RoverSat
Prepared By:
J. Perkins and M. Albertson
University of Colorado at Colorado Springs
Mechanical and Aerospace Engineering, UH3
1420 AustinBluffs Pkwy
P.O. Box 7150
Colorado Springs, CO80933
In Association With:
Anna Stanko
University of Colorado at Colorado Springs
Mechanical and Aerospace Engineering, UH3
1420 AustinBluffs Pkwy
P.O. Box 7150
Colorado Springs, CO80933
March28, 2005
OUTLINE
- Introduction
- Payload Box
- Box
- Internal Circuitry
- Heater
- Opening Mechanism
- Payload Testing
- Rover
- Materials
- Construction
- Testing
- Launch Day
- Conclusion
INTRODUCTION
The goal of the RoverSat project is to design and construct a small weather balloon satellite capable of driving autonomously and imaging a landing site. The other requirement is that the payload could not weigh more than 1.8 kilograms. To meet these requirements within the given timeframe, our design group was split into two teams; one team focused their efforts on the design and testing of the payload box, while the other concentrated on the design and construction of the rover. The goal of the payload box design was to be able to sustain the rover through in-flight conditions and deploy the rover upon landing. The payload box design was viewed as a critical priority based on the failure of past RoverSat payload boxes. The goal of the rover was to deploy autonomously and collect digital images.
PAYLOAD BOX
To accomplish the goal of sustaining and deploying the rover during and after flight the payload box was divided into four subsystems. The box, internal circuitry, heater and opening mechanism were the necessary components of the payload box, each was designed and implemented separately.
BOX
The design of the box was the topmost priority of the RoverSat project. The box was designed to be structurally sound and capable of withstanding large temperature differentials as well as the impact of landing. The first aspect of design considered was how the box would physically deploy the rover after landing. It was necessary to determine the simplest way for the rover to deploy from the box. The proposed solution was to design the box based on a four sided pyramid with the top of the pyramid removed to form a flat top parallel to the base.Foam core was used for the box construction because of its lightweight, thermal and structural properties. The rover would then be mounted to the base on the inside of the box and could easily deploy after the sides had opened. The next area of concern with the pyramid design was how the box would land. In order for the rover to deploy properly the box would have to land on the base. This presented a problem because since previous DemoSat projects demonstrated that the payloads have a tendency to rotate and drag along the ground upon landing. The exact side the payload box would land on was very difficult to determine. To solve the problem two areas were examined. The first solution was a way for the box to right itself upon landing. The pyramid box design had five possible sides that it could land on. The base was the ideal landing position but the sides were also realistic possibilities. The proposed solution to having the box land on one of the sides was to mount extension springs to the box externally. Two springs were connected from each side to the base of the box. When the box was in the closed position the springs would be in tension. If the box were to land on a side the springs would act to “flip” the box to the base side down position and the rover could deploy easily. The other area of concern in landing position was where to connect the box to the tether. The first idea was to connect a tube to one of the joints joining two sides. The difficulty of this design is the uncertainty that the connection would be able to withstand the forces exerted on it by the rotation of the payload during flight. The next idea was to mount the box to the tether via the base. Two screw eyes were inserted into the box on one of the diagonals. The tether cord could then be easily threaded through the two screw eyes and a tube would not need to be used. The concern again with mounting the tether in this fashion was strength. The proposed solution was to mount aluminum plates to the inside of the box and have the screw eyes anchor into the plates. That would distribute the force more evenly across the base instead of having all the force directed at one point in the foam core. In addition to the aluminum plates it was decided that the best way to alleviate some of the stress on the screw eye joints was to mount as many of the internal box components to the base as possible. The idea was to keep the weight as close to the tether cord as possible to reduce the mass moment of inertia of the payload.
