Test Results of some Low Cost GPS Antennas

Recently I obtained some low-cost GPS antennas and a perfect platform for test : the popular Garmin 12XL hand-held GPS receiver (s.w. v 3.62). The antennas are:

Check my Overview of Low-Cost GPS Antennas for the manufacturers data.

It's my aim to compare the low-cost antennas and try to pin-point the differences. This may help you to decide which antenna is suited best for your application. I am not the only one playing around with antenna's ,don't forget to visit David Martindale's pages with extensive test results.

Please contact me for errors, improvements, etc., and help me enhance the quality of the page.

1. Actual power consumption test

The actual current delivered by the Garmin to the antenna has been measured directly by interrupting the cable between receiver and antenna.
The results are:

Mighty Mouse 10 mA
Big Brother 28 mA
Lowe active 20 mA
NovAtel 511 20mA

The Mighty Mouse scores very good, Big Brother's higher gain comes at a cost ...

2. Azimuth sensitivity test

The purpose of this test is to determine the effect of irregularities in the axial radiation pattern of the antenna on the receiver signal to noise ratio (SNR), and obtain an idea about the radiation pattern of the antenna as a function of the satellite elevation.
The test is carried out as follows:

  1. The antenna under test is mounted on a small, round ground plane (diameter 27 cm) in order to suppress multipath effects which may obscure the test result. The groud plane is mounted on a pole on the roof of my house, with an unobstructed view to the sky.
  2. The antenna is connected to the Garmin 12XL with its own connector (Tri-M antennas) or a small adaptor cable (Lowe).
  3. The Garmin is set up to output NMEA messages - the message reporting satellite elevation, azimuth and SNR is the one of interest.
  4. The 12XL is connected to a PC running a proper download program.
  5. The antenna is rotated to the startposition with the pigtail pointing towards the North direction.
  6. The antenna is rotated over 360 degrees with steps of 10 deg, the rotation is recorded (good old pencil and paper).

This procedure is repeated for all antennas under test within a short time interval, aiming at marginal satellite configuration changes only.

The results are listed in the table below. For each antenna and for each satellite in view the satellite elevation and azimuth are listed with the SNR averaged over one full rotation (36 values), the minimum SNR and the maximum SNR.

Big Brother Mighty Mouse Lowe Active NovAtel 511
prn el az avg min max el az avg min max el az avg min max el az avg min max
1 36 77 52.2 51 54 36 81 49.1 47 51 35 86 51.5 50 54 36 70 50.5 49 52
2 22 297 48.9 48 50 22 301 45.3 43 46 22 304 48.1 46 50 21 292 46.9 46 49
3 10 150 45.9 44 48 13 150 43.8 40 47 17 149 46.7 42 49 5 151 41.7 39 43
9 3 12 42.3 39 46 3 15 39.7 36 42 2 18 41.6 37 45 3 8 40.7 38 42
14 25 221 50 48 51 22 220 46.6 44 48 18 219 47.6 45 50 30 223 49.8 49 51
15 68 287 53.3 53 55 65 285 51.3 50 53 60 281 53.6 52 54 73 293 50 49 51
21 23 55 48.3 46 49 26 55 46.6 44 48 29 55 49.2 48 51 18 55 45.5 44 47
29 37 127 52.1 50 53 34 129 48.9 47 50 31 132 50.1 49 52 40 121 50.8 49 52
31 49 171 53.1 52 55 52 171 51.2 49 53 57 173 52.9 48 54 44 170 51.4 50 54

Some observations on the table:

1. For low elevation satellites the minimum and maximum SNR values vary significantly : 6 to 7 SNR units for the low-cost antennas and 4 units for the reference. Under sub-optimal reception conditions (e.g. under heavy foliage) the 'dip' in sensitivity may result in a loss of lock. As an example the SNR values of satellite 9 as received by Big Brother are given in the radial plot below. The outer ring represents a value of 46 SNR units, then a ring representing 42 units, then 38 and 34.
A symmetry around the Y-axis can clearly be seen. The misclosure at about 11 deg is due to increasing SNR because of the increasing satellite elevation during the measurement.

2. Big Brother's SNR values are roughly 3 units larger than Mighty Mouse, Lowe active lies in between.This last observation is detailed below in the plot of avarage SNR values against the satellite elevation.

