Академический Документы
Профессиональный Документы
Культура Документы
Brian Fischer
E
very year, many requests for pro-
posals for an anechoic chamber editors Note
are generated by companies and Most readers of the Measurements Corner column will be very familiar with anechoic
institutions that perform antenna mea- chambers as users, but many may not understand the numerous design trades that go
surements. The task of adequately into their construction. This issues column captures key requirements and trades that
specifying performance for an indoor one must consider when preparing to procure a new chamber.
anechoic chamber without driving
unnecessary costs or specifying contra-
dictory requirements requires insight institution facilities can then define chamber provides an environment that is
that is not always available to the author the appropriate square footage neces- free of echoing or other radiated signals
of the specification. Although there are sary to house the required antenna and reduces the effects of these undesir-
some articles and books that address range. This article intends to avoid able signals during the testing of a given
anechoic chamber design [1][3], a con- some of the common contradictory antenna. This article covers applications
cise compendium of reference informa- requirements. Some of these contra- where an antenna is radiating or receiv-
tion and rules of thumb on the subject of dictory requirements are not enough ing a given signal while we want to know
specifying ranges would be useful. real estate to accommodate a chamber its performance as a function of direc-
This article intends to be a helpful operating at low frequencies or levels tion. Excluded from this discussion are
tool in that regard. It starts by recom- that are not possible given available anechoic chambers for measuring radar
mending the proper type of range for absorber technology. cross section. The general range geometry
different antenna types and frequen- is shown in Figure 1.
cies of operation. Rules of thumb are THE anechoic chamber
provided to select the best approach The ability to measure an antenna is Range Type Selection
for the required test or antenna type. an important design requirement. It is There are several methods of measur-
Information is provided on the deriva- imperative to know whether the energy ing the radiation patterns of antennas
tions needed for other ranges, such is radiating properly and in the desired indoors: far-field illumination, near-field
as compact ranges and near-field test direction per the design specifications. It measurements, and compact range (CR).
facilities. Simple approximations are is also important to know how much ener- Like many other solutions, there are pros
used for absorber performance to gen- gy is traveling in undesired directions. and cons for each of these. There is not
erate a series of charts that can be used To measure antennas (like many other a single solution that is ideal for all types
as a guide to specify anechoic chamber devices that are being measured), there of antennas and situations. The type of
performance and size. Company and is the desire to have the antenna unaf- range that is most suitable for a given type
fected by its surrounding environment. of antenna is determined based on two
Digital Object Identifier 10.1109/MAP.2016.2609821
This is where the anechoic chamber parameters: the frequency and the electri-
Date of publication: 1 December 2016 becomes a viable solution. The anechoic cal size of the antenna under test (AUT).
Equation (2) can be used when the not support some indoor measurements selecting the range, it is important to
antenna is under one wavelength in at those low frequencies. Thus, there is think of the typical electrical size of
electrical size. Going back to (1), the far- a lower frequency boundary for indoor the AUT. When creating an anechoic
field distance can be plotted as a func- ranges. In general, for frequencies below chamber, the goal is to obtain a volume
tion of the electrical size of the antenna. 100 MHz, an outdoor range is a better in the chamber where any reflected
This is shown in Figure 2. As a rule of approach. Indoor ranges can be built, but energy from the walls of the range
thumb for indoor ranges, we can state the antenna size should be kept lower (ceiling and floor) will be much lower
that far-field illumination techniques than 2m , which limits the far-field distance than any of the features of interest on
are better suited for antenna sizes under to 8m (24 m). This distance is close to the radiation pattern. This volume is
ten wavelengths. This rule is related the 10m given by (2); Table 1 provides an what it is known as the quiet zone (QZ).
to the electrical antenna size. The fre- approximate guide for the different anten- Figure 1 shows that as one antenna
quency of operation adds another factor na sizes and frequencies of operation. transmits, it illuminates the receive
that will influence the type of range. The values in Table 1 are general antenna and all the walls and surfaces
Figure 2 shows that an antenna with a guidelines. Spherical near-field (SNF) of the range. As described in [1] and
size of 10m will have a far-field distance ranges can test antennas as small as [3], the energy incident on these sur-
of 200m , so the test distance is 20 times m/2 , but for such a small antenna, faces will be reflected toward the QZ.
