the board is fried. Now, Lv and me got the copter resoldered we were looking at the next qtestion

Motors don't beep and IMU sends three zero x as data

Right now, feeling really weird and sort of screwed up. The unbelievable bit is this. I committed and I really mean committed to getting the entire operation i.e. the penta running, in these three days. Unsurprisingly, we got stuck at step 1. Our idea was to divide our work into the following stuff that we need to do-
1.Calibrate the motors.
2. Get data from the IMU and this time, meaningful data i.e. not a single stream of 30x thatdoes not change no matter how much you turn or swoosh the IMU in midair like you were making it takeoff instead of the penta!
3. Make the PID loops get this data and get the outputs. Now, transform the output to meaningful PWM
4. Write two kalman filters.(The one thing that I really understood about the kalman filter is that it can predict a position(the one that we want), then correct our present path to reach that position and then estimate how close we are to our defined path)
5. Get data from the UltraSonic sensors, the GPS and then see how this data fits in with everything else
 Where did the screw up happen?
1. Ok. The motors didn't beep. What the hell? Is that even possible? The beep means that the motor is ready to be calibrated. Ok. Calm down, I said to myself. Then, I checked each motor and ESC. None of them beeped. Yes I am using the Turnigy product D2836/6 motor with a 30A ESC also bought from HobbyKing. None of them worked!
Ok.Calm down. There has got to be a rational explanation for this. You know you didn't touch the motors, or the ESC or the penta until now. Eureka! The battery must be out of charge. Then, again, didn't have the battery charger. Ok, what else could I do? I remembered that I had another motor that was for a different spec, and voila! It worked like clockwork, and beeped as well as roadrunner on the looney toons. Now, I am getting weird. How and what could this mean? There is no way that all the ESCs can do the same thing.
2. Ok, that didn't work. So, I tried to do the second best thing. Getting data from the IMU was pissing off I had weird errors with my ARDUINO. One, it would forget the COM port after two trials of uploading the code and receiving the values. Two, there was no way of knowing what data I had. I mean that the data was constantly in the same range no matter what. Ok, that didn't work. The next best thing was to use the breakout pins, and get some data off of it. Now, this time, I did find something interesting. Hopefully, this meant that the board worked. two things
-1-The VCC pin gave a constant reading of 1023
-2-The ground gave a constant reading of 0.
It wasn't much but, it gives me hope.
Now, the values I got from SDATA and SCLK weren't too interesting. It might as well have not been done. The data shows no change when you move it in the air, abruptly; no change when you rotate it about the three axes. I mean a meaningful stream of data. And, that proved that the data did not come from the gyros or the acceleros.AAAAAAAAARRRRRRRRRRGGGGGGGGGGGHHHHHHH!! I feel like Juan Martin Del Potro, though I know I'm a noob right now at this whole business, but, I'm gonna get better.