Ashraf HZ
Ashraf HZ
Communications
08 Nov 2007

linked sensor problem in Line Following Robot

Hey CEans!

*looks at xheavenlyx's posting template* 😛

Im working on a line following robot in a group that uses the PIC16F877A uC and the L293 motor driver.

The sensors are simply three pairs of LED and LDR in a triangle shape layout. The bottom two act as the line detectors, while the front one is used when turning 90[sup]o[/sup]. The power supply is a normal 9V battery that is regulated to around 5V, which powers both the PIC and the sensors. The line we want to detect is black, the background is yellow. All sensors share the same 5V power and GND, but seperate outputs. The pins on the uC we used for the sensors were AN0 (front) ,AN1 (left) and AN2 (right). Components used for each sensor is a standard LDR in series with 20K ohm resistor and LED in series with 1k ohm resistance. See below:

[​IMG]

The problem arose when we wanted to calibrate the sensors to see at what level would the sensor detect the black line. The program we use to see the uC output is PICBootloader +. When we placed the front sensor on black, we get an increased reading on the left sensor as well, even though that one is on the yellow background. If we put both the front and left on black, the left registers a VERY high output, as if it was a summation of the front and left outputs. I asked a friend in the lab, he theorized it could be due to current drain or something. Basically, they are linked somehow. Ive checked the wiring and soldering, the outputs of the sensors are not touching each other. We even switched the analog ports for the uC, but still got the same problem.

Anyone has a clue on why the sensors affect each other? 😔 one thing to note is that the resistance of the front sensor is slightly larger than the left sensor, cause we couldnt get the exact resistors for each. Could this cause a problem?
Ashraf HZ

Ashraf HZ

Communications
11 years ago
Hm, I found this link out:
https://www.seattlerobotics.org/encoder/200011/Line Following.htm

My first attempt at having a finer level of feedback from the line following array was simply a linear array of sensors. This didn’t work out very well for several reasons. First of all, the emitters for adjacent detectors were spilling over and making the signal for the line position very faint. I was using analog readings and the high/low reading of the center detectors was simply not very good. The second reason was that I soldered in the sensors directly onto my breadboard. It turns out that the heat from the soldering iron was affecting the matching between sensors. I would get unpredictable and wildly varying readings from individual sensors.
I wonder if this was the case with my sensor? Perhaps my friend's soldering gun was too hot compared to normal ones. In any case, his arrangement of sensors is identical to ours.. just that ours was really crude compared to his slick version using photoreflective detectors. Gotta love his differential detecting method 😉 If anyone else has encountered this prob, please post your thoughts!

Share this content on your social channels -

Only logged in users can reply.