|

Installation guidelines for static GPS Antennas
|
Background
Unlike a TV broadcast satellite,
GPS satellites are not stationary, but are in near-circular Earth-orbits. They
take approximately 12 hours to circle the globe, but because the Earth is also
rotating, a stationary Antenna will see a particular satellite reappear at
intervals of around 24 hours.
There are up to 24 satellites in six different orbits equispaced around the
equator, which have been described as being reminiscent of a spherical birdcage.
Their orbits are inclined to the earth's axis, and therefore the satellites do not
pass overhead at the poles. There are three or four satellites spaced around
each orbital plane.
A GPS Antenna is to a first approximation, omnidirectional above the horizontal
plane, and it is capable of receiving signals from as many satellites as are
above its visible horizon at a particular time, provided that its view of them
is not obstructed by local objects.
Satellites that appear near the horizon will provide only weak signals, and the
Antenna and receiver are usually programmed to reject those that are below an
angle of 5 to 10 degrees from the horizontal plane.
The number of satellites visible at a particular time will vary from as few as 4
to as many as 10. Because they are in different orbits, some will appear for a
short time by just rising briefly over the horizon and others may track almost
overhead and therefore be visible for several hours,
The signal level available from a GPS Antenna sited anywhere near the Earth's
surface does not vary significantly with the Antennas' height above the local
ground level. In a flat desert region it would work well if mounted only a few
centimetres above the ground, and there would be no advantage in mounting it on
a tall mast.
However it is essential that the Antenna has a 'line-of-sight' relationship with
available satellites, because at the frequencies used for GPS (around 1.6GHz),
signals are blocked by solid or semi-solid objects in the same way that light is
blocked.
In typical urban locations it is therefore frequently necessary to mount the
Antenna in a high position in order that it can 'see' a lot of the sky, despite
the presence of signal-opaque objects such as adjacent buildings.
When viewed from the Northern
Hemisphere, and because of the inclined orbits mentioned above, the sky-tracks
described by satellites viewed from the Earth's surface tend to concentrate in
the Southern sky. If, at a particular site, a large obstruction to an Antennas
view is unavoidable, it is usually advantageous to site the Antenna so that the
obstruction is on its Northern side.
Although apparently complex, the
visibility of all the satellites at a particular point on the Earth is precisely
predictable, and there are (PC) computer programs available that will accurately
forecast the situation at any point in time, past, present or future, and at any
place on the globe.
In addition to the basic geometry
of the GPS system, it is also necessary to consider the likelihood of the
Antenna receiving signals that have bounced off nearby reflecting surfaces,
concrete walls for example. This multi-path effect, as it is known, can
seriously degrade the performance of a GPS system, and may cause complete loss
of signals from otherwise 'good' satellites, due to the time-delayed reflected
signal cancelling out the direct signal. Whilst nothing can usually be done
about reflection from distant objects, it is wise to avoid siting an Antenna at
the same horizontal level as a wall or metallic structure such as a lift-motor
housing or water tank, which are common objects on many large buildings.
Unlike the short-term 'flutter'
caused by passing aircraft, that affects VHF-FM radio reception, the multi-path
symptoms in stationary GPS installations are frequently very long-term effects
caused by the movement of the satellites and not the reflecting objects. Signal
levels can be observed to fluctuate with periods of several hours as the
satellites position changes relative to the reflecting surface.
In addition the amount of energy reflected from a particular surface may vary
for example according to whether it is dry or wet, so the situation can change
from day-to-day.
Finally, it is important to
remember that the signal power received by a GPS receiver is very low indeed,
and it is prudent to avoid, as far as possible, siting Antennas next to
significant sources of interference; electrical switchgear, and Radio
(transmission) Antennas are obvious candidates.
Also, power transformers, particularly large ones used in sub-stations,
frequently appear to be a 'magnet' for rf energy from various sources and are
items to be viewed with suspicion.
The notes above are intended to
illustrate problems that may occur, rather than those that will
occur.
GPS receivers are designed to
work with small fluctuating signals, to survive sudden obscuration of a
satellites signal, and in Timing applications, can even operate for long periods
with only one satellite in view. It is because of these remarkable capabilites
that the vast majority of Antenna sites, if chosen intelligently and well
engineered will operate correctly from the day of installation onwards.
|
|
Handling GPS Antenna Assemblies
Most GPS Antennas assemblies
contain both an Antenna element and a Low Noise amplifier (LNA). They are
usually contained in a plastic Radome cover that is transparent to radio
frequencies, but hopefully impervious to weather
Shapes vary, but are dependent on the type of Antenna element inside. The three
most common are:
The 'double helix' type, which
have long tubular covers,
The 'crossed-dipole' type which
have egg-shaped covers,
The 'patch' type which use
microstrip (idg-dipole technology) and may have low-profile flat-top housings.
For obvious aerodynamic reasons the latter are often used on aircraft, but also
appear frequently in ground-based applications
Antennas with domed or pointed tops are generally more capable of rejecting the
build-up of winter snow or autumn leaf-debris and are also less attractive to
perching birds -- this may seem funny, but it is a real problem in some places.
All these types are fairly rugged as far as the Antenna element itself is
concerned, but sometimes the LNA is more vulnerable both to electrical (static)
damage and mechanical shock.
Despite some optimistic manufacturers' specifications, it is advisable to avoid
dropping any type of GPS Antenna during installation, and in particular to avoid
fitting any unit known to have suffered this fate.
The types with egg-shaped or domed tops are particularly easy to drop.
To provide the best protection, both from static and mechanical problems, it is
advisable to keep Antenna units in their packing until the mounting arrangements
such as poles, masts, brackets etc. have been completed.
