Go directly to conclusions

THE BREADCRUMB PROBLEM

There have been some reports of problems with the 276C GPS recording tracks after coming to rest. I set out to determine the circumstances of the the problem and to see if it constitutes a threat to my data recordings.

I recorded tracks with my data logger connected to my 276C, the 276C is set to record tracks on RESOLUTION mode and the resolution is set to 0m, this I have found produces conveniently sensitive recordings for doing accurate backtracking. I do not normally record tracks using my GPS, but this resolution would produce too much data for normal track recording and result in the GPS memory filling up prematurely.

Track recording setup page on my 276

Typical track record sample

136 2/26/2005 4:22:07 PM 21 m 14 m 0:00:01 52 kph 124° true S28 44 06.9 E32 04 03.8
137 2/26/2005 4:22:08 PM 20 m 16 m 0:00:01 57 kph 124° true S28 44 07.2 E32 04 04.2
138 2/26/2005 4:22:09 PM 20 m 17 m 0:00:01 60 kph 124° true S28 44 07.5 E32 04 04.7
139 2/26/2005 4:22:10 PM 20 m 17 m 0:00:01 62 kph 124° true S28 44 07.8 E32 04 05.2
140 2/26/2005 4:22:11 PM 20 m 17 m 0:00:01 60 kph 124° true S28 44 08.1 E32 04 05.8
141 2/26/2005 4:22:12 PM 20 m 16 m 0:00:01 56 kph 124° true S28 44 08.4 E32 04 06.3
142 2/26/2005 4:22:13 PM 21 m 15 m 0:00:01 54 kph 124° true S28 44 08.7 E32 04 06.7
143 2/26/2005 4:22:14 PM 21 m 14 m 0:00:01 51 kph 124° true S28 44 09.0 E32 04 07.2
144 2/26/2005 4:22:15 PM 21 m 13 m 0:00:01 48 kph 124° true S28 44 09.2 E32 04 07.6
145 2/26/2005 4:22:16 PM 21 m 12 m 0:00:01 44 kph 123° true S28 44 09.5 E32 04 08.0
146 2/26/2005 4:22:17 PM 21 m 11 m 0:00:01 40 kph 125° true S28 44 09.7 E32 04 08.4
 

The following tracks were recorded :-

Data logger connected to GPSMAP 276C

The following track from the GPS memory:-

In order to determine the nature of the problem two tests were conducted.

1) Stop and then drive away in a straight line, then return to the same spot at 90 degrees and turn back onto the line.

2) Stop and pull away on a curved route.

These tests conducted at #1 and #2 marked on the above map, respectively.

TEST 1

In this recording the blue line shows the direction of travel of the blue track. The vehicle was stopped on the west side and after a short while, pulling away there were no breadcrumbs recorded until a change in direction of travel occurred.

The Red line shows an approach to the same point from the North and stop, followed by a sharp right turn after being stationary for about 2 minutes. Once again no tracks are logged until a change of direction occurs.

The following tracks from the data logger show that the position was accurately transmitted by the GPS even while the GPS was trying to determine if it was really moving or not.

TEST  2

This test was to determine if the lost breadcrumb was still experienced when moving away in a curve.

I drove up to the arrow at point #1 and stopped for about 2 minutes, then drove away around the circle, it is quite clear that the data logger track below recorded breadcrumbs immediately when pulling away. The second pass was recorded without stopping, i.e. I stopped at point 1, pulled away, drove around the circle and passed point 1 again without stopping.

and zooming in on this area of the track

Shows that the track with stopping on the first pass, and then without stopping on the second pass are sufficiently close to say there is no appreciable difference.

Surprisingly enough the GPS data log below shows no loss of breadcrumbs and additionally the extra points recorded on the data logger while stopped are suppressed in the GPS track log :-

CONCLUSIONS

I believe that the GPS276 is programmed to detect standstill and it requires two non-linear position samples after stopping, to determine that the GPS is in fact moving again. It disregards the data lost while it makes up its mind that it is in fact moving again, and as this is in a straight line, Garmin figures this to be O.K. ?

Clearly the problem does not exist with the serial output position to the data logger.

Donald Pittendrigh

Resolution of the problem using the speed filter

I found that by playing with the speed filter I could get the missing breadcrumb problem to go away.

To recap, the Garmin 276C stops tracking when the vehicle is stationary, this means if you have a shaky satellite or two, the scatter of breadcrumbs resulting from changes in your fix postion, although stopped, cause a clutter on your screen. If the GPS determines you have stopped moving it stops recording, but then has the problem to distinguish between a valid resumption of movemment or a movment due to error in your fix.

The auto setting APPEARS to deal with this by waiting until it has 2 points which move away from the original point of stopping and constitute a change in direction. Theoretically this should not be a problem as when you start to move again, if the next breadcrumb recorded is where you changed direction, then your track is not seriously asdversely affected, as the point of stopping to that breadcrumb is a straight line anyway. This works sometimes but makes a mess of your tracks when you are mapping.

Taking the filter off completely causes a mess every time you stop, but by setting the filter to manual, it is possible to find an optimum setting which allows the recording of tracks to begin at an appropriate distance from the point of stopping.