A 70cm BeeLine GPS Transmit Amplifier

Project Status

V1.2 Flight Tested on S08L01
V2.0 Not Flight Tested

Problem Statement

The BeeLine GPS Transmitters are 440Mhz APRS telemetry units. The power output is only 15mW, its not surprising to see the signal fade away to nothing around 10,000ft + some horizontal distance. The units cost $350 apiece and we have two of them so reclaiming use to 100,000ft would prove economical as well as convenient.

Solution Proposal [ppt]

The BRBs are transmit only units. This means that a PA can be used without the worry of a T/R switch or other bypass networks. We propose the use of a GSM amplifier chip to boost the power levels to a more convenient level.

The RF5110G GSM Amplifier IC:

1. 32dB gain at 450Mhz
2. 36dBm (4W) output max
3. Output can be reduced down to -10dBm or turned off completely to save power
4. Costs $6 per chip in single unit quantities
5. Uses standard lithium-polymer/ion cell voltages (3.6V nominal, up to 4.1V)
6. Draws 2A max during TX at 36dBm
7. Draws 200mA Iq without idle power shutdown

The circuit:

1. Has a total cost of $65.44 including PCB (without shipping). Many of the passives require 10 qty minimum so incremental costs are much smaller $34.50
2. Allows the output to be turned off by applying a high level signal to the *EN pin
3. Allows the calibration of the analog power control pin through a trimpot and a jumper (this cannot go above 2.8V)
4. Has a voltage-follower to buffer any input voltage to the analog power control pins to remove any current dependencies
5. Has a 3.0V regulator which is only used for the input to the voltage follower (So the output power doesn't change as the battery drains)
6. We have custom firmware compiled for an amplifier enable signal
7. Board size is 1.0x1.15"
8. Weighs 6.6 grams + 41.6 grams for a 2AH lipoly cell
9. Does not have an onboard regulator, apply ONLY a single lipoly cell (large enough to supply the 2A required for TX)

Work completed thus far

1. Two Rev 1.2 boards ordered and one assembled
2. Two Rev 2.0 boards ordered and one assembled

• REV 1.2

1. Schematic Diagram
2. PCB preview!
3. PCB board layout (with fill)
4. PCB board layout (without fill for visibility)
5. Excel parts list with cost for a single board

• REV 2

1. Eagle layout
2. PCB preview!
3. Photograph of the bare PCB

Test Results

• REV 1.2 - tested in the lab and field

1. Outputs 1 Watt into a 50ohm magnetic base dual band antenna (center loaded quarter wave)
2. Packets are decodable at 1 Watt
3. During S08L01 the packets came in from over 36 miles away and were decodable using the 70cm j-pole antenna antenna!

• REV 2.0 - tested in the lab

1. Outputs 4 Watts into a 50ohm magnetic base dual band antenna (center loaded quarter wave)
2. Packets are decodable at 4 Watts

Todo list

1. Assemble the V2 board

Bugs

• V1.2

1. R1 should go to ground, not be in series! (fixed in REV2)
2. Striplines should be real calculated striplines (fixed in REV2)

• V2.0

1. For some reason the silkscreen labels didn't print…
2. More decoupling was needed (68uF tant x2, 100nF tant x1)
3. Output power isn't stable — this is either a real bug caused by biasing voltage changes, or the power meter is stupid and we need a new digital one.

Project Contact

Jasper Nance