It is almost impossible to test a GPS Antenna
without connecting it to a GPS receiver, so attempts to diagnose problems using
a test-meter (or even a Spectrum Analysers) are at best pointless. In some
cases, attempting to measure impedance at the Antenna connector can be fatal to
the internal LNA, which is often a sensitive low-voltage Gallium-Arsenide
semiconductor chip.
In particular it is essential that if rf
cable-testing equipment is in use at a site, that the GPS Antenna is left
disconnected from the cable until after all cable testing is complete. The power
level required to destroy the LNA is very low. It is also important that the
Receiver unit at the other end of the cable is disconnected during cable
testing.
|
|
Cables (Downleads)
Most GPS Antenna use coaxial
cables as downleads, the exceptions being so-called 'Smart' Antennas that have
multiple twisted-pair cables.
Because the GPS signal levels are
very low, it is often necessary to employ ultra low-loss cables to avoid
attenuation of the signals to unusable levels.
If the cable run is short (less
than 12 metres total), it is possible to use RG58 cable for a temporary rig, but
this is usually inadequate for permanent use because RG58 is only single
shielded and is vulnerable to noise pick-up.
A better cable, for runs up to
about 30metres, is RG214; this is a 10mm diameter double-shielded type that has
better screening efficiency and lower losses than RG58.
For runs of 30 to 70 metres it is
possible to use ultra-low-loss types such as Westflex 103 or Belden 9913. These
cables are also double-shielded, but have much lower rf losses than RG214. The
internal construction is air-spaced, and they have a solid (ie non-stranded)
inner conductor which means that they are far more vulnerable to damage due to
handling. During installation, it is particularly important to avoid sharp bends
or accidental kinking to less than the specified minimum-bend-radius, which is
about 600mm. This is not a very sharp bend!!
Once a cable has been bent to less than its specified minimum bend radius,
straightening it out will not help if it is already damaged internally.
Although they appear to be very strong, all co-axial cables can be damaged by
stretching and should not be pulled through long ducts without considerable
care. This applies in particular to the very-low-loss types; crushing them with
excessively tight cable straps should also be avoided.
If necessary, Andrew 'Heliax'
cable can be used successfully for GPS receiver applications. This cable is
stronger and has lower losses than the types discussed above but it is heavy and
as a result, may require more mechanical support along the run. In addition it
is larger in diameter, and is relatively expensive.
Many different grades of cable
are known by the generic title of 'Heliax'. The most satisfactory type for GPS
downleads is LDF4-50, which has sufficiently low losses to allow runs of
100metres or more with typical Antennas and Receivers. This is a 13mm o/d cable,
and it can be made up to N-type connectors, although both 'Heliax' and other
low-loss cables may require special N-type connectors and special cutting and
stripping tools for satisfactory termination.
|
Do's and don'ts summary
|
The GPS antenna that is specified for use
with the 1804 in ground installations is a fixed-pattern type which has a
built-in preamplifier, powered via the co-axial RF downlead cable from a
low-voltage d.c. source
(< 5 volts) that is built into the 1804.
The signal connector on the antenna head assembly, although sealed at its
mounting face, is not a weatherproof connector.
In order to provide weather protection for the interface
between this connector and the antenna downlead, and to avoid mechanical strain
on the connector when the downlead is connected, a waterproof housing with an
entry conduit is provided.
Rapco Antenna-downleads, both the standard 20 metre version
(E43) and the heavier 50 metre type (E44), are ready-fitted with a screw-on
cable gland which allows the assembly to be weatherproofed at the point of cable
exit.
If longer cable lengths are required then consult Rapco for
details.
In extreme environments the user should provide further
protection for the co-axial cable itself from its point of exit on the antenna
assembly.
Information regarding antenna installation is given in the
pictures below, which depicts the Rapco antenna assembly (Type 4/4B).
GPS Antenna Installation
If the unit is operated with antennas not supplied by Rapco,
the following points should be noted.
There is a dc voltage ( <+5V )on the co-axial cable
connector J4. This should not be shorted to ground, or connected to any dc load,
other than an antenna with a dc current consumption of less than 30mA.
If sharing an antenna with other equipment, any power
splitter that is used must provide a dc block (capacitive coupling) in the feed
to the 1804 to prevent the dc source in the 1804 becoming connected to another
dc source in other equipment.
Failure to observe these precautions may result in permanent
damage to the unit.
The GPS L1 signal level provided to the 1804 in these
circumstances must be approximately equivalent to that from an isotropic antenna
with a 20 - 25dB gain preamplifier connected to the unit via a loss-free cable.
Poor signal-to-noise ratios in the GPS receiver can result in
difficulties in maintaining lock. This can be caused by excessive signal
attenuation, and in some circumstances, from an excess of signal. These
situations may arise if the overall signal gain in the antenna circuit
i.e.(antenna preamp gain) - (cables + splitter losses) is significantly greater
than +25dB.
The 'signal level' readout from the RS232 port ( RL Command)
on the 1804 is in fact a measure of the correlator SNR, and is a good guide to
the rf behaviour of the 'front-end'.
In a typical scenario with a well-sited ground antenna there will normally be
several satellites in view that are yielding signal-level readouts of above 10
and possibly up to 20.
This range of levels will also apply with an airborne antenna
situation, but there are sometimes more visible satellites at altitude, so the
receiver (which favours the better signals) may have more choice and may be able
to better optimise the situation.
If it is found that there are a large number of satellite
signals that are (simultaneously) higher than (say) 18 on the signal-level
readout, this may be an indication of excessive signal gain in the antenna
system. This could give rise to problems in the presence of interference or multi-path reception.
|
|
|