the size of the antenna; at some micro- the better approach may be to use a The level of reflected energy must be
wave frequencies, this may be a test far-field illumination range. When a given number of decibels below the
distance of 200 in (5 m), so an indoor
range may be easy to implement. How-
1,000
ever, a 20m antenna will have a test
700
distance that is 800m . 500
For example, consider an 18-in dish
300
that is used by a popular satellite televi-
200
sion (TV) service. This satellite service
operates at 18.55 GHz. The dish antenna 100
is 28.29m in size. The far field is at ap 70
proximately 1600m or 25.86 m (84.84 ft). 50
r /
Reflectivity (dB)
the thickness. When the thickness of the 1.8
25 2
absorber is such that 0.25m # t # 2m, 2.5
then the coefficients of (4) are given by 30
3
the following polynomials: 35 4
5
40 6
A 1 ^ t h = 1.5252 - 4.8243t + 6.9479 t 2 7
- 3.8332 t 3 + 0.7333 t 4, (4a) 45 8
9
50 10
A 2 ^ t h = - 0.0754 + 0.24782t - 0.3984 t 2
15
+ 0.2285 - 0.0442 t 4, (4b) 55 20
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85
A 3 ^ t h = 0.0016 - 0.00502 t + 0.00938 t 2 Angle of Incidence ()
- 0.00577 t 3 + 0.001155 t 4,
(4c) Figure 4. The estimated reflectivity of an RF absorber as a function of an angle
of incidence.
A 4 ^ t h = - 1.58 $ 10 -5 + 4.91 $ 10 -5 t
- 1.015 $ 10 -4 t 2 + 6.58 $ 10 -5 t 3 polynomials. Hence, (4) is valid when and the polynomial results. However,
- 1.35 $ 10 -5 t 4, (4d) 0.25m # t # 20m, as stated previously. that null in the reflectivity may shift
The range of (4) should also be limited depending on the material in which
A 5 ^ t h = 5.84 $ 10 -8 - 1.78 $ 10 -7 t to 55 R(dB) 0. For an absorber of the absorber is mounted, as shown in
+ 4.02 $ 10 -7 t 2 - 2.71 $ 10 -7 t 3 thickness larger than 20m , the reflec- [10]. In general, the polynomials are
+ 5.7 $ 10 -8 t 4 . (4e) tivity can be approximated using the a safe approximation for the perfor-
results for a 20m -thick absorber. Fig- mance of RF materials at different
When the thickness of the treatment is ure4 shows the bistatic performance as angles of incidence.
such that 2m # t # 20m, then the coef- a function of the angle of incidence for a The largest typical absorber size cur-
ficients are given by the following set of series of different electrical thicknesses rently available is 72 in (1.82 m). As was
polynomials: of the absorber. Figure 5 shows a com- mentioned previously, this size provides
parison of computed results using the a frequency limit for the use of indoor
A 1 ^ t h = 0.1751 + 0.149 t - 0.0119 t 2 method in [9], given manufacturer speci- ranges. At 100 MHz, the thickness of
fications and the results from (4) for a this absorber is 1.64m . Its normal inci-
+ 0.00028 t 3, (4f)
material of thickness equal to m and 2m . dence performance is about 33 dB. In
A 2 ^ t h = - 0.0105 - 0.00824 t + 0.0007 t 2 If we compare the results of the an indoor range lined with this material
- 1.61 $ 10 -5 t 3, (4g) polynomials presented in this article pattern, features 20 dB from the peak
to those from numerical computa- will be difficult to measure accurately.
A 3 ^ t h = 0.00029 + 0.000123t tions, we can see that the polynomi- There are hybrid absorbers merging fer-
- 1.13 $ 10 -5 t 2 + 2.57 $ 10 -7 t 3, als provide a conservative number for rite tiles and lossy substrate pyramids of
(4h)
the reflectivity that may be higher wedges that operate down to 30 MHz
A 4 ^ t h = - 1.69 $ 10 -6 - 4.77 $ 10 -7 t by about 10 dB. The manufacturer or even 20 MHz. These types are more
+ 5.08 $ 10 -8 t 2 - 1.14 $ 10 -9 t 3, specifications were only provided suitable for electromagnetic compatibil-
(4i) from 45c to 80c and for a normal ity applications as their normal-incidence
A 5 ^ t h = 0. (4j) incidence. Computed results were absorption is typically limited to 25 to
obtained only at a few angles. For the 35 dB.
