Shop OBEX P1 Docs P2 Docs Learn Events
Modify Navigate the Maze for 3 QTIs — Parallax Forums

Modify Navigate the Maze for 3 QTIs

To cut costs, I'm only using 3 QTIs on each boe-bot for my class. Last year we had so many problems with our IR pairs that I am really excited about using QTIs for our maze project.

Would this work to modify the example Navigate the Maze for three QTIs?

CASE % 0110 becomes CASE % 010
CASE % 1000 becomes CASE % 100
CASE % 1100 becomes CASE % 110
CASE % 0001 becomes CASE % 001
CASE % 0011 becomes CASE % 011

Any pointers or advice?

Thanks,
Diane

http://learn.parallax.com/project/boe-bot-maze-navigation-qtis/maze-setup-and-calibration





Comments

  • And I would omit CASE %1110 and CASE %0111
  • Hi Diane,

    It works fine for basic line following. Intersection detection like in the link you sent is probably going to get more challenging. It can be done with 3, (or even 1), but the code will have to be modified. To really simplify intersection detection, wide stripes and the outer 2 forward of the inner 2 makes the code the simplest. The code would then use the inner 2 to follow the line, and the outer 2 to detect intersections. With 3, the outer 2 might be making course corrections OR running into an intersection. So, the added challenge is making the program discern whether the Boe-Bot needs to make a course adjustment or respond to an intersection.

    Each stock Boe-Bot kit has parts for an additional 2 student-made QTIs. All a QTI is is an LED shining light at the stripe and a resistor-phototransistor-capacitor connected like the light sensing circuit in chapter 6. More info after lunch.

    Andy
  • Okay, lunch was good... Note the schematic in on page 2 of this doc: https://www.parallax.com/sites/default/files/downloads/555-27401-QTI-Line-Sensor-Documentation-v1.0.pdf. The QRD1114 is just an infrared LED (like Robotics with the Boe-Bot, chapter 7) and a phototransistor (like chapter 6). Although the charge transfer circuit looks different from the one in chapter 6, it should still work with the test code that uses the RCTIME command in chapter 6.

    Each LED should be in its own tube, and likewise for each phototransistor. The tubes should be opaque, like a straw wrapped in electrical tape, or heat shrink tubing from electronics suppliers. Heat shrink tubing does not have to be heated to make it shrink, it's just an alternate construction material. The tubes should be arranged to make the LED point at and illuminate a spot that the phototransistor is also pointing at, up close, just like the QTIs. Start testing them on the breadboard first with electrical tape on poster board. Make sure to cast shade on the system, so that the ambient lighting is similar to what it'll be under the bot. The IR LEDs will make the circuits more sensitive than the red LEDs. Other options for increasing/decreasing sensitivity include using larger/smaller distances, or larger/smaller resisters in series with the LEDs. Maybe a red LED with a 220 resistor will work best, or maybe an IR with a 1 or 2 k resistor. Testing will tell.

    Mounting can be done with Popsicle stick or 3d printer technology as well as anything in between. Wires and connectors are available through electronics outlets, though I typically prefer these: https://www.parallax.com/product/800-00048.

    Hope this helps, and that it enhances the activity. I would be interested in what happens if three QTIs focus on line following, and then the two DIY sensors are mounted further in front (and to right/left) for early intersection detection.

    Andy
  • Thanks for your help. I think that sounds more advanced than my abilities. I'll just have the kids partner for this section this year, and buy more QTIs next semester.

    Thanks again!

    Diane
Sign In or Register to comment.