Starting my first scratch built robot

Community General Discussions
1

Hello All,

Very informative forum you have here. I have been lurking for a couple of weeks and decided it was time to join up.

I built my first and second robots about 4 weeks ago from an I-Box "Robo-Box" kit, an O2 obstacle avoider, and a "NanaTank" edge detector. I later modified them to be line followers and object seekers and pushed the limits of the RoboBox controller. It seems like the Robo-Box LogoCricket microcontroller is sort of a dead end, but good enough to get started. I learned some elements of programming, played with infrared sensors and line following. Had a bit of difficulty because I switched all of our home computers to Ubuntu last October only to find that I could not make the CricketLogo IDE run under Linux or WINE.

I ended up resurrecting an OLD PC from a garage sale just to have a Win2k box to program the robot. Actually that worked out OK because I was never able to get the USB to RS232 adapter to run properly, so running from a real serial port helped alot. Eventually I hope I learn enough to install a different microcontroller chip in the I-Box...maybe something that will do C++. Anyway, the RoboBox kit came with a really nice set of sensors and building components that I can use on future robots, so I remain quite satisfied with the kit even though I have already outgrown it.

My third robot is on the boards. It's a "Sandwich" line follower from David Cook's Robot Building For Beginners. I'm really enjoying building this robot with the Solarbotics parts pack, and I highly recommend Cook's book for anyone who is starting out in robotics and who knows little or nothing about electronics. Even though I knew slightly more than nothing, the book has been a great teacher about things like how to use a comparator, how to calculate and measure voltage in a voltage divider circuit, how a transistor works, and how to really use my digital multimeter. I dug out my soldering irons which had been in boxes for about 25 years

My SANDWICH is not yet complete, but I have all the circuitry running well on a bread boad and am in the process of transferring it to the PCB. Meanwhile I take a break from the electronics by building up the chassis. The last step will be to mate the "body" with the "brains", and for that, I am waiting for a package of Molex connectors I ordered from Jameco. I want to be able to easily swap out things like controllers, motors, battery packs, sensor packages and so on, so it seems logical to me to start out right away by using Molex connectors to connect any component that is not mounted directly to the PCB. Eventually I think I'll want to replace Sandwich's primitive brain with a microcontroller and I want to be able to do that without having to unsolder anything.

Anyway, I can see the light at the end of the tunnel for my Sandwich, so it is time to start planning my fourth, and all of the skills and knowledge I gained from my first three robots are now coming together as I plan my. I found an article on SERB, a servo powered, Arduino based robot. I toyed with the idea of buying the kit, but decided that it was a bit pricey. Instead I downloaded the cutting template and am having a friend who is good with a scroll saw cut the chassis components for me from acrylic sheet. I've been fooling around with an Arduino for about 6 weeks and learning a lot from it, so I think I will build another Arduino and mate it to a small breadboard to create a simple platform for experimenting with the the Wiring and Processing dialects of C++, and using Bluetooth, or something, to send sensor and maybe even mapping data back to a PC .

Anyway, I really enjoy reading about everyone's projects on this forum.

Cheers,

Chris

2

Sandwich is finished

We (my seven year old and I) put the finishing touches on our “sandwich” robot on Saturday. We didn’t find a Ziplock container that would fit all the components, but we found a pencil box that works pretty well, so we call our robot “Pencil Box”. We powered up our Pencil Box with 6 AA Nimh cells wired in series.

We did initial testing on the kitchen floor, but mommy wasn’t too happy about having a loop of electrical tape in the middle of her kitchen so after we had Pencil Box tuned up a bit to follow a dark line on a light surface, we moved to the family room and tested him out on the carpet. I was surprised that he was able to follow a black line line over the textured cut pile carpet. We set him up and he ran around a kidney bean shaped course for a couple of hours, handling both left and right turns with ease.

Since we didn’t have a dark surface to test him on, we put down a second tape line around the outside of the first, building a “race track” so we could test his ability to follow a light line over a dark surface. Pencil Box remained on track just fine.

We did discover that after a couple of hours of running, Pencil Box starts to lose track of the line on the sharper curves. He uses a pair of white LEDs to light up the floor beneath him, and the light from the LED’s reflects back to two symmetrical pairs of CdS photo-resistors to sense the line. We think that as the voltage drops, the light output fades a bit and because our line is black electrical tape over a brown carpet, the contrast is not as apparent as the light fades. Pencil Box does have a 20K trimmer pot that allows us to dial the brightness up a bit, and this seemed to correct the problem, but we don’t think it’s a good idea to be fiddling with the trimmer pot repeatedly, we know these are kind of fragile and not really made to be constantly adjusted.

We are thinking about adding a switching voltage regulator to the power circuit that will limit the supply voltage to around 5 volts, then we will re-tune the sensors and headlights to the best performance at a lower voltage level and see just how long we can keep him running.

Our next goal is to learn how to duplicate the functions of Pencil Box I using an Arduino micro-controller. We are working on a chassis for Pencil Box II. We’ll give Pencil Box II a pair of motors that are capable of a bit more speed and then initially we’ll dial the speed down with Arduino. We know the reason we lose the line is that sometimes Pencil Box overshoot it. We’ll try to program Pencil Box II to follow the line more smoothly to reduce the oscillations as the left and right motors switch on and off. We will try to maximize speed around the course, and then we will try to build in some sort of a memory function that will help Pencil Box II rediscover the line if it overshoots.

3

Congratulations!!!

I’s good to know about your experiences with the book and your observations and projects. I had and I still  have good times with my kids and sandwich. Congratulations to you both and I hope you can share some photos or video of your pencil box. I can’t wait to see it.