The domain of (4) is limited by those 1m -thick absorber, the different meth-
angles of incidence where 0c # i # 85c ods follow similar trends, with the Range Size
and where i = 0c is a normal inci- polynomials providing the most con- The discussion regarding sizing the
dence. Additionally, the domain is lim- servative number. There is a large dif- range begins with rectangular far-
ited by the domain of the coefficient ference at 35c between the computed field ranges, which have a test distance
25
W = ^2n 2 cot ^i h + 2t h m. (8)
30
Parameters i and t must be chosen to
35 obtain the required reflectivity. It is impor-
tant to check that the minimum 2m spac-
40
ing from the QZ to the absorber is kept.
45 The length of the rectangular far-field
chamber is mainly given by the far-field
50 distance and the QZ size plus the absorb-
er thickness and at least a spacing of 2m
55
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 between the QZ and the absorber. Added
Angle of Incidence () space should be included for the range
antenna and for the absorber behind it.
The total chamber or range length
Figure 5. A comparison of bistatic reflectivity from a computational approach,
manufacturers specifications, and (4). (L) is given by
L = ^2n 2 + n + 2 + t + K h m , (9)
is maintained. This minimum spacing
avoids loading of the AUT. The far-field where K is a factor large enough to
distance is given by include the source antenna, the spacing
to the absorber, and the absorber behind
r = 2n 2 ,(5) the source. It should be noted that these
x m
equations provide a minimum require-
n where n is the number of wavelengths ment. Work must be performed inside
2n 2
in the size of the AUT. The QZ must the chamber, mounting and connect-
be large enough to encompass the AUT. ing the antenna, switching range anten-
Figure 6. The geometry of a far-field Hence, the QZ is nm . Figure 6 shows a nas, etc. The space should be checked to
range. typical rectangular range geometry. allow for people to perform these tasks
From the geometry in Figure 6, we inside the anechoic range.
determined by (1). Following the discus- can derive an equation for the distance x. Let us enter some values into the
sion on far-field ranges, CRs and near- The distance x is the distance from the prior equations to observe the expected
field ranges will be discussed. range centerline to the absorber tips. chamber sizes. It will be assumed that
the source antenna is directive and that
FAR-FIELD CHAMBERS x = n 2 $ cot i.(6) it has a sufficient front-to-back ratio. The
m
It is common to find sources stating the absorber behind the source antenna will
rules of thumb for sizing a rectangular Equation (6) gives the distance in terms be one wavelength in thickness, and the
anechoic chamber for far-field illumi- of wavelengths. If we look at Figure 3, the factor K will be set to four. In Figure 7,
nation; these can be found in [1]. The value of i can be chosen to give a desired the width and length of a series of rect-
general rule is that the width and height reflectivity. The curves in Figure 4 angular chambers have been plotted ver-
of an anechoic chamber should be three will also provide a value for the thickness sus the lowest frequency of operation. In
times the diameter of the minimum of the absorber. Hence, if the AUT has addition, the electrical size of the AUT at
sphere that contains the largest antenna features that need to be measured in the the lowest frequency is indicated by the
being tested. It is important to check 25 dB level, then the bistatic reflectivity value of n for each plotted chamber size.
that a minimum spacing of 2 m between of the absorber must exceed that level. From Figure 7, there are some con-
the AUT and the tips of the absorber An absorber 2m in thickness will exceed clusions that can be drawn. The first is
4 Absorber
Trimmed Wedge
12 ft
QZ 3 ft
12 ft
6 ft 3 ft 3 ft
LFB
24 ft
12 ft
2-10
7-0 26-0 17-4 MI-644-M-013 PiT
Vertex to Focal Length Focal to
Wall CL of QZ
74-0
Inside Walls
2-GHz Reflector Pattern Approximate
90
120 60
150 30
180 43-0
60 40 20 0 Inside Walls
150 30
120 60
90
Figure 10. A typical CR layout with the reflector pattern superimposed and showing that the energy incident on the side
walls, floor, and ceiling is already more than 40 dB down at 2 GHz.
Y (m)
1
0
1 1
1 90
2 60
0.5 0.5
3
4 30
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5
X (m) X (m)
Pulse Transmitted Pulse Transmitted 0
0
0
0
4
2
6
Y (m)
1
0
1 1 1
0.5 0.5 2
3
4
0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 5
X (m) X (m)
Figure 11. Radiation from the reflector compared to the far-field pattern.
