My Robot Lawnmower's Main page

Hi. My name is Eddie Leighton and I live in a historical small town called Lichtenburg in South Africa. I enjoy designing and building simple electronic projects , technical computer programming , hobby robotics and amateur radio (Call sign ZS6BNE since 1975). This is a snapshot of me taken in February 2001 in my home electronics workshop and "Ham shack".

The first desktop robot I ever built is shown below. This robot has an "Umbilical Cord" which is plugged into a printer port of any standard IBM or compatible PC.

It used to have a 386DX40 onboard PC and an infrared sensor mounted on a third stepper motor to allow it to "Look around" but the Power requirement to Weight ratio was just too high and the stepper motors battled to handle the load.

The coils of the two stepper motors were controlled via the printer port by software for the desired movements of the robot. This provided me with a very good basis for experimenting with robotics and stepper motors in general. Power was supplied to the coils of the stepper motors by a 6 volt Gel battery carried on the robot. There are also facilities for sending back five signals from the robot to the controlling PC printer port for sensor experimentation.

This is where it all started. A floor full of parts shown above. I managed to purchase two identical second hand Nissan Skyline wiper motors and a 12 volt blower motor from a motor scrap yard dealer in Potchefstroom not too far from my home in Lichtenburg , Northwest Province , South Africa. The wheels are standard 170 mm diameter lawnmower wheels obtained from the local hardware store. Their traction is not all that good but they made a good start for the robot lawnmower project. The only component which I could not use for the robot lawnmower was the dustbin lid which was too small , but I used it on my first desktop robot and it gave the robot a nice "Turtle" effect. Although the MACH 2001 Robot Lawnmower (The name was derived by it's month and year of being built) does not look like a lawnmower that will eventually be it's goal.

Here is a snapshot of how I mounted the lawnmower wheels to the Nissan Skyline wiper motor shafts by means of a threaded 12 mm bolt. This was done with the help of a friend who could do the necessary turning for me on his back yard lathe. The chassis was made from an aluminum cover taken from an old piece of electronic equipment.

The various components evolved over time by trial and error , doing sketches here and there and somehow they all fitted together as though they were designed as such!

I used an old 386DX40 small sized computer for the "brains" of the robot lawnmower. It's maybe a bit of an overkill but it was simply available and I enjoy controlling the outside world via the printer port. It is a cheap and easy method for computer control. As can be seen above , I cut the motherboard mounting area from an old desktop computer case for use as supporting hardware for the video / printer port card , the floppy disk controller card and also the 386 motherboard.

The 386 PC Controller has now been removed and will be replaced by a PIC16F84 Micro controller which solves the problem of having to boot from the floppy drive and also solves the power supply problem. Programming the PIC is not as straightforward as programming the 386 PC's printer port though as it has to be done using PIC assembler. (I use the MPLAB assembler from Microchip)

Here is an example program to send a signal to one of the robot's motors via the motor relay driver board. (5 Seconds ON / 5 Seconds OFF) IOControl.ASM

I used second hand YUASA Gel batteries which were replaced by new ones in computer UPS systems. Two 6 volt batteries are wired in parallel to supply sufficient power to the 386 PC via a silicon diode bridge to drop the voltage down to approximately 5 volts required by the PC. No fancy voltage regulation at this stage but it must get the highest priority for the future. The third battery is a 12 volt battery used for the wiper motors , cutter motor and automotive switching relays shown below. I may even add a second 12 volt battery to power the cutter motor alone seeing this motor can draw currents in excess of 5 amps and over 20 amps on startup!

A 7805 voltage regulator is used to drop the 12 volt supply down to 5 volts to provide power to the PIC Micro controller. An important point is the use of 33 microfarad tantalum capacitors on the input and output of the regulator and a 100 nanofarad capacitor across the supply terminals of the PIC IC to prevent noise affecting the PIC Micro controller due to the switching of the wiper motors.

The whole setup is not as efficient as one would like it to be , but this was to get something going with no unnecessary problems to get it done. No electronic H Bridges and speed control circuitry although I have plans to use an output of the printer port for a software controlled pulse width modulated signal for possible speed control when applying power to the wiper motors and possibly even the cutter motor too. The cutter itself will probably be in the form of a bush cutter's blade or edge trimmer. This still has to be experimented with.

All I/O control is done via the printer port (LPT1) of the onboard 386 PC. The outputs from the infrared detector kits are fed from the open collector transistor output to the input interface module. Further inputs will be read from various other sensors via the input interface module when the budget allows it.

The input interface module is basically just a biasing network for the input lines on the printer port. The lines are biased to +5 volts via 1 to 10K resistors and can also provide a supply to the open collector sensor output circuitry.

When the PC is not running , I have a computer command simulator to switch the motors on , off , forward and reverse to give desired robot movements. This is simply a box of switches interfaced to the output control circuitry.

Originally , the software was written using Visual Basic for DOS which I found on the internet and although it is slow , it allows me to do event driven programming which is a little better than using languages like PASCAL or normal BASIC. I just have to plug in a standard IBM PC Keyboard and Mono screen. I use a removable 1.44 Mb floppy drive to boot the PC with DOS and load the program into memory. Thereafter I remove the floppy drive and the program continues normally running in the 2 Mb of onboard RAM.

See above , concerning the new PIC16F84 Micro controller. A program I wrote in PIC16F84 assembler to simulate all the Mach 2001 Model II robot movements can be viewed here.

I have built a wooden stand as can be seen above under the robot to prevent traction of the wheels while simulating behaviors in response to sensor input (Stimulus). Almost sounds like psychology but is it at all anything else? The infrared detector units are not shown in this picture , but have they been mounted on both sides in the front , with adjustable brackets to be able to get the correct angle to be effective in detecting objects from the front and sides of the robot.

I thought it would be of interest to mention what my robot lawnmower cost to make it a reality and it is by no means complete.

Chassis (Cover from an old electronics unit) Free

Second hand Gel batteries Free

Angle iron for mounting brackets Free

386DX40 Motherboard + 2Mb RAM (No longer used) Free

Video / LPT card and IDE controller card Free

PIC16F84 and programmer unit Donated

DC Trip switches for the 12 an 5 volt supplies Free

Additional hardware (Bolts and nuts) R 30-00

1.44 Mb Floppy drive (No longer used) Reusable

Two second hand wiper motors R 300-00

One second hand fan motor R 100-00

Lawnmower wheels R 140-00

Castors (Had to buy 4 of) R 60-00

Automotive relays R 100-00

Electronic components and Vero board R 100-00

Two Infrared TX/RX kits R 111-00

TOTAL R 941-00

To compare internationally , South Africa's Rand is equivalent to approximately R 8-30 to the dollar
and to date (25 October 2001) R 9-30.

I would love to share ideas with other robotics enthusiasts. I know my robot is very primitive and still needs a lot of attention. The important thing to do in the near future is to replace the two balancing castors with a large third castor to assist with movement over grass. The small ones tend to get in the way.