Antenna design

Slot Antennas
Slot antennas are used extensively in aircraft and radar applications. The basic slot antenna is a 1/2 wave slot cut in a conducting sheet of metal. The feed point is across the center of the slot and it is balanced. The feed impedance is high, typically several hundred Ohms. Because the slot antenna is the opposite of a dipole, that is, the slot antenna is a non-conducting slot in a sheet of metal as opposed to a conducting rod in free air, the slot antenna is similar to a dipole. However, it does exhibit some differences as follows:
• The feed point is across the center instead of in series, so the feed point impedance is high instead
of low
• E and H fields are switched so that the polarity is opposite
• A horizontal slot is equivalent to a vertical dipole
• The slot antenna may be of interest if the RF unit must be placed in a metal enclosure where the
slot antenna could be made in the enclosure itself
• If the slot antenna is cut in the center, a 1/4 wave slot antenna is created which is equivalent to the
• Impedance matching is accomplished by tapping across the slot close to the shorted end
The slot antenna can be used if a metal enclosure is required or if considerable board area is available. If a slot antenna is implemented in a PCB made with FR4 material, considerable dielectric loading occurs which causes the physical length to be shorter than expected.

Chip Antennas
Numerous commercial chip antennas are available. At first glance, chip antennas appear to work for no apparent reason. However, careful investigation reveals that most of these antennas are based on a helix, meander, or patch design. To ensure proper operation, it is very important to follow the manufacturer’s recommendations regarding footprint, ground areas, and mounting of the chip antenna. The “keep out” area around the antenna is especially important. Even following the recommendations does not always guarantee good performance due to de-tuning by nearby objects. It is expected that fine tuning of the antenna and/or a matching network is required to ensure satisfactory performance. Because chip antennas normally, but not always, use a ceramic material with higher dielectric constant and lower loss than the usual FR4, it is possible to build smaller antennas with reasonable efficiency.

Efficiency is not exceptionally high and is typically in the range of 10-50%, which corresponds to 3-10 dB loss (-3 to –10 dBi). The lower number being inferior products with high inherent losses. As already stated, buying a chip antenna does not guarantee good performance. However, while they provide the smallest antenna solution possible, the size reduction comes at a cost both in performance and pricing. If a slightly larger PCB area is available than is required by the chip antenna and the “keep out” area can be allocated to a PCB antenna, it is possible to implement a PCB antenna with the same or better performance than a chip antenna.

Many of the antennas already mentioned in this note are single-ended and designed to have a feed point impedance close to 50 Ohms. A balun is required to interface these antennas to a balanced output/input. The balun converts a single ended input to a balanced output together with an optional impedance transformation. The output is differential. That is, the output voltage on each pin is of equal magnitude, but of opposite phase. The output impedance is normally stated as the differential impedance. That is, measured between the two output pins. The balun is bidirectional. The balanced port can be both input or output.

Several discrete circuits are available that perform as baluns but most of them are sensitive to input and
output loading and PCB layout issues which requires cumbersome fine tuning. Also, all of these require at
least two chip inductors. In the 2.4 GHz band, there are small ceramic baluns which are easy to use and
are less sensitive to the PCB layout with standard output impedances of 50, 100, and 200 Ohms.
The cost of a discrete balun is comparable to the ceramic balun and the ceramic balun requires less board
space. Therefore, the ceramic balun is recommended for most designs.

Miniaturization Trade-offs
As previously stated, reducing antenna size results in reduced performance. Some of the parameters that
suffer are:
• Reduced efficiency (or gain)
• Shorter range
• Smaller useful bandwidth
• More critical tuning
• Increased sensitivity to component and PCB spread
• Increased sensitivity to external factors
Several performance factors deteriorate with miniaturization, but some antenna types tolerate miniaturization better than others. How much a given antenna can be reduced in size depends on the actual requirements for range, bandwidth, and repeatability. In general, an antenna can be reduced to half its natural size with moderate impact on performance. However, after a 1/2 reduction, performance becomes progressively worse as the radiation resistance drops off rapidly. As loading and antenna losses often increase with reduced size, it is clear that efficiency drops off quite rapidly.

The amount of loss that can be tolerated depends on the range requirements. Bandwidth also decreases, which causes additional mismatch losses at the band ends. The bandwidth can be increased by resistive loading, but this often introduces even more loss than the mismatch loss. The low bandwidth combined with heavy loading requires a spread analysis to ensure adequate performance with variations in component values and PCB parameters. So, it is often better not to reduce antenna size too much if board space allows. Even if range requirements do not require optimum antenna performance, production problems and spread are minimized. It is also best to keep some clearance between the antenna and nearby objects. Although the antenna may be retuned to compensate for the loading introduced by the surrounding objects, tuning becomes more critical, and the radiation pattern can be heavily distorted.

Potential Issues
Numerous things can go wrong with an antenna design. The following list provides a few do’s and don’t’s
which may serve as a good checklist in a final design. Many of these items seem obvious to the
experienced antenna designer, but many of these issues are routinely encountered in practice. This is
obviously not a complete list.
• Never place ground plane or tracks underneath the antenna
• Never place the antenna very close to metallic objects
• In the final product, ensure that the wiring and components do not get too close to the antenna
• A monopole antenna will need a reasonable ground plane area to be efficient
• Do the final tuning in the end product enclosure, not in open air
• Never install a chip antenna in a vastly different layout than the reference design and expect it to
work without tuning
• Do not use a metallic enclosure or metallized plastic for the antenna
• Test the plastic casing for high RF losses, preferably before production
• Never use low-Q loading components, or change manufacturer without retesting
• Do not use very narrow PCB tracks. The tracks should be relatively wide as space allows


COMMENT Uncategorized

  1. Guru

    SMART antenna

    The field of RFIC design and manufacturing always faces constant pressure for increased performance while still decreasing the chip size and accommodating more functionality in the die and in the package. The long term goal has always been to integrate everything on the same die (i.e, System-on-chip SOC). However, cost and performance issues have forced the industry to rethink its strategy towards using the package to incorporate some of the functions. For example, while inductors in today’s technology have seen tremendous increase in quality factor compared to a decade ago, their performance is still limited. Several system-in-a-package (SiP) and silicon-on-a-package (SOP) techniques are currently being pursued to incorporate passive elements in the package for module-level integration. The above approaches target mainly low density circuit architectures and limited numbers of passives.
    Multi-antenna systems on the other hand, operate with multiple wideband transceivers with Tx and Rx RF paths functioning simultaneously. Clock distribution and I/Q matching usually pose major challenges as the lateral dimensions increase on chip (for SOC). They also require digital detection techniques whose complexity grows with the antenna and constellation size. As we move towards supporting millimeter-wave standards (24GHz and 60GHz) for future communication systems, the increase in frequency will complicate controlling interconnect parasitics due to the small wavelength. Howver, the reduced antenna size and spacing will make it possible to integrate them directly on chip and in the package