3. Elevation sensitivity test

Setup during this test is identical to above, the antenna cable points towards North.

I connected the NovAtel 511 to my Garmin on April 29 and recorded data from roughly 19:00 to 23:00 local time (UTC + 1 hour). The next day (our queen mothers birth day) I hooked up the Mighty Mouse but recorded during a smaller interval because I drained the batteries in the Garmin (curve Mighty Mouse 2). On May 1 (our labour day) Big Brother was under test, May 2 (just a normal Saturday) saw the Lowe Active, May (Sunday) was my day off and May 4 (we commemorate the WWII victims) saw the Mighty Mouse again. Finally on May 5 (we celebrate our WWII liberation) I processed the data.

The plot below shows the signal strength for satellite 15 on the successive recording days. I took May 1 as reference day and shifted the other curves by 4 minutes per day (the satellite configuration repeats itself after 23 hour 56 minutes). The SNR values have been obtained by averaging the raw SNR values from the Garmin over 2 minutes. The time axis shows UTC time of day (seconds).                                        

For reference I plotted below the elevation and azimuth of satellite 15 in the same time interval (obtain the azimuth by multiplying the plotted value by 4).

A number of remarks can be made:

4. May 13, 1998: A Beautiful Evening in the Amsterdam Forest

First about the weather. I believe that we have the best weather since 1945 or so. Up to 30 deg C during the day, cooling down to 15 deg in the night. Very, very unusual for a typical May month in The Netherlands (and pleasant). So running the comparative 'forest' test on the antennas was more than a pleasure !

I took my car and drove to the only place within a radius of 50 km with some decent trees: the Amsterdam Forest. I parked my car on a location which was fully covered by trees. I placed the receiver on the roof of my car, hooked it up to a laptop to record the NMEA output.

I collected data for about one minute using the 12XL's internal antenna. Then I connected Big Brother to the receiver, placed Big Brother on the same location on the roof of my car, recorded for about one minute, hooked up the Lowe Active, recorded, hooked up the Mighty Mouse, recorded, and re-run the whole sequence two times just to be sure. What I try to establish with this test is the behaviour of the antennas under typical, i.e. sub-optimal conditions and try to pin-point some differences.

First some statistics on the number of satellites tracked. The receiver tracked the 6 highest satellites during all tests, although more were available during the test interval, but at a too low elevation. When using the Garmin internal antenna, the signal strength value was often at or close to the critical low value of 30 units, and sometimes 0 units (the receiver lost lock). The number of observations with the Mighty Mouse, Lowe Active and Big Brother were identical almost no loss of lock occurred with the external antennas.

Next the signal strength analysis. I averaged the signal strength values for each satellite during the observation interval of 1 minute. Then I compared the external antenna average with the internal antenna average, the table gives the additional signal strength of the external antennas.

Big Brother Lowe Active Mighty Mouse
run 1 +6.0 +4.1 +4.7
run 2 +5.8 +4.7 +4.4
run 3 +5.0 +3.4 +2.1




The trend of the earlier tests is again visible: Big Brother delivers a stronger signal than the Lowe Active and the Mighty Mouse, the latter two perform nearly equal.

Another conclusion (which is evident and not surprising): the external antennas perform significantly better than the 12XL's internal antenna, although during this particular test the number of satellites tracked was only marginally lower with the internal antenna compared to the external antennas.

5. Strong points and weak points

Below I listed my impression of the strong and weak points of the three low-cost antennas.

Big Brother Mighty Mouse Lowe active
+ - + - + -
high gain
great connector set
high power
larger size
very low power
great connector set
larger size very small medium power

All antennas show a fairly large amount of radial assymetry with satellites at low elevations. The external antennas show their advantage when working under sub-optimal reception conditions.
Prices are nearly identical, around US$ 60.

6. When use which antenna ?

If you have a clear view to the sky, don't bother about external antennas. The 12XL's internal antenna performs well.
If you use your receiver in a location with an obstructed view (inside a building or vehicle, obstruction by your body, to name a few examples), or if you expect serious attenuation of the signal (in the forest), you'll be better of with an external antenna.
When you have to run your equipment on batteries, the Mighty Mouse is your choice. If you need those extra dB's because of a very long cable to your antenna, or in very dense forest, take Big Brother. The Lowe Active scores with its lowest weight at a moderate power consumption.

Back to index