The next area of the box design and construction was to make an inside frame for the box. The inside frame was necessary to hold the rover in place during flight and secure all the internal components. The frame needed to be lightweight and strong enough to withstand in flight rotation and landing. The shape of the frame was essentially the same as that of the box except a bit smaller. A half inch gap was placed between the frame and the box on all sides. This was to allow for the servo arms to rotate and springs to operate. There was also a 1/2 inch gap between the top of the frame and the box. The frame was constructed out of 5/16 inch aluminum tubing, and the sides of the frame were constructed out of electronic circuit board. The circuit board was chosen because it was lightweight and easy to assemble. The base of the frame was constructed out of aluminum sheet and positioned one inch above the base of the box. The last area of concern was keeping the rover secure during flight. A gate was fastened to the side of the frame wherethe rover would deploy from. The gate also had an extension spring mounted similar to the springs on the outside of the box and was kept closed by a servo arm during flight.
INTERNAL CIRCUITRY
The operations of the payload box were performed by a Basic Stamp 1. The stamp required a 9V power supply and had eight I/O pins to connect devices to. Opening the box was the only operation the stamp was used for. Connected to the stamp were the servo controller and the servos. The servo controller used was the Mini SSC II Serial Servo Controller from Scott Edwards Electronics, Inc. The servos used were HS-55 Feather servos from Hitec. The servo controller required a 9V power supply and the servos required a 6V power supply. Initially the stamp was required to read and store voltages coming in from a pressure transducer but due to time restrictions and the limited electronic knowledge of the team, the pressure idea was discarded. The next idea for the stamp was to cause the box to open based on a timer. When the timer reached a certain value, the stamp would then activate the servo controller which would in turn activate the servos.
HEATER
In preparation to develop a feasible rover design, past attempts at rovers, and other experiments were investigated. Upon analyzing the past launch data, regulating the internal temperature of the launch capsule was deemed a top priority due to the causal effects upon electrical equipment.
In designing a heating apparatus for the capsule, several options were examined while specifically noting their cost, size, heat output, and required power. After weighing several different devices, only two devices successfully met the requirements. The first, a daisy-chain linkage of ceramic resistors; designed to dissipate excess heat when subjected to excess current. The second, a heating coil constructed from nichrome wire, designed to rapidly heat the ambient air in the capsule. Due to weight limitations and requirements of higher power consumption by the ceramic resistor design, the heating coil design was chosen.
The design of the heating coil aimed at providing intermittent quick bursts of heat to the internal chamber of the capsule. Intermittent usage of the heating mechanism was intended to conserve battery life, and prevent any damages that may result due to prolonged close range exposure to the device. To achieve this objective, a wall mount thermostat was included in the circuit. This provided a constant monitoring of the internal temperature, as well as acted as a breaker switch when the target temperature was reached and usage of the heater was no longer required.
OPENING MECHANISM
In designing an autonomous capsule for the rover, a reliable mechanism was required to open the casing upon landing. Due to the design of the containment unit, a mechanism was required to secure the four sides of the box together; releasing them simultaneously would upright the box and allow the rover to exit the capsule. To accomplish the requirements set during the design phase, addressing the time frame of release was dealt with first. A timing loop was created on a Basic Stamp 1, a processing chip, which upon completion would activate a servo that initialized the release of the tethered walls.
To open the launch capsule, the opening mechanism essentially acted as a spring-loaded lock. The four individual capsule walls were aligned such that each would overlap on the top of the container. Through this overlap, a pin was placed from inside the casing, protruding outward.This provided equal tension in all four walls, keeping the box from opening. This pin was connected to a circular disk inside the capsule. Connecting from the inner cage of the capsule to this circular “release disk” was a spring. The spring was kept in an elongated state, prevented from contracting by the servo arm, wedged beneath the release disk.
When the timing circuit initialized the opening device, the designated servo rotated 90º, allowing the spring to contract downward towards the inner cage of the capsule. When the spring contracted, it lowered the tethering pin. By lowering this pin inside of the casing, the walls of the box were no longer tethered and opened outward like a blooming flower. This simultaneous opening of the four individual casing walls would in turn upright the capsule (if necessary) allowing the rover to dismount correctly every time.
PAYLOAD TESTING
Each component of the payload was tested separately. The box was first tested and was simply subjected to a drop test. The box was dropped from the top of a three story building with a simulated payload inside. The frame was not installed in the box and a small spool of copper wire was attached to the inside of the box. The wire spool weighed approximately 800 grams and was similar to the projected weight of the rover. Upon completion of the test, the box was examined for failures. Overall the box held up to the drop but was cracked in a few places on the base. The cracking was attributed to the fact that the inside frame of the box was not yet in place. The frame would provide the stability and strength the box lacked. The test was repeated with the interior frame and the results were better; the box was not noticeably cracked and the payload was intact. The only blemishes were minor indentations expected in the drop.