0
ond near-field measurement approach is Point
5
1 0
cylindrical scanning. In this technique, 150
2 5
0
5
the field is measured on the surface of
2
30
a cylinder excluding the top and bottom Vertex
surfaces. This technique is ideal for long 180
60
antennas that are omnidirectional or
150
have a wide beam on one of the principal 90 120
planes but a narrow beam in the per-
pendicular plane. The last near-field and
more general measurement approach is Figure 12. The absorber on the floor between the feed positioner and the reflector
spherical scanning. In this technique, the is critical to reduce the reflected energy from illuminating the reflector.
field is measured on a sphere that con-
tains the entire antenna being measured. This is a rough approximation. For used in SNF, hp should include the
In general, the test distance for planar both (16) and (17), an additional 1 m height of the floor slide, the azimuth
near-field measurements is between 3m minimum should be added to prevent positioner, and the offset slide. The
and 10m . For SNF, however, the probe the positioning equipment from hitting reader should be aware that we have not
can be further away. The SNF is dis- the probe as it rotates the antenna being taken into account the positioning equip-
cussed in detail in [18]. measured. Also, this space should provide ment in the far-field chamber equations
The same equations developed for room for people to perform work inside or the CR equations (except for the feed
far-field chambers can be used for SNF the chamber during the set up of the positioning). The reason being that, in
with the exception of equations for the measurement. This is more critical for those types of ranges, other dimensions
test distance. In general, the equation is higher frequencies (above 2 GHz) where are so dominant (the far-field test dis-
given by the 4m separation may not be enough for tance or the reflector size) that the posi-
the positioner to clear the probe. tioner is not an issue.
L = d pp + ^ n + 6 + 2t e h m ,(16) We notice that the angle of incidence PNF systems use a planar scanner
on the side absorber is given by to measure highly directive antennas
where d pp is the depth of the probe (i.e., gain > 20 dB). The high gain
(measuring antenna) and its positioner. i = arctan c
4n + 16 .(18) of the AUT benefits in the design
m
2n + 16
The variable n is the diameter in of the range, as some regions of the
wavelengths of the minimum sphere Taking the limit as n " 3, it follows that range (those behind the AUT) do not
that contains the AUT. The absorber i 1 63.4c . Using Figure 4, we can esti- need to be treated with an absorber.
on the two end walls will have a mate that t s . 2t e . To do this, we check The test distance, as stated above, is
thickness of t e m , where t e is the the reflectivity of the end wall absorber at between three and ten wavelengths.
thickness, in wavelengths, of the end the normal incidence, and we select the The dominant factor for sizing a PNF
wall absorber. As customary, 2m is thickness of the absorber that will provide range is the scanner. The scan size is
added between the minimum sphere similar reflectivity for the 63.4 incident given by
and the absorber tips, and, finally, we angle. The ceiling and the floor will have
estimate 4m as the distance between the same absorber as the side walls. L x = (n + 2k tan ^i s h) m ,(20)
the probe and the sphere containing The chamber height can be estimat-
the antenna. The width of the SNF ed using where i s is the maximum angle for an
chamber is given by accurate far field, and nm is the elec-
H = h p + ^ n + 4 + t sh m.(19) trical size of the antenna being tested
W = ^ n + 4 + 2t s h m ,(17) (see Figure 13). The variable k is the
The variable hp accounts for the height test distance in wavelengths, hence,
where, in this case, t s is the thickness of the positioning equipment. In a typi- 3 1 k 1 10. The physical scanner will
in wavelengths of the side wall absorber. cal rollover azimuth positioner that is usually be slightly larger than the scan
7.5 Conclusions
The goal of this article is to provide
an overview of the rules and physics
5
that guide the selection and sizing of
indoor anechoic chambers. All of the
2.5 equations provided in this article are
30 35 40 45 50 55 60 65 70 75 80 approximations. The chamber length
s Maximum Angle for Accurate Far Field () calculated using (9), (13), (16), (20), and
(23) is a minimum, and more space
Figure 14. The angles of incidence on the absorber as a function of the maximum may be required to help with the load-
angle for accurate far-field pattern. Plotted for different test distances. ing and unloading of the device-under
Catania, Italy, since 2014. His research Trans. Antennas Propag, vol. 62, no. 5, pp. 2407 [12] L. Kosuru and D. Deavours, Optimum per-
interests include the development of 2417, Feb. 2014. formance for RFID tag immersed in dielectric
[6] D. Pinchera, M. D. Migliore, and F. Schet- media, in Proc. IEEE Int. Conf. Radio-Frequency
ultra-wideband compact antennas, tino, An ultra wide permittivity antenna (UWPA) Identification, Orlando, FL, 2011, pp. 1118.
