Pictured left: The RS-550 motor

The single passenger F-150 that we’re using contains two RS-550 motors which, according to the datasheet, can draw 70A of current when stalled. Typical performance is around 9A, but is still much higher than I expected. I should note that I’m not quite sure precisely what I expected, but I’m a little concerned about creating a motor controller for something that drives so much current. I may take a serious look into purchasing a pre-fabricated motor driver.

It is also noteworthy that that RS-550 motor is actually capable of running on voltages as high as 14.4V constant. And while I have no immediate plans to upgrade the battery, it’s nice to know that we can unlock some serious potential for speed. I’m already fantasizing about hosting my own little autonomous NASCAR circuit on a quarter mile running track. But one thing at a time.

I had my brush with compromise the other day. It seems that while I presumed that our autonomous vehicle would run solely on the power of microcontrollers, my partner was under the impression that we would be using a PC as as the brains of the operation. And to make matters worse, he insisted that the PC run Windows.

I did my best to argue against the Windows PC, but in the end my friend proved to be doggedly relentless. Over time, I’ve actually grown to appreciate the unexpected change however; the beauty of compromise. Obviously a PC will provide us with a much more powerful platform on which to build.

In the end, we agreed to use a Dual Core 1.8GHz Dell laptop that has been lying around my apartment unused for over a year. This is proving to be an excellent experience on working in a group setting where the goals of the individual members of the group vary widely.

The article in the April 2008 issue of Servo Magazine titled “Turn a Kid’s Ride-on Car into a GPS Guided Robot” changed my mind. I had never considered using such a simple platform to build a (relatively) low cost autonomous vehicle. My interest was peaked.

As luck would have it, a close friend and colleague expressed interest in a similar project. We decided to undertake the project together. And as it turns out, I didn’t need more time to get ready. What I needed someone to share my interest with. Although my friend certainly brings a great deal of technical knowledge to the table, what I am most excited about is the opportunity to share in the experience with someone who can appreciate my enthusiasm.

I will be documenting our little experiment on this blog. It is my greatest hope that someone will read these entries when building their own autonomous vehicle and perhaps avoid some of the (many) mistakes that I am certain we will make along the way. If even one person can be spared the agony of watching their hard earned time and money go up in smoke, even at my own expense, then the time and effort spent writing this journal will have been well worth it.

Enjoy!

For BongoPong there isn’t really much to say.

The Velleman Pong kit was (one of?) the first kit(s) that I ever soldered together. It was a fun little kit to build, but I wanted more from it. I knew that I would someday find a fun or interesting mod for it.

A couple of weeks ago I came across an article on HackADay where a clever young hacker controlled his pong kit by flexing his own muscles. That hack got me thinking about new ways to interface with the oldest video game.

Being the worlds first video game, pong has been hacked to death. There are laser pong hacks, pong that plays against itself, and even pong on the side of buildings.

Clearly, finding something new and interesting in this saturated field would not be an easy task.

The ideas floated around my head for about a week. All of them were interesting, but each suffered the same flaw. They’d been done to death. Not necessarily with pong, but with video games in general. I needed to put a fresh twist on the old game. More importantly, it needed to be simple. My meager background in electronics certainly wouldn’t enable to run right out and build a mind controlled pong game.

The idea to use an Indo (Bongo) board for pong came out of the clear blue sky. Really. I have no idea what prompted me to think about it, I was driving when I thought of it. Within the hour I was at the hardware store, trying to find parts to make a tilt activated switch.

Things just kind of came together from there. I found the metal cylinders (see previous post), some PVC pipe, and a three foot piece of wood that looked sturdy enough to hold me.

The whole thing was functional in another hour or so. I was pretty terrible on the board that first night, and i noticed that whenever I would make a quick motion and the board would move very quickly to the left or right, the tilt switches would not react properly.

A quick google search taught me that this is a common problem, even for commercial mercury-based tilt switches. There is simply no way for the switch to differentiate between gravity and g-force. It makes sense when you stop and think about it. Luckily the problem became less apparent as I got better at using the board and the movements became more fluid.

The major flaw in my design was that there was no way to serve the ball without the controller. Using the Velleman pong kit, you serve the ball by pressing both buttons simultaneously. To fix this, I grabbed a cheap skateboard. Skateboards have the nose and tail curved upward slightly at the ends. I swapped the switches on the left and right side of the board, and flipped them so the trigger faced inward, and slightly down. This resulted in the same functionality as before, but now when the board is roughly parallel with the ground, both switches are activated and the ball is served without any additional user input.

As always, there are a million things I could do to make the project better. Imagine using an accelerometer to get analog values and then interfacing the board with a console or PC to play games like SSX, or even breakout.

It’s interesting to point out that I lost my entire weekend playing this game. At no point did I stop because I was tired and/or bored. Each gaming session ended when something on the board broke. On the first night one of the wires snapped off of a tilt switch. On the second night the PVC pipe cracked. I’d like to get a real Indo board, but the prices start at around $80 and work their way up. That just seems too expensive for a piece of wood and a rubber coated roller.

Be sure to check out the BongoPong project site for videos of BongoPong in action.