So a store I was at was having a sale on solar garden lights. They were so cheap, it was almost worth buying them for the white LEDs alone, the solar cell was a bonus! So I picked up a bunch, ripped the guts out and tried to figure out what I would do with a ton of solar cells. My first thought was to make a dual 555 timer photopopper using two gear motors salvaged from a hand held camcorder from the 1980s. This was initial design layout.
I really wanted the solar panels to be the backbone and have a cool curve, so I took a random piece of cardboard and bent it into the general shape I was going for, then I arranged the solar panels on top to get an idea of where I was headed.
I was still planning to make a photopopper at this point using the stepper motor heads from two different handheld camcorders as the wheels.
I liked the way the design looked so I took some scrap metal I had downstairs, bent it into shape and glued everything together! The solar cells are wired up so that four groups of two panels in series are connected in parallel.
I still think this is one of my cooler designs!
Then, I was thinking, all that cool and its just gonna roll around all slow like, seems like a waste. So after some reading of www.beam-wiki.org I decided that I wanted to go for a walker, but what kind of walker? Since this was going to be my first walker ever, I had quite the challenge ahead of me. So I got my hands on some small, lightweight hobby servos and hacked them, as outlined here: www.beam-wiki.org/wiki/Hacking_a_Servo.
Still not sure what design I'd run with, I stumbled across Wilf Ritger's uCrawler, www.beam-wiki.org/wiki/Ucrawler. This design is genius, a one motor, two legged walker, that should be an easy first project and fit nicely under my solar cells. On Wilf Ritger's page, he outlines three design iterations for the micro crawler. The first being a simple feedback design using the wiper pot to provide the feedback for a single neuron. The second one is a bit more complicated and adds a centering trimmer and two LDR's for phototropic behavior. The third is a bit beyond me, haha.
At any rate, he outlined using a 74HC14 or a 74HC240 in all of his designs. I do not have either of these chips, but I do have a ton of transistors and 555 chips. So I wired up one 555 to work as an inverter. The input for this 555 is the wiper of the servo and the output feeds both one side of the 6 transistor h-bridge and the input of the second 555 chip. The second 555 chip is set up as an inverter as well and its output feeds the other half of the h-bridge. This got my servo swinging back and forth just like intended and the LDRs have enough of an impact on the system that you can visibly see the servo moving differently depending on which one has light.
I ended up copying his second design and adding in essentially a second neuron doing the opposite of the first neuron to help control the Tilden based 6-transistor h-bridge. Although, to be honest, I think it's more of a mix of his second design and his third design, haha. I really wanted to make use of the solar panels, so I had to figure out a way to make the bot stop in bright light, cut off current to the main circuit and allow the solar cells to charge the rechargeable batteries. Enter the LM339! I could have done it with a smaller comparator, but this was all I had, haha.
When the photocell isn't receiving enough light, it kicks the main circuit on and the bot uses it's phototropic behavior to search around for the brightest light source! Or at least that's what I hope it does, haha. Here's the final circuit before I started soldering it together!
Head to Page 2 for pictures of the construction!
Check out the rest of my robots here!