Solar Bot

In Energy, ITP

For the solar project, we were asked to build a Solar Engine Beam circuit (see Jeff’s Video) to provide some analog logic to little solar bots. Marcelo and I worked together and explored a variety of different configurations to see what would work best for this idea we had of a spinning bug.


To start out, we knew from the measurements we did in class that the open circuit voltage of the little solar panels we were using was about 4.6-4.8 volts for us. However, as we’ve learned, once you add a load to your circuit the voltage decreases form the OCV. This is relevant because for the first circuit we actually got working, we were using a 4.2 voltage detector. The panels that we connected in paralell never actually reached the necessary threshold to trigger the voltage detector since it was really firing at around 4.5/4.6 voltage based on our multimeter readings. So we decided to put the panels in series to increase the voltage we were generating. At that point, things were pretty rosey. We had two DC motors triggering as expected. In this circuit below, we used a 1 Farad capacitor for energy storage and a 47 micro Farad capacitor as a the timing capacitor on the voltage detector.




Unfortunately, when it came time to fabricating the solar bot, this circuit that we replicated on a proto-shield no longer worked as desired. It some how lost the logic that we created through the voltage detector and transistor. Instead of turning the load on and off, the load would just stay on. Jeff later told us that it’s possible these circuits can behave unpredictably and then ultimately reset at some point. But without knowing that, we decided to follow the suggestion of a classmate and switch our NPM transistor from the pn2222 to the tip120, which solved the problem. One more thing to mention is that we concluded current was going to be more important to us given our low voltage motors and that a higher current would result in our little bot making more substantial spins. Therefore, we switched the panels from in series to in parallel and replaced the 4.2 voltage detector with the 3.2.



Now that everything was working nicely again, we stared fabricating the rest of the bot.



We thought we were done at this point but it turns out that the weight of the panels and the enclosure was too much for our motors to handle. So while the wheels would spin a ton when we lifted the bot off the ground, the minute we put it back on the table, nothing would happen. Our solution was to reduce the friction of the bottom part of the enclosure dragging on the ground and distribute the weight more evenly to allow the motors to move more easily. We temporarily achieved this by giving our bot some “resistor” legs that the enclosure could rest on as you see below. The only problem was that sometimes it would work great and other times not at all based on the angle of the resistor wire ect.





The demo above worked great but it was nearly impossible to recreate that success again. Therefore we decided to create a support that would provide the least amount of friction and give us consistent results. Thankfully, Brady had a couple extra ball-bearings and we used that instead of the “resistor” legs.




The results were less dramatic then in the first video, however, the ball-bearing support allows us to get more consistent results. I hope it will work well in the terrarium!


Lastly, I wanted to break down how much energy our circuit was providing our two DC motors. Based on the fact that we are using a 1 Farad capacitor and that the voltage release is 1.1/1.2 volts when the voltage detector triggers (cap drops from 3.13 to 2 as you see from the video of the oscilloscope and multimeter), the circuit is providing the motors with 0.565 joules.



Given that the motors stayed on for roughly 7 secs, the circuit provided the motors with 0.081 Watts.



Submit a comment