New Navigation and Fried Boards

So I tried to implement battery voltage measurements into the current configuration. That turned out to be a bad idea and I guess one good surge must have fried several pins in the Atmega2560. Now there is a Arduino Mega at the controls and life is looking better. Also, for those that don't know the Mega had a price drop recently - 50 bucks.

Am really excited about the announcement from Arduino about the Due - an Arm based processor that is 32-bit and 90Mhz. Should be available in the next few months and that is really cool. http://blog.arduino.cc/

Also have implemented 'navigate to a gps point' . Managed to send HARVE down the street, to a particular lat/lon and turn around and return to starting point. I will try and post video soon.

HARVE Mk I

So I finally decided on a name - HARVE (an nod to an old friend) - and it makes for a good acronym. Half Assed Robotic Vehicular Explorer.

Video

This is far from my finest work with a video camera (actually a Canon Elph) but pretty cool.

Holy Crap It WORKS!

Well now I have to admit that going back and forth endlessly is not really all that exciting but its a start. Really just needed some sort of quick program to load in and demonstrate that it actually works. Managed to wire the motors to the wrong sides at first which left the thing going in endless circles. Also had to figure out the front from the back but I think that the 'tail' has decided that for me. I hope to add a second higher capacity battery that will provide the motors with power while the smaller one runs the Arduino Uno. I also have an LCD screen with a couple of buttons that will get added to the stack. Right now its running on 7.2V RC car batteries and it would make sense to move up to Lithium Polymers for the weight but that will happen when I fall into a bathtub filled with money.

Took long enough but all things considered (like my relative ignorance and lack of a research budget) I think its off to a rollicking start. Now I just have to make it do something interesting.

Close but no cigar

So it is now almost reality. I did a test assembly of the frame, motors and wheels. Things look good. I still have to put on some sort of top plate for the electronics and batteries, wire the motors and finish the side plates. Oh, an program it to do something. Its the little stuff. Hopefully by this weekend it will at least be moving on its own power!

motors, motors everywhere

So it turns out there is a bewildering array of DC gear head motors out there that one might employ in a robotic adventure. I finally decided on these guys from Electronic Goldmine (they are a surplus dealer so no guarantee of how long the link is good)
http://www.goldmine-elec-products.com/prodinfo.asp?number=G18120
They mount with 3mm metric screws and at 7.5 volts they draw about 0.5 amps and seem to draw less than an amp with a pretty good load on them. We shall see how they fare in actual use. My guess is total stall current is around 2-3 amps. With four of them that makes total stall around 8-12 amps, which is really high but that is a concern with any differential drive or multimotor robot.

So just for those who dont know there are a couple of different motor types and things to consider. Cheap DC motors are designed to run efficiently around 1800+ RPM. The speed of a motor varies directly with the voltage applied (more voltage more speed - until you burn it out). The current a motor draws is a different story. At "free running speed" meaning optimum voltage and no load, is where a motor draws the LEAST amount of current. As the loading increases the current draw increases until the motor hits stall (which is what it sounds like). This is where the most current load occurs. This is true for both brushed and brushless designs.

To slow these little guys down you use a motor controller and pulse-width modulation but even then they tend to run a little fast so you employ a gear head to slow them down. This also has the benefit of increasing the torque and the slower you run the more shaft torque you will have. For really accurate positioning you need to employ some sort of wheel encoder to track the number of wheel revolutions. I dont have time to talk about stepper motors but I am curious to try one for informations sake.

More later about the electronics to run these guys.

Adventures in anodizing

So there is a lot of information on the web about anodizing aluminum at home. I own a power supply so I thought I would try my hand. Well the first thing is I wanted to see if normal dyes work, and they do, sort of. Didnt want to spent 20 bucks on real anodizing dye. Cool part is that the anodizing process worked and really isnt all that difficult. Two of the wheels came out nice and even and the other two have some darker staining. The dark brown dye I used thinned down to an odd olive/gold but I am actually ok with that.

Other cool part is the finish. I tumbled the wheels in a vibrating tumbler for about three hours (I started with a pretty good machined finish) and using some abrasive plastic pyramids the parts took on a great matte finish that is really smooth. We will see if the abrasives opened up the axle holes a few mils.

And for those who are thinking of trying anodizing at home. There is a lot of good info on the web so I wont rehash that here but its a pretty rewarding thing to do. Also, even less than optimal anodizing is harder and more scratch resistant than most paints out there so its a great way to go. It does involve battery acid so make sure you know what you are doing there.

