Nu Blue v .002 By Michael Owings Last Updated Nov. 18, 1999

Introduction

IT LIVES! I haven’t had much time to put this report together, but basically, she’s a-working. A more detailed report, along with more photos and schematics/PCB layouts is planned in the near future. Currently, the robot is little more than a remote controlled wheeled platform -- but hopefully this will change as soon as time permits. For photos skip down to photos.

Mechanical: While MUCH more is planned for this bot, currently it consists of three decks. A bottom deck reinforced with an angle iron frame holds the batteries and motors. A middle rectangular deck holds the electronics, and another round upper deck at the top can hold a notebook computer and camera. It can also hold beers, chafing dishes, small children, etc. The robot is a "dual differential steering" -type, driven by two independently controlled gear motors and 6-inch wheels. The small mid-deck with the electronics will be replaced with a round one like the others at some point.. Multiple mid-decks are possible.

The stucture of the robot is pretty much completely made of stuff you can get from any Home Depot, with the exception of the 1/16" aluminum sheet. This was given to me by a friend, but it can be had fairly cheaply from McMaster Carr. The sheet was cut with a standard jigsaw.

Sensors: None at the moment. Multiple Sharp IR distance/obstacle detectors will be installed in the near future, along with some kind of physical obstacle detection (bumpers or whiskers). This robot is designed to help me do some machine vision research and experimentation (thus the rather large size) and will ultimately have at least one camera, ultrasonics and possibly microphones.

Drive: Nu Blue is driven by two Zagros Robotics motor/gearbox/wheel assemblies. I have been pretty happy with these. I replaced the old motor/wheel units which were from H&R Enterprises. The H&R units had a lot of shortcomings for my application, including speed vs. torque, apparent plastic gear box assembly and noise. The motors are controlled via a home-built 2-channel optically isolated FET/Relay controller. The relay is used to change the motor polarity to reverse a wheel, and the FET provides a PWM signal to the motor output. Seems to work quite nicely so far – these motors can draw up to 8.2 amps stalled (at 12 volts). I will post schematics and PCB layout for the controller in the near future. The controller itself should be able to handle at least 15 amps per channel with a little care.

Logic: Not much yet. A 10MHZ PIC handles speed and direction control. Commands are input to the PIC serially at 9600 baud. The main controller is currently a basic Stamp II. This part is currently responsible for decoding commands sent to it via an IR TV remote controller, and passing them to the PIC. Currently, using the remote control, the robot can be told to STOP, go FORWARD, go BACKWARD, rotate LEFT, rotate RIGHT, go FASTER or go SLOWER.

Eventually a PIC will be assigned the task of dealing with the remote commands. Currently all control logic is breadboarded – when I’m happy with it, it will be transferred to a PCB.

Power: The motors run off a single 12V 6.5ah Lead acid cell mounted on the bottom. I will undoubtedly add another of these in the near future. The logic runs off of a separate power pack (just 4 AA cells at the moment) – this will also be replaced by a separate logic power subsystem.

Photo Gallery

I have a few photos available. Click on the thumbnails below for a larger shot.

	Nu Blue side view, The remote I use to control it is shown. Just a standard $10 universal remote. Note jaunty angle of robot. While this looks good, it's actually bad design and needs correcting. If anything, the robot should be angled down -- otherwise, the drive wheels lose traction when the robot is traversing even small ridges or bumps.
	Front view. The robot is supported on a simple structure made from casters and preformed metal parts I found at Home Depot The decks are supported on risers made out of 1/2" threaded rod. The decks are joined to the rod by nuts and large "fender washers". The arrangement is fairly stable, but the threaded rod is somewhat heavy. McMaster-Carr sells aluminum threaded rod which should be MUCH lighter and perhaps a better choice.
	Close up of the middle deck, which supports the motor controller and associated breadboarded logic circuitry. The Basic Stamp II responsible for decoding the IR commands is also shown. This is due to be replaced shortly by a PIC, and a new NetMedia BasicX-24 will serve as the heart of the system.
	Bottom view. A 12-volt drive battery is suspended from a "cage" built out of bent and drilled metal straps. You can also see the angle iron frame. Note that the wheels protrude past the edge of the deck. This will be fixed in a later revision. Ideally, the wheels should be inside the circumfrence of the deck.

Back to SwampGas