|GPSR Tests on Waldo Canyon Trail|
I've hiked the Waldo Canyon Trail many times with various
of Garmin GPS receivers. The receivers were the 12XL, eTrex
Venture, GPS 76, and GPSmap 76. The 12XL was used 18 times, the
Venture was used 14 times, the GPS 76 was used 20 times, the GPSmap 76
was used 12 times. These counts are somewhat approximate since some
early hikes were not done with testing in mind and other hikes were
done with different test goals in mind. A few points are included
in the overall average from a couple of hikes in the area that only did
part of the Waldo Canyon Loop. So when comparing results from one
receiver to the next keep in mind they did not get equal time!
For a valid comparison, it would have to be done many more times.
Nevertheless it is still informative.
The locations (waypoints) along the common leg of the trail were marked on the way in and some hours later on the way out. On a few hikes, I hiked the loop and also went back up and down the canyon for some additional marks in the canyon. Any location where I noticed the coordinates on the receiver had changed after the initial mark was probably marked a second time to record the "latest" new value. This generally happens when the satellite constellation is poor - usually in the canyon. One mark is usually done when the reading is stable. That is when the coordinates don't change at all or change only slightly within seconds to a few minutes.
The reference trail map was created by averaging many of the track logs the way I would if I were creating a trail map. That is I visually do an approximate average of the track logs keeping in mind my knowledge of the trail (minor input) and the quality of the track logs (ie. ignore parts that are clearly wrong). The averaged waypoints are included in the background so that the tracks may be drawn through them when appropriate. All of the tracks were not used because they created to much of an unreadable mess.
Pictures are included to give you an idea of what the environment looked like.
The reference trail (in purple on the topo maps) was overlaid on USGS Ortho Photos to give you an idea of the accuracy.
Check out the following links for more information:
The following link is a web page with 32 pictures. The pictures are around 100KB each for a total of about 3.43MB. Each picture is available individually through the picture index link near the bottom of this page.
Waypoint lists are broken down in various ways and include the maximum offsets from the average and the average offset from the average for the eastings and northings. Also listed is the standard deviation "of the population" (STDP in Microsoft Excel). That is it was divided by N instead of (N - 1). So the standard deviation will be a bit larger. Someday I will change my waypoint processing software to (N - 1) but it is low on the priority list.
Important! Note that since the eastings and northings are in meters the statistics for them are in meters. The elevations are in feet so the statistics for them are in feet. All data logging was not done to the same satellite constellations on different days. In particular note that the GPSMap 76 had the fewest test runs and therefore, is in need of more data.
An additional look at points recorded at the "Gold Camp Road: AE4292" survey location is available at:
The purple line is a common reference line and an approximate
average intended to represent the trail.
data logging was not done to the same satellite
constellations on different days. In particular note that the
GPSMap 76 had the fewest test runs and therefore, is in need of more
Southern section track closeups:
Tracks with the 12XL connected to an external antenna:
Tracks with the GPS 76 connected to an external antenna:
Maps and Orthophotos of the area (use the scroll bars especially when you get a "blank" screen):
USGS OrthoPhoto of the area
USGS OrthoPhoto with the reference track
USGS Topographic map of the central area
USGS Topographic map of the northern area
USGS Topographic map of the southwestern area
Test Dates and a Few Comments
Index to Pictures and Maps
For the past two years I've used a GPSmap 76 in my hand (internal antenna) and a GPS 76 in a belt holster connected to an external antenna. Sometimes one will do better and sometimes the other will do better. If I enter an outdoor/park type restroom with concrete walls and maybe a metal roof, the GPSmap 76 using the internal antenna will often loose lock but the one with the external antenna won't. That shows an advantage of the amplified antenna.
However, by amplifying weak signals, the external antenna will also amplify weak multipath signals which can result in less accurate data.
When out hiking I try to keep an eye on the satellite page. When a satellite starts coming into view, it can take one receiver a
minute to lock on but five minutes for the other receiver to lock on
even though I may have moved slightly to maximize the S/N bar on the
one that hasn't locked on yet. I've never counted the number of
times but it's closer to 50/50 than 10/90. So I can't say that an
helps much in that case. on occasion I till the external antenna
a bit so the satellite is above the plane of the antenna.
In terms of reception I don't find a MAJOR difference in the
above list. I don't really use the 12XL anymore because it lacks
other features (far less detailed track logs and no elevation).
In some difficult reception cases, I'd take the 12XL (with or without an external antenna) over the Venture.
The antenna is often being blamed for poor reception (inability to obtain a lock on the satellites) when the problem is more likely in the digital signal processing area. The problem is in finding the correlation peak in the frequency/code space. Consumer GPS receivers, with perhaps only two (or three) correlators per channel, can spend only a millisecond or so at each point in the frequency/code space. The more time spent at each point (dwell time), the greater the sensitivity.
Global Locate® makes a chip with over 16,000 correlators (GL16000) that can search all code delays in parallel with long dwell times and finish in a reasonable amount of time. It was designed to receive GPS signals indoors. Long dwell times allows it to pick up very weak signals. (Code space of 1023 times 12 channels times 1 correlator per
|Return to the Test Index Page|
|Return to the Index (Home) Page|
|Copyright © 2003, 2004, 2005 Dan Anderson. All rights reserved.|