I hope to have the axles done this week and the side plates next. Will be getting pretty close to a working prototype soon I hope!

For anyone that has suffered through a schematic

This guys comics crack me up but this almost made me do a spit take.

http://xkcd.com/

A breather and some progress

Well now that life has settled down a bit I was able to model the basic platform for the robot. I chose some 100:1 gear motors that I will detail later. Basic dimensions for it are 8" by 5" (chassis) and the wheels add about 3" to the width. I will be adding a platform on the top, at least, or enclosing the whole thing for battery storage and such. I am thinking about venturing into carbon fiber molding for that part. I also hope to have the wheels done by the end of the week and maybe anodized at that.

It will obviously be a differential drive with a motor at each wheel. I hope to eventually add roller bearings to the axle so the motor shaft does not have to handle the entire load. My guess is that most of these cheap gear motors fail due to lack of shaft support.

As for the electronics I will be using an Arduino Uno or Mega depending on the program size and storage requirements.

Duh

So I do get to work on my machine skills - cut one of the wheels too deep. Bonus is finished wheel is now down to 2.66 ounces. Lighter is better right. Never knew dumb could actually work out.

Am trying to figure out the suspension right now. Probably going to use RC off road shock absorbers and some sort of unequal length heim jointed rods. Sounds good in my head. Also really trying to narrow down my motor choices. Probably brushed gear heads for now. Guess I gotta finish the platform, calculate weight and figure out what size I need.

Remember, as I have learned. With motors, its the stall current that matters since a locked up motor could fry a bunch. Some small motors that draw less than 1/2 amp can easily exceed 5 amps at stall. And also remember that a motor that isnt moving must quickly go through stall therefore there is a momentary current draw spike at the stall current. Four motors like that would be 20 amps total!

3D Modeling

So for those of you who have used or have access to a 3D modeling suite like Autodesk Inventor, you are probably aware of the capabilities of this software. For those of you who dont, here is a little example. The wheel I created out of 6061 aluminum weighed 2.71 ounces. Inventor says it should weigh 2.704 ounces. Really amazing. Guess I need to work on my machining skills!

Let there be wheels

So any project presents the dilemma, buy or make... Well I look at making as kinda the fun part and really labor costs are the worst in any project and relative to the rest of the world, my labor is cheap.

So now there is a real prototype wheel for the as yet unnamed robot. Several hours slaving over a benchtop lathe and voila. The specifics so far are as such. Extruded billet started around 9 ounces. The mid point was down to 7.5 and the final form, speed holes and all, was a scant 2.71 ounces. Could always be lighter but this is pretty good. With the Traxxas tire the unit weight is 3.96 ounces which will make the full set right at a pound.

The next model might have something better but this will work for now. Keep it light keep it simple right?

Wthere the rubber meets the road

So for anyone aspiring to build a robot the question of tracked vs wheeled comes up. Well for my purposes the wheeled idea won out. I hope to have independent drive motors, and control, for each wheel eliminating the need for transmissions or differentials.

Now, that said, I thought I would start in in the R/C car world and man... There are a TON of tires and wheels available. I ended up hitting the local shop and with their help I found a Traxxas Anaconda 2.2 inch tire that will suit my needs, for now. The R/C tires use foam inserts for support, come in all varieties and sizes. There are also wheels ranging from cheap plastic to CNC bead lock varieties that will set you back a ton (100+ bucks!)

Being a real adventurer I am going to make my own aluminum wheels with the help of a cheap, benchtop lathe. More on that process as it evolves.

For now I will post the drawings of the 3D wheel model, both as it will be made and the 'cool' version that I dont have the machinery for. Both are intended to help anyone that wants to follow along on their own.

The Beginning

So I have decided to document publicly my attempt to build a robot. I hope it will be a good learning experience for me and for anyone that chooses to read this.

A little background. I am a CS student at the University of New Mexico and hope to use this as a platform for autonomous control and machine development.

The specs, as I have them in my head, so far are.

Controller - Arduino Uno (a great, low cost, open-source hardware processor)

Chassis - four wheeled with independent drive motors on each wheel. Platform will be a flat plate that will carry electronics and sensors.

Power - proof of concept will probably be corded power supply that will migrate to batteries and hopefully solar charging.

Wish list - I hope to add wireless programming, onboard SD card, GPS and some sort of display for diagnostics.

As yet the design is in the early stages of 3D modeling with Autodesk Inventor and as I develop more ideas I will post them here.