The internal circuitry was initially tested outside of the box. A program in Basic was written that would activate the servos after the timing loop had run through. The times were initially set at just a few minutes to test the servo operations. The times were then extended to ensure that the servos would indeed activate after being dormant for a period of hours. The anticipated flight time was estimated to approach three hours so a timer was set accordingly. The servos worked as anticipated and were activated after the three hour time was through.
The heater was also tested outside of the box. The main issue in testing the heater was the required voltage to produce an effective amount of heat. The hope was that the less voltage the heater needed the fewer batteries would need to be used and the weight could be reduced. The heater was tested using one 9V battery all the way up to four 9V batteries. After testing the different arrangements of batteries, three 9V batteries were deemed adequate to power the heater.
The opening mechanism was a crucial component to test. If the latch were to fail, the project could be lost. The mechanism was setup as described previously and the servos were also setup via the stamp. The timer was set and tested in much the same way as the internal circuitry was tested. The initial tests went well and the latch was released as desired. The mechanism was then implemented in the box and tested. The next tests were to determine if the latch could hold up to the rotation experienced during flight. Two tests were performed in this manner and the mechanism held up each time.
ROVER
The task of the rover was to drive autonomously around and image the landing site. The rover had a programmable stamp that allowed it “think” on its own and employed a digital camera for imaging purposes.
MATERIALS
From a bulldozer rover kit, the gear box and motors assembly, a small portion of the wood frame, axels,wheels, and track were used. The assembly was hooked up to a dual H-bridge driver to control and power the motors. The dual H-bridge driver needed two 6-volt batteries to operate.
For the rover’s vision, bumper switches and an infrared proximity detector were purchased. However, only the bumper switches were used on the final rover. Each bumper switch kit had three prongs, a switch, sleeve tube and cap, and a 10 kΩ resistor.
To image the landing site autonomously, we used an HP 435 digital camera powered by two 1.5-volt batteries and a mini dual H-bridge driver. The mini dual H-bridge driver replaced a timing circuit and contained a connection strip, 8 PNP transistors, 8 NPN transistors, 4 1-kΩ resistors, and 8 diodes.
The bumper switches and mini dual H-bridge driver used the same 6-volt battery for power.
A P-Basic computer program written to the Basic Stamp II controlled all the devices together by way of input-output pins: four for the rover’s dual motor H-bridge driver, two for the two bumper switches, and two for the camera’s mini dual motor H-bridge driver.
The frame’s materials were PCB board, 3-32 lock-tite® nuts, 4-inch-long 3-32 bolts, and brackets.
CONSTRUCTION
The parts for the gear box came from a kit. Each gear had to be properly placed and greased. Each motor controlled one side of the split axel that was put on the front of the gear box. A track connected each front wheel with its corresponding back wheel to make the back wheels turn the same direction as the front wheels. The back wheels were connected by a solid axel through two holes in the rear of the gear box.
The dual H-bridge driver used to control the gear box and motors assembly required no construction.
For each bumper switch, the 10 kΩ resistor connected the 6-volt battery (positive side) to the input-output Basic Stamp II pin. The input-output pin then connected to the middle prong of one bumper switch. The golden-colored (ground) prong connected to the negative side of the 6-volt battery.
After the camera case was removed, two wires were soldered to the part of the circuitry corresponding to the button pressed to take pictures. Also, the two power wires were soldered together to create a short so that the camera would always be in the “on” position. However, the two power wires kept falling off creating problems.
The components of the mini dual H-bridge driver needed to be soldered into the correct position. The connection strip was soldered at the top of the board. One end of each resistor was soldered to the middle pins of each side. The other end of the resistors were soldered to the top four transistors (each PNP on the outside and two NPN transistors in the middle). Four more PNP and NPN transistors in the same order connected the top transistor row to the four diodes of motor A. Motor B was soldered in the same way as motor A.