Robot Wars Home Page Southampton University
Robot Wars 1999/2000

RC BANDIT
UoS Home Page
PAGE INDEX
MENTORN REGULATIONS
MOTION CONTROL
WEAPON CONTROL
TESTING
HOME
PAGE INDEX
SITE INDEX
THE COMPETITION
HOUSE ROBOTS
COMPETITOR ROBOTS
THE CHASSIS
DRIVE TRAIN
PRIME MOVER
PWM MOTOR CONTROL
THE ARM
ACTIVE WEAPON
RADIO CONTROL
SPONSORS
CONTACT US
SITE INDEX
RADIO COTROL
MENTORN REGULATIONS
Mentorn Regulations encourage the use of 40MHz radio communication systems, whilst the use of 27MHz and 458MHz bands are also permitted. Mentorn also stipulate the no AM sets shall be used to control heavy weight robots such as the team's design.

Mentorn also require that a number of "fail safe" features be included aboard the robot. They require that in a failure/error scenario the robot will come to a halt, with all weaponry powered down, within a few seconds. Broadly speaking some error scenarios are:

1.Transmitter failure in which the transmitter either transmits nothing or meaningless/corrupted data

2.Transmitter becoming detached from its operator

3.Receiver failure in which the receiver either receives meaningless/corrupted data

4.A robot sub-system failure


MOTION CONTROL
In controlling the robot's tank style tracks, three main control schemes are possible:

Firstly, a simple three command (either reverse, off and forwards) could be sent to each track motor. However human reaction time would mean that steering the robot would be practically impossible.

Secondly a progressive scheme could be used. In such a scheme, the user would operate a pair of 'increase speed' and 'reduce speed' buttons, which would change the speed of each robot track by some predefined step. Under this system, however, the robot would not in general come to a standstill if the controller were removed from its operator - i.e. failing to comply with fail-safe requirements.

Thirdly, a fully proportional scheme could be implemented, whereby the user would operate a pair of control joysticks which would transmit an absolute velocity command to each track. This fully proportional scheme was deemed the most appropriate because it allows the user commanding the robot to make rapid acceleration / deceleration manoeuvres in a competent manner. Many such manoeuvres are not possible with a progressive system as to get from one speed to another each intermediate speed must be stepped through.


WEAPON CONTROL
In terms of components that need to be controlled, there exists a pump that provides a pressure that the hydraulic fluid can transmit. This is driven by an integral electric motor. A main solenoid then controls a valve, which enables/disables flow from the pump. Four further solenoids/valves then direct fluid flow to make each of the two pistons either push, pull or hold their positions. The five solenoids mentioned are on/off devices, and so ultimately require digital signals to control them. Initially it was thought that the speed of the electric motor/pump could, in principle be controlled. However, on a closer inspection of the hydraulic system, it became apparent that the motor is an AC device fitted with a mechanical, and therefore slowly switching, DC-AC converter. It is therefore a difficult, if not impossible matter, to make such a converter operate reliably and without causing undue wear, with a relatively high-speed Pulse Width Modulation (PWM) control scheme. Furthermore, since the motor is a particularly high current device, it instigates the use of large switching transistors. Assuming that mosfets would be used, these would have a large gate capacitance and therefore lengthy switching times, which would lead to large switching losses.

Finally the DC-AC converter seems to require a fair quota of its 12V supply before it starts working properly and the motor begins to turn, meaning that only a small range of speeds could be obtained in any event. One further possibility, to enhance arm/spike control would be to attach position feedback potentiometers to the pistons, whereby the user's command would be an intended position for the arm/spike. This command would be compared with a voltage from the feedback potentiometer and corrective signals sent to the hydraulic unit. This plan, however, was abandoned, as control would lost if the feedback potentiometer were destroyed, which would be quite likely in a combat situation.


TESTING
Some preliminary testing has been conducted with the TX and RX units. The units appeared to operate admirably, all signals being correct, and a range of about 200m was achieved with both the transmitter and receiver using quarter wave whip aerials. In addition, the fail-safe time-out duration substantially confirmed its design, it lasting about one second.
RADIO CONTROL