Lab FollowBot

Lab: FollowBot Introduction For the great 111 final project challenge, my partner and I decided we would attempt to design a simple robot which followed objects in front of it. It would accomplish this by sending out an infared signal from an LED and then it would pick up the reflected signal with a phototransistor which luckily peaks in the IR part of the electromagnetic spectrum. The robot would then compare the two signals and based on which one was higher, it would drive a DC motor accordingly so that the robot would approach the object in front of it. In a similar manner, we would use a third photo transistor and compare it to some predetermined offset. The output would then tell the robot whether or not it was still okay to go forward. Driving the LED At first we decided to drive the LED with the signal generator to test it in order to make sure we could measure the initial signal and also the reflected signal. These both worked. However, in my personal foolishness, I accidentally connected the special IR LED between the two terminals of our 6 V battery power supply. In doing so, our LED exceeded its maximum power rating and was no longer usable. We still had one more though and I vowed to be more careful. I realized then that we would need to drive it with some oscillatory signal so that it would not overheat. We decided to try to make an LC oscillator. We purchased an inductor but we could not really figure out the way to actually use the LC circuit to generate any sort of sine wave. Next, I decided to scrap the whole oscillatory idea and just do it with DC signals so I began testing current limiting resistors to find the best one to use with our IR LED while still having it produce intense enough light for us to measure. This was, of course, futile and resulted in me burning up our final good IR LED. We had purchased several backup IR LEDs but their output intensity was pitiful in comparison to what we were seeing before. We were able to pick up signals but only like an inch away which was like still on the mini robot thing we had so it was kind of silly. Thus we decided we would use a white LED and place it far enough away to prove our point. Once more, I realized we needed to drive the LED with an oscillating signal or it would be futile. We decided instead to make a relaxation Oscillator which produced a 7 V amplitude square wave which we used to drive the LED. The schematic for said Oscillator is shown below in Figure 1. It oscillated at 9.289 kHz. The output of said circuit across the LED is shown in Figure 2. 1 Figure 1: Relaxation Oscillator Figure 2: Relaxation Oscillator output across LED Measuring the Signals Our original intention was to measure any sort of IR signal with the phototransistors we used in one of the previous labs. However, since I burned out both of our good IR LEDS and we had to switch to a white LED, we decided to switch to photocells instead. In principle, they accomplish the same goal in that they measure the amount of light received. In using both, however, there seemed to be differences. The phototransistor output would oscillate from 0 to whatever the max was for some given distance at the same frequency of the relaxation oscillator. The photocells had a different output. They work in that they have some resistance based on the amount of light they receive. We placed them in between our power source and the next stage of our circuit and measured their output in reaction to the LED. It was quite interesting actually. It would hop up to some output then oscillate some small amount above that. This is shown in the scope trace in Figure 3. It appears to be acting as some sort of integrator. The specifics are not entirely important as long as both photocells seem to have a similar output and our goal is to compare the two signals. We ended up passing the output of these signals through an RC bandpass filter to try to get rid of any unwanted signals. The gain for the two filters is shown in Figure 4 as a function of frequency which we made to make sure the two filters matched well enough that it would not really affect the measurements, though it cannot be perfect. The cutoff points were supposed to be at around 8000 Hz and 12000 Hz with the resistors and capacitors we used. The schematic for this part of the circuit is shown in Figure 5. In the schematic I have included the original phototransistor as it would still work if it were there and we had an actual IR LED. For now, however, just imagine there is a photocell there which acts sort of as a variable resistor, which is high when there is little light and very low when there is a lot of light. 2 Figure 3: Band Pass Gain Figure 4: Output across Photo Cell Figure 5: Bandpass Stage Comparing the Signals Now that we have the signals we want, we should turn them into usable signals that we can compare with a comparator op amp. We do this by using transimpedance amplifiers on both signals. Then we feed the outputs of these signals into the two inputs of an op-amp. The op-amp acts as a comparator and will output to one of its rails, depending on which input is greater. The schematic for this part of our circuit is shown below in Figure 6. The inputs are taken from the outputs of the bandpass circuit. 3 We can then use the output of the comparator to drive our car right? Wrong. The comparator outputs too low of a current to actually drive the motors. After discussing this problem with a GSI, we decided to use a push-pull circuit like the one we used in Lab 8. Figure 6: Comparator Stage Driving the Motors We built a push-pull circuit as shown in Figure 7. It’s output on the scope for an input 1 Volt peak to peak sine wave is shown below in Figure 8. We tried using it on one of the motors and then fiddled with some wiring so it would turn the correct way for the light’s position. However, if only one wheel turned it would be kind of lame so we threw two diodes onto the output of the push-pull circuit in parallel. One was reverse biased while the other was normal. As is such, we had to flip the inputs of one of the motors so that they could both turn forward. Figure 7: Push Pull Stage 4 Figure 8: Push Pull Output We threw in some switches just to make it easier to turn on/off and it marginally worked in that it did indeed follow the light source. Theoretically it would also work by following whatever was reflecting the infared signals we originally wished to implement. However, it worked with what we had so I was very happy. Further Improvements After finishing this, we were essentially out of time that we could spend in the lab. However, before leaving we came up with an idea to make the robot even better. We decided it would be simple enough to add another sensor stage to just measure the overall light received. This was in an effort to make the robot move forward or backwards until it reached some designated stopping point. We decided that it would be a good idea to measure the light, create an offset, compare the light to said offset, and move the wheels accordingly using the same push-pull circuit mechanism we used previously. It would of course be a much better idea to turn the AC signal we measure into a DC signal but I think this would probably work well enough to accomplish what we wanted to originally. The offset is completely arbitrary and could be replaced with just some Vof f set for practical purposes. In the current version I wrote the schematic for in Figure 9, it would be approximately .6 V which is actually probably too low. A better guess would probably be around 2 V. If the measured amount was lower than the offset, then the robot would continue to move forward, whereas if it were higher it would begin to move backwards using the push-pull mechanism and the comparator as shown in Figure 10. Alas, as I write this, I have realized that the way I made the circuit diagram it would do the exact opposite of what I just described and I unfortunately will not have access to the program I used to create the schematic before I have to turn this in so I’m making a note of it here. It would of course be a simple fix as you would notice it doing the exact opposite of what you wanted and you would just need to switch the motor leads. Figure 9: Offset and Sensor 5 Figure 10: Offset and Sensor Conclusions This project was actually quite a lot of fun when we weren’t trying to figure out why our push-pull circuit wasn’t working. I learned a great deal about electronics in general and found it really cool to be able to do something very physical. Ultimately we wanted it to be able to play music when it detected something and do other strange things for novelty which I actually feel very confident doing now from this class and especially this final project. 6