|
APRS/Rino?
Using the Garmin GPS/FRS Radio in APRS
Bob Bruninga, WB4APR
UPDATE. Waypoints Downloaded Automatically
Steve Bragg, KA9MVA, has managed to download 'hidden' waypoints from the
RINO onto his PC without user intervention. This leaves the way open to
integrate the RINO into the APRS data stream. You can find out more at http://www.hamhud.net/rino-frs.html
We need help! The Garmin FRS/GPS radio would be a fantastic addition to
the tools available to APRS operators in support of public service events if
we only knew how to get the RINO data being transmitted over the air into APRS. Then we
could decode the positions of all such handhelds at an event and integrate
that data into the
APRS Tactical Map displays and other
HAM radio communications systems. We have written a
SPEC on how we would integrate
the position information heard from a RINO into the local APRS network for
the event.
APPROACHES: There are two approaches:
First is simply download RINO data as waypoints from a central RINO receiver
Second is to receive the RINO tones into a PC sound card and demodulate them.
RINO WAYPOINTS: This was the initial idea and what started all this project.
But almost immediately it was explained to us that a RINO contact does not
automatically go into the WAYPOINT list until it has been manually accepted by
the receiving RINO operator
(see original Email). THis obviates an automatic system and was abandoned.
(Though recently it has been suggested that our first source of that manual
process may have been wrong! We are awaiting a trusted agent to perform a test).
DECODING THE RINO DATA:
What we have so far, is a WAV file
that lets you hear what the 1/4 second data
burst sounds like. The first burst is just a kerchunk of the PTT button
to force the RINO to send a position. Then the data follows.
Note: this WAV file is not good enough for decoding. We have another file
later recorded off of a discriminator that is suitable for decoding.
Above is the captured waveform.
Click here for a captured waveform file that you might be
able to help us decode. This is a text file capture of over 2000 data points
from a A/D converter. The 1st column is the sample number, the 2nd column
is the sample data, and the 3rd column is the unused 2nd channel of the A/D converter.
Bring it into EXCEL and you can plot it or do whatever you want with it.
IDEALLY, what we want is someone to write a sound card decoder for
this data, so that we can then take the data and inject it into any of the
many APRS client programs to integrate the positions of the RINO's into the
same APRS tactical picture.
What We Know So Far:
The burst lasts 275 ms (168 bits +/-)
The burst has periodic peaks every 1.625 ms (600 per second)
In the TEN-BYTE name field of this RINO (above) were the two letters "AJ".
The long string of identical bits are the remaining 8 null NAME 6-bit bytes.
The RINO was indoors, so the GPS had no lock and we do not know what posit was transmitted.
The last LOCK position was 38 deg 59 min north and 76 deg 29 min W.
The total information conveyed in the burst is NAME, ICON, and LAT/LONG.
We assume these take 10, 1 and 18 or so bytes respectively...
The above graph is an extended plot of the first 9 obvious data "symbols" where it
is easy to see that a symbol length is 1.625 ms. All of these symbols appear to
be either right side up or upside down, only two types exist. Byon figured out
that an UP peak is a "0" and a down peak is a "1", or so it seems based on
our interpretations below. Anyway,
your help in determining the type of modulation and how to decode this in DSP
on a sound card would be greatly appreciated by thousands of APRS operators.
Objective: Our objective is to see if we can use the RINO for APRS
applications and so we have no intent to undermine Garmin's proprietary protocol.
In fact if we can decode this, then Garmin will sell thousands of RINO's to the
potential market of 20,000 hams worldwide that use APRS. But if we cannot receive
the RINO data into our existing APRS system, then we are not interested in it...
and it will remain as a kid's toy as it appears to be currently marketed.
NAME FIELD FOUND AND DECODED!
I saved another file with presumably the same LAT/LONG as before, but
I changed the name from AJ-------- to ABCDEFGHIJ. This changed 60 bits...
confirming the NAME field must consist of 6 bit characters.
Download the new A/D capture fileRinoABC.txt in text format.
From this file, Byon, then has figured out that he initially had the bits
inverted and the
6 bit characters are LSB first, 6 bits each with the letter "A" starting at
Octal 42. See his latest Email... ID Report!
If the above image of Byon's decoding doesn't show up (Netscape),
try this Link
Default Position: We dont know for sure the position being transmitted because
we must do the capture down in the lab with quick hand-eye coordination sitting at the
digital OScope, but indoors, the GPS loses lock and ERASES the LAT/LONG display!
But just outside the building, the LAT/LONG is
38 deg 59 mins NORTH and 076 deg 29 mins WEST.
New WAV file: Doug Quagliana has rebuilt the RinoABC.TXT file into a
new WAV file.
And OH2BMS has also generated a WAV file for comparison.
Phil Karn, KA9Q suggests it is MSK at 600 or 1200 baud because
lots of business and police
2-way radio use this and there are chips available to do this.
See his Email.
MORE DATA! I Captured two new files WITH position data. This is very difficult
to do because when you come indoors, it Deadreckons the posit until it disappears.
For this test, I crept towards the door
at inches per minute and then shoved the GPS under my ARM pit and RAN inside. This
stopped it from Dead reckoning and I was able to record the positions just prior to them
saying "lost lock" and then it erases the screen! SO here are the new data files:
ABCDEF-1.txt with callsign of "ABCDEFGHIJ"
and position of 38 59.089N by 076 29.117W with radio-1.
BOB-2.txt with callsign of "BOB"
and position of 38 59.091N by 076 29.116W with radio-2.
NEW MULTIPLE WAVE FILES:
With the progress being made by Darryl, Byon and Brent, here are three 60 second WAV files
recorded via the speaker jack on another FRS radio under the conditions as shown.
[The waveforms here are drastically alterted from the original off the Discriminator
so these may not be at all useful]. I had two RINO's and alternately pressed PTT on each so as to get some data
every 5 seconds (instead of every ten). In all cases I alternated between the two radios.
One has the callsign of AJ and the other has BOBS1. AJ's has the FIRST icon available in
list and BOBS1 has the last ICON available. The general location of the tests were
38deg 59 mins north and 076deg 29mins west.
Test-1 taken with both HT's by the laptop, not moving.
Test-2 Walked West about 150'
Test-3 Walked South about 150'
Play for a full minute. If someone wants to extract all the DEAD TIME between bursts
and combine these files into one WAV file, go ahead if it would be useful for testing...
The
If you are successful, contact me at WB4APR@AMSAT.ORG.
Can anyone take this ASCII text A/D file capture and turn it into a WAV file
so we can distribute it as a clean WAV file?
Note: they may have dropped some
bytes in the LAT/LONG since these FRS radios only have a range of about 2 miles.
Maybe they don't even transmit the Degrees of LAT/LONG, but only transmit
the minutes? But we do have some one reporting a 20 mile DX contact, so they
have to have preserved at least all the minutes...
____
Welcome to our Techical Site. If you are interested in an overview, then visit our
Marketing Site
Copyright © 1994-2005
Radioactive Networks ,
darryl@radio-active.net.au
This page was last updated 2005-09-02 15:14:05
This page was last compiled 2005-11-15 18:52:13
Question or Comment? Click
here
|
