GPS Receiver Expansion Module for Handspring Visor™ PDA

Members

Paul Niemczyk, pniemczy@uiuc.edu, 365-0458

Tim Wilson, twilson@uiuc.edu

David Marshall, dmarshll@uiuc.edu

Goals

Handspring is a recent startup company, founded in late 1999 by the members of the original Palm Pilot team from 3Com. Handspring’s PDA, Visor, is revolutionary because it offers a plug and play hot swappable expansion module for quick hardware upgrades and miscellaneous add-ons. Handspring encourages entreprenuership and development of its expansion cards by OEM’s. The market is young, certainly not saturated, so there is definite entrepreneurial possibilities.

Currently, a company is planning to market a GPS receiver hardware module for the Handspring. My group would like to design one as well. The GPS would be a compact card that would plug into the Handspring, taking advantage of its software capabilities and onboard CPU, providing the user with the typical options available to GPS subscribers: position, heading, hyperaccurate time, and a simple map function that details paths already taken. Thus, this project has two goals:

Show the entrepreneurial possibilities of this project by quickly delivering a cheap and functional GPS receiver in the form of an expansion module;

Show our engineering skills by learning and utilizing a current industry followed computer interface to build a useful product. As we do not know about the software API, the specifications of the Handspring modules, or about how the GPS receiver works, integrating these three will prove to be a challenging project.

Block Diagram of Design

Because we have not yet determined which GPS receiver we are using, we cannot give a detailed block diagram at this time. The device consists of three interconnected units: the Visor software (further broken into application software, which interacts with the user, and device software to control the hardware interface), the module hardware that connects the GPS to the Visor and provides the Visor with information about the status of the GPS receiver, and finally the GPS receiver itself (we plan to use a receiver board, not a black box already-build consumer GPS unit).

Visor

Application software

Module Hardware

GPS Receiver

CPU

ROM

Serial – Parallel Conv.

Device driver software

& Hardware Module

Serial Data Out

Module Logic

Currently, the specifications for the Visor are known. General information about Visor:

133 MHz 32-bit Motorola Dragonball CPU. Expansion interface is extension of CPU address/data bus with significant bit inoperational; that is, expansion bus is 16 bit address, 16 bit data. 3.3 V power supply on-board, capable of driving module hardware. Guaranteed maximum 100mA source. Docking voltage is maximum 5 V (buffered inputs not necessary; can use TTL). Sample timing diagrams from developer manual are in lab book.

Since we do not know which GPS receiver we are using, all that we know about the GPS is:

Output 9600 baud RS-232 serial, capable of decreased speeds for synch with external hardware. Onboard programmable MCU. Typically driven by 5 V source. Printed microstrip GPS antenna with optional input from external GPS RHCP (right hand circularly polarized) antenna.

Resulting Performance

Because this is intended to be a consumer device, performance will be based on accuracy, dependability, speed of operation/ease of use, and battery use.

Dependability should not be a concern provided that the unit adheres strictly to the guidelines set by Handspring for their expansion modules.

Accuracy is unfortunately a parameter that will be difficult to determine. To make sure the GPS works as it should, we will test our GPS along side an existing black-box consumer GPS. Resulting headings should be approximately 5% within each other, the given times should be within 0.01% of each other (as each of the 26 GPS satellites keeps exactly the same time), and the given locations will be similar with a large and unpredictable error (due to the inaccuracies interspersed into the consumer GPS system to protect military interests). The government guarantees that consumer GPS systems (i.e., those not using military P/Y code) should have tolerances no greater than 22 meters of error. The bottom line is that government sets the errors and tolerances on the GPS systems, so there will be little to test other than functionality.