Servo - May 2009

REDUX

This month, I had originally planned to move on to a new TankBot project but I quickly realized that in order to do so, I would first have to dismantle our previous project because it took up virtually all the space on the TankBot’s breadboard. I couldn’t bring myself to do that. Instead, I opted to simplify the IR circuitry so that it would be smaller and therefore able to coexist on the breadboard with the next project (or two) that I have in mind.

Simplifying the IR Object-Detection Circuit

In the March issue of SERVO Magazine, we implemented a simple IR obstacle-detection system that enabled the TankBot to explore his environment on his own, without any intervention on our part. The project used a PICAXE-08M “slave” processor that emitted short bursts of infrared light and then “watched” for echoes in much the same way that a sonar system detects the sounds that are echoed back to it from nearby objects. One of the things we discovered along the way was that the IR system was surprisingly sensitive; it required some detuning and other techniques to decrease the probability of false-positive echoes being detected when there was no obstacle in front of the TankBot.

Figure 1 shows the new circuit schematic and Figure 2 is a photo of the new parts layout on the TankBot’s breadboard. There’s no need for a parts list; you just have to remove the previous project’s extra parts that we won’t be using this time. Actually, there is one small addition to the circuit: a 100K resistor that ties the 08M Serin line to ground (we’ll get to that shortly). As you can see, the circuit is considerably smaller than our previous version. For one thing, I reduced the number of IR LEDs from two to one. Besides saving some space, I also hoped this change would reduce the circuit’s sensitivity and therefore further decrease the

72 SERVO 05.2009

number of false-positive echoes. This, in fact, turned out to be true; the revised circuit is definitely an improvement in this regard. I was able to save even more space by removing the 08M’s programming circuit from the breadboard. This may sound like a strange choice, since we certainly need a way to program the 08M. One possibility would be to do so on a separate breadboard and then move the 08M back to the TankBot’s breadboard. However,

this approach would quickly become a nuisance because, as you well know, program development always involves many cycles of testing and program modification.

I opted instead for a different approach – one that involves constructing the programming circuit on a separate board that can be easily inserted and removed from the breadboard. This decision is what necessitated the addition of the 100K resistor mentioned above.

In order to run a program, the “Serial In” (Sin) line of any PICAXE processor must be tied to ground. Normally, the combination of the 10K and 22K resistor in the programming circuit is what fulfills this requirement. However, if you remove the programming circuit (as we are about to do),

the Sin line is left floating and the PICAXE can’t function. I have found that adding a 100K resistor to ground solves the problem and also doesn’t interfere with downloads from an external programming circuit; it can just be left in place all the time.