microwave devices, and computational for reliable through-wall communications, IEEE [13] M. Deshpande and M. Bailey, Input imped-
electromagnetism, with particular interest Trans. Antennas Propag., vol. 61, no. 2, pp. 957 ance of microstrip antennas, IEEE Trans. Anten-
960, Oct. 2012. nas Propag., vol. 30, no. 4, pp. 645650, July 1982.
in radio-frequency plasma interactions and [14] C. A. Balanis, Antenna Theory: Analysis and
[7] M. Rad and L. Shafai, Embedded microstrip
particle accelerators. patch antenna for structural health monitoring appli- Design. Hoboken, NJ: Wiley, 2012.
cations, presented at the IEEE Antennas and Prop- [15] R. Garg and S. Long, Resonant frequency of
agation Society Int. Symp., San Diego, CA, July 2008. electrically thick rectangular microstrip antennas,
REFERENCES [8] X. Jin and M. Ali, Reflection and transmission Electron. Lett., vol. 23, no. 21, pp. 11491151,
[1] H. Stockman, Communication by means of properties of embedded dipoles and PIFAs inside July 1987.
reflected power, Proc. Inst. Radio Engineers, vol. 36, concrete at 915 MHz, presented at the IEEE [16] R. Sorrentino and G. Bianchi, Microwave and
no. 10, pp. 11961204, 1948. Antennas and Propagation Society Int. Symp., RF Engineering. Hoboken, NJ: Wiley, 2010.
[2] G. S. Mauro, G. Castorina, A. F. Morabito, L. Charleston, SC, June 2008. [17] F. Abboud, J. Damiano, and A. Papiernik,
Di Donato, and G. Sorbello, Effects of lossy back- Simple model for the input impedance of coax-fed
[9] S.-H. Jeong and H.-W. Son, UHF RFID tag
ground and rebars on antennas embedded in concrete rectangular microstrip patch antenna for CAD,
antenna for embedded use in a concrete floor,
structures, Microwave Opt. Technol. Lett., vol. 58, IEE Proc. H Microwaves Antennas Propagat., vol.
IEEE Antennas Wireless Propag. Lett., vol. 10, pp.
no. 11, Nov. 2016. 135, no. 5., pp. 323326, Nov. 1988.
11581161, Oct. 2011.
[3] R. W. P. King, G. S. Smith, M. Owens, and T. T. [18] I. Wolff and N. Knoppik, Rectangular and
[10] S. Jiang and S. V. Georgakopoulos, Opti-
Wu, Antennas in Matter: Fundamentals, Theory, and circular microstrip disk capacitors and resonators,
mum wireless power transmission through rein-
Applications. Cambridge, MA: MIT Press, 1981. IEEE Trans. Microw. Theory Techn., vol. 22, no. 10,
forced concrete structure, in Proc. IEEE Int.
pp. 857864, Oct. 1974.
[4] H.-Y. Liang, H.-C. Yang, and J. Zhang, A cylin- Conf. Radio-Frequency Identification, Orlando,
[19] J. J. Wang, Y. P. Zhang, K. M. Chua, and A. C.
drical conformal directional monopole antenna FL, 2011, pp. 5056.
W. Lu, Circuit model of microstrip patch antenna
for borehole radar application, IEEE Antennas [11] C. Thajudeen, A. Hoorfar, F. Ahmad, and on ceramic land grid array package for antenna-
Wireless Propag. Lett., vol. 11, pp. 15251528, T. Dogaru, Measured complex permittivity of chip codesign of highly integrated RF transceiv-
Dec. 2012. walls with different hydration levels and the effect ers, IEEE Trans. Antennas Propag., vol. 53, no. 12,
[5] C. Liu, Y. Guo, and S. Xiao, Capacitively load- on power estimation of TWRI target returns, pp. 38773883, Dec. 2005.
ed circularly polarized implantable patch antenna Progress in Electromagnetics Res. B, vol. 30, pp.
for ISM-band biomedical applications, IEEE 177199, Jan. 2011.