Maker Pro
Maker Pro

NMEA GPS Signal Generator

What aspect of the receiver do you want to test? I suppose one could

go so far as to deploy an array of controllable pseudolites but that

would be overkill if you're just checking the front end, and way

overkill if you only need a periodic NMEA source.

We're are building a GPS into one of our instruments. It may be a module ora PCB based one with an antenna. We'll we need to test this thing in manufacturing and we would like to test the performance of the antenna location in the instrument. GPS doesn't work in the building unless we're near a window and varies with what satellites are in the sky. So it is hard to compare the performance of different devices and even develop a simple pass fail test in manufacturing.

What I would like is a device that sends a fake NMEA string with a range ofabout a couple feet. I could then create a manufacturing test stand that puts the DUT at a certain position and passes if it receives the string and fails if it doesn't. I could also test different antennas and antenna positions in the device by moving the device closer or farther away from the transmitter and find the minimum distance for communication.

National Instruments and Agilent have test setups but they seem awfully expensive. You can get other RF dongles to do Wi-Fi and Bluetooth. I'd think transmitting GPS would be simpler than those. I'm surprised there isn't somesimple tester available with a satellite code that indicates it is a fake and only for testing.
 
A

Adrian Jansen

Jan 1, 1970
0
We're are building a GPS into one of our instruments. It may be a module or a PCB based one with an antenna. We'll we need to test this thing in manufacturing and we would like to test the performance of the antenna location in the instrument. GPS doesn't work in the building unless we're near a window and varies with what satellites are in the sky. So it is hard to compare the performance of different devices and even develop a simple pass fail test in manufacturing.

What I would like is a device that sends a fake NMEA string with a range of about a couple feet. I could then create a manufacturing test stand that puts the DUT at a certain position and passes if it receives the string and fails if it doesn't. I could also test different antennas and antenna positions in the device by moving the device closer or farther away from the transmitter and find the minimum distance for communication.

National Instruments and Agilent have test setups but they seem awfully expensive. You can get other RF dongles to do Wi-Fi and Bluetooth. I'd think transmitting GPS would be simpler than those. I'm surprised there isn't some simple tester available with a satellite code that indicates it is a fake and only for testing.

You are under several misapprehensions. The satellites do not send NMEA
strings to be picked up by a receiver. The receiver decodes a very
complex set of signals from several sats, and computes the local
position from that, then outputs the NMEA string ( usually serially )
based on that computed data.

Its very easy to generate NMEA strings, even a very simple script
running on a PC terminal program will do that. But that does not help
you test a GPS reveiver front-end.

For that you need a pseudolite generator, ie a 'thing' which generates
the satellite signals. Available, but very expensive, as you found.

Why not use one of the many active antennas, set up as a repeater, so
you mount that on the outside of your building, with a clear view of the
sky, and hence a reliable source of the existing sat signals. Feed the
signal inside, to your GPS receiver, or even just another antenna acting
as a local transmitter to your DUT ?
 
You are under several misapprehensions. The satellites do not send NMEA

strings to be picked up by a receiver. The receiver decodes a very

complex set of signals from several sats, and computes the local

position from that, then outputs the NMEA string ( usually serially )

based on that computed data.



Its very easy to generate NMEA strings, even a very simple script

running on a PC terminal program will do that. But that does not help

you test a GPS reveiver front-end.



For that you need a pseudolite generator, ie a 'thing' which generates

the satellite signals. Available, but very expensive, as you found.



Why not use one of the many active antennas, set up as a repeater, so

you mount that on the outside of your building, with a clear view of the

sky, and hence a reliable source of the existing sat signals. Feed the

signal inside, to your GPS receiver, or even just another antenna acting

as a local transmitter to your DUT ?





--

Regards,



Adrian Jansen adrianjansen at internode dot on dot net

Note reply address is invalid, convert address above to machine form.

That explains why I can't find what I'm looking for. I don't know what I'm talking about.
 
That's exactly what I setup at one company. Choke ring style GPS

antenna on the roof. 30dB amp inside the antenna. About 150ft of

LMR-240 coax. I used a power splitter and an AGC controlled amplifier

on the test line with a switchable RF attenuator to control the signal

level. L1+L2+WAIS patch antenna radiator under the test fixture to

minimize leakage to other test stations. I originally installed a GPS

bandpass filter to eliminate potential LightSquared interference but

that proved to be un-necessary and was removed. The building was

fairly well shielded as long as the doors and windows were closed. I

also installed a DGPS LF receiver and distribution system on the test

line, but since that dramatically increased the test time and few

users need DGPS, someone wisely decided that sample testing was good

enough.



The key part of the puzzle was the LMR-240 coax and associated

amplifiers. Any leakage from the coax or the amps would screw up the

test measurements. The switchable RF attenuator produced signal

levels down to about -145dBm, which could not be seen by any piece of

test equipment available. Typical operating levels at the receiver

input were about -135dBm. Watch your shielding carefully. However,

if you don't want to deal it, there are shielded boxes that will work.

<http://www.rfshieldbox.com>



The GPS location indicated by such a system is the location of the

rooftop antenna, which was surveyed to an accuracy where continental

drift became part of the reference position calculation. With DGPS

active, the largest errors came from the phase errors of the rooftop

antenna measured in mm. You could sorta see earthquakes and

subsidence when big heavy trucks arrived at the loading dock.



Also, I recommend you get a GPSDO (GPS Disciplined Oscillator) to run

the test equipment clocks. Some early head scratching and weirdness

was eliminated by syncing everything to a common frequency (and

phase).



There was some mumbling about a NEMA-0183 simulator. This can be

easily done with a computah and commonly available GPS test programs.

<http://avangardo.com/software/gps-generator-pro.html>

<http://sourceforge.net/projects/gpsfeed/>

etc... Google for "NEMA test generator". Someone else scribbled a

test program based on gpsfeed+, which included simulations of radical

position excursions, reflection simulation, leap second function,

random garbage, loss of signal, and other oddities difficult to test

with a live GPS signal.



Good luck.







--

Jeff Liebermann

150 Felker St #D http://www.LearnByDestroying.com

Santa Cruz CA 95060 http://802.11junk.com

Skype: JeffLiebermann AE6KS 831-336-2558

Great Info. I'll definitely look into this.

Thanks
 
Top