How GPS works:
In it’s most basic form, radio signals are beamed (transmitted) down to the earth’s surface from satellites, usually made up of around 30+ satellites in what is known as a constellation. These are spaced apart, and for us to get a lock on our current position, you need a minimum of three signals (from different satellites in the constellation) to get a triangulated position fix (in reality you need four, the extra one is for timing/error correction). However most GPS receivers will lock on to between 4 and 15+ satellites to get data from at a time; these may change as you move around an area/locality, etc.
There are three main GPS constellations in wide use, these are: GPS (the original US constellation), GLONAS (the newer Russian constellation) and GALILEO (the new and still incomplete European constellation), there are also others being deployed.
Just to add confusion, there are also ground based systems that are used to augment the GPS constellations, these include systems such as EGNOS and WAAS. These were very useful in the early days when GPS chip sensitivity was limited and it was not unusual to have significant errors in position accuracy. WAAS and EGNOS are only supported by devices that have them as an option.
You can find more detail, here: https://en.wikipedia.org/wiki/Global_Positioning_System
The different types of antennae used in most devices:
There are two main types of GPS receiver antennae in use in almost all consumer devices with GPS receivers in them, these are:
The quad helix antennae is widely regarded as the most sensitive and accurate of the antennae used by commercial grade GPS Receivers. The problem is that this antennae is huge, so is only used on handheld GPS receivers, such as the GPSMap60, 62, 64, 66 etc. This type of antennae is also widely used in military grade GPS receivers.
The rest of the devices from Garmin, such as the Nuvi, Zumo, Fenix series, Forerunner, Oregon, Edge, Etrex, Montana, Monterra, etc. all use a patch antennae instead. This is an internal antennae, unlike the quad helix one. Because it is internal, it works best when the screen is faced up to the sky (horizontal), unlike the quad helix antennae which works best when the device is vertical, with the antennae facing straight up. This type of antennae is also used in smartphones.
Some GPS receivers also have a port (interface) for external antennae. This can improve the signal strength in challenging conditions; weather, city centers (with high buildings), deep forest, canyons, etc.
The “Canyon Effect”, what it is and what it means:
As it’s name would suggest this is a phenomena that affects all GPS receivers when in canyons; be they natural ones, or so-called “urban” canyons (where you have lots of tall buildings). This can also occur in deep forests/woodland where the view of the sky is almost non-existent. All antennae types will struggle in these conditions. I have seen (on older devices) accuracy be out by 30-50m (100-160ft+). Most modern devices will have accuracy errors between 10 and 30m (33ft -100ft) in those difficult conditions.
Also GPS does not work well indoors (no proper direct view of the sky), or at all underwater/underground (duh!) 😉
The “Canyon Effect” is also known under a different name, this being “Multipath Errors”. This describes the issue well as the problem is that the GPS signal is being reflected from other surfaces causing what you might call a GPS signal echo as it bounces off multiple surfaces and the GPS Receiver gets multiple sets of position data and gets confused.
GPS Receiver Chips
There are a number of different chipsets used in GPS enabled devices, and these have significantly improved over the last 20+ years. All GPS chips used in today’s devices are high-sensitivity ones. The chip used is as important as the antennae used.
Many modern consumer devices use software to fix common errors, these include “lock to road” so that you don’t look like you are driving in the hedge, field or through a building, when on a road. There are others that use “smoothing” techniques to make the recorded GPS track look more uniform (smooth), and there are numerous others, they all “process” the raw GPS track data and change it to suit their, or your requirements.
This is, in my opinion (especially for off-road use) disingenuous, and causes much of the “my smartwatch/GPS/phone, etc. records better GPS tracks than yours.” that you see in the many vendor forums.
Please only compare like with like, raw “unfiltered/un-edited” GPS track data, otherwise you are comparing oranges to a bowl of petunias!
GPS Receiver accuracy is a mixture of chip used, antennae type, conditions, software controls (such as smoothing, lock to road, etc.) It is amazing just how good many modern GPS Receivers are nowadays, even using tiny patch antennae in a watch is better than handheld devices (with a quad helix antennae) from just 15 years ago. Many modern GPS Receivers also support using more than one GPS constellation at a time, this can improve accuracy of you use the correct ones for the region/area you are currently in.
In general use, you can expect GPS accuracy to have an error margin of less than 5m (18ft), and often sub 3m (9ft).
Lots of accuracy errors can be reduced by just changing the mode or recording settings used, as I show in the testing below.
Testing and the Test Results:
As I may have mentioned I have quite a number of Garmin devices (both used for testing and when out hiking).
I am often asked how accurate they are, especially the Fenix 6 series with the new Sony chip.
So, I went out for a 12 mile hike this week and had both an Oregon 700 and my Fenix 6X Sapphire set to record my track, and I have some screen shots of the two tracks (overlaid on my British Isles map).
Both devices were set to use GPS+Glonas and the Fenix was set to use 1 second recording. The results (differences) are quite interesting. Both devices were exposed to the sky (not in a pack or covered up in any way) all the time.
The hike included deep woodland, open fields, urban areas and overgrown footpaths and bridleways where the sky was impossible to see.
First the Oregon 700 screen shot (recorded track is in grey):
And here’s the recorded track for my Fenix 6X Sapphire (recorded track is in grey):
The screen shots only cover a small part of the whole 12 mile hike.
In quite a few places the Fenix 6X has recorded a more accurate track than the Oregon, quite impressive! Likewise in some places the Oregon is more accurate (which is what I expected as it has a larger patch antennae than the Fenix 6X). Conditions were excellent clear skies almost all the hike.
Update: Another track test, this time with just GPS (rather than GPS+GLONAS), screen shots below:
Orgeon 700 (just part of the recorded 12.5 mile track, recorded track is in grey)
Fenix 6X Sapphire (just part of the recorded 12.5 mile track, recorded track is in grey)
In quite a few places the Fenix 6X has recorded a more accurate track than the Oregon, quite impressive! Likewise in some places the Oregon is more accurate (which is what I expected as it has a larger patch antennae than the Fenix 6X). Just using GPS on the Fenix 6X Sapphire shows it struggling (slightly) in deep wooded areas, but still acceptable. Conditions were changeable for this hike, with large parts of the hike in overcast conditions and occasional clear sky.
For those of you with a Fenix 5X, 5 Plus (all models), 6 Pro, Sapphire or Solar (you can probably do this on a Tactix, MARQ, Forerunner 945 or other similar devices) who want to know how to set the recording to 1 second, here are the instructions:
1. Press and hold the middle-left button. Select System from the menu shown.
2. Select Data Recording from the menu shown, and set it to Every Second.
That’s it, any activity that you have set to use System Settings will use this recording setting.