J. Mike Rollins (Sparky) [rollins@wfu.edu]
  CISSP, GIAC GPEN
Hyperbola New
My Cats New
Kitty New
Mike is on
  LinkedIn
  FaceBook
BackYardGreen.Net
HappyPiDay.com
Green Cycle Design Group
CamoTruck.Net
  Resume  
  My Stuff  
  Art  
  My Truck  
  People  
Electronics
   AC Inverter
   Taz (My EV)
   Ethernet
   PIC Switch
   LM3524 Motor Control
   Microcontroller
   PWM and MOSFET
   SMP with 317
   Power Supply
   Function Gen
   Simulator: Zener
   Charge Controller
   IR Audio
   PWM with 555
   Solar
      Truck
      Tracker 1
      Tracker 2
      Tracker 3
   Entertainment Center
   BLDC
   GPS & WD-C2401P
Jacob's Ladder
Scripts
Math
Notes
My House
My Cars
My Cats New
My Jokes
Pi Poetry
pumpkin
Toro Mower
Development
Speed of a Piston
Not a Pipe
Linux
















Tracker 3

Dec 17, 2009

I continued the solar tracking system development into a two-axis version. I figured it was time to see if the tracker could be powered from the actual solar panel. So, the tracker is much bigger than in my previous experiments. I began by making the two-axis mechanical system. I used square tube aluminum and some bolts. Nothing fancy here.

click to enlarge/reduce
I attached the linear actuators as I built the mechanical setup. I had to take some time to do the math to figure out how to place the actuators to get a good range of motion. I planned for close to 180 degrees horizontal and close to 90 degrees vertical.

I complimented the sensor with two additional sensors. I call them Sensor Right, Left, Top, Bottom and Center. The algorithm attempts to keep the voltage of the L/R and T/B equal, and keep the Center greater than all the other sensors. As the system narrows in on the direction, Pulse Width Modulation prevent overshooting the position.

The linear actuators contain built-in potentiometers that indicates the position of the linear actuator. This feedback to the microcontroller prevents the linear actuator from over-extending. I also use this to adjust the panel in a neutral position when the light levels are too low. Since I have two linear actuators, I now how two position sensors: Sensor Position 1 and Sensor Position 2. Throughout this project, my biggest challenge was keeping the naming consistent and intuitive.

I added a sleep/wake cycle to conserve power. The system will sleep for a few seconds then wake for a few seconds. During the sleep mode, the power is off to the sensors. The microcontroller turns this external power system on and off. This power system contains a 5 volt source for the sensors and a 17 volt source for the actuators. The H-Bridge is setup should reduce this 17 volts to about 12 volts.

It currently has an occasional Myoclonus jerk when it wakes up.

During the testing, I really needed to know more debugging information than what a couple of LEDs could indicate. So, I investigated the UART feature of the PIC18F chip. I had ordered some MAX232 chips about a year ago. I finally got a chance to use them. This addition was fairly simple. So, the device now has a serial debugging port for an old-fashion serial terminal.

Here is a video and the full schematic diagram.

click to enlarge/reduce


Place the mouse over the image to view a time-laps photo.
It may take it a second or two to start.
The Pepsi bottle is a counter-weight. When it is balanced, it is easier to turn.