SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
S. K. Padhi, M. W. Waterson, Budi Juswardy, Aziz Jiwani, Adrian Sutinjo and Peter Hall
ICRAR/CIRA, Curtin University, Australia
Do we need ground plane for
SKA-low Radio Telescope?
SKA Technical Progress Meeting, October 22-23, 2012, Bologna, Italy
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Outlines:
Introduction
Soil characterization
Measurements:
Ground plane option:
Sensitivity with ground plane
Antenna noise calculation with mesh ground
Antenna noise on dispersive soil
Antenna Measurements:
Antenna performance
Conclusions
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Introduction:
― Soil modelling
― Very important to understand the environmental influence on
antenna performance.
― Accurate characterization sensitivity of antenna (array)
― Helpful in calculation of power budget
― Ground plane choice
― Cost
― System level performances
― Mesh-grid design (for improved performance)
― Sensitivity of antenna
― Antenna metrology
― Challenges in fabrication of Spiral antenna fabrication.
― measurements
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Soil Characterization
― We classify the morphology of soil as two
layers structure.
― Upper layer: mixture of gravels, sands, silt
and clay.
― Lower layer is granite base
― Thickness of upper layer: 30 cm to 80 cm.
― 5 samples were collected during sunny day
― 4 samples were taken from each corner of
200 m length square area.
― The 5th one collected from the centre of
square.
― The samples were sieved and oven dried
before measurements.
Lower layer
Upper layer
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Soil Measurements:
• Chemical analysis + mineral composition
• Measured relative permittivity of the soil
samples, in CSIRO facility.
• Permittivity measurements with different
moisture contents.
• Nicholson-Ross-Weir algorithm is used in
the calculation of permittivity.
• Debye 2 parameter model is used for
fitting data (model).
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
0.0 0.2 0.4 0.6 0.8 1.00
2
4
6
8
10
Imag(r):
measured
fitted
Real(r):
measured
fitted
Per
mit
tivi
ty (R
e/Im
)
Freq (GHz)
2 % moisture contents:
Soil: measurements:
0.0 0.2 0.4 0.6 0.8 1.0 1.210
20
30
40
Soil data(10% moisture):
' meas.
' model
'' meas.
'' model
Freq (GHz)
' (mea
s. a
nd m
odel
)
1
10
100
'' (mea
s. a
nd m
odel
)
0.0 0.2 0.4 0.6 0.8 1.0 1.2
2.0
2.5
3.0
3.5
4.0
4.5
5.0
5.5
6.0
Measurements:
real(r)
imag(r)
Freq(GHz)
Per
mit
tiv
ity
(re
al)
0.0
0.5
1.0
1.5
2.0
Per
mit
tivit
y (
imag
)
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Do we need a ground plane??
― Ground planes are critical components of antenna.
― GP influence (depends on various types of antenna)
― Impedance match
― Gain/sensitivity of antenna ― GP needed for mechanical support or robust placement/fixing
antenna
― Proper grounding of circuits etc..
― Cost is important for GP size ??
― Alternate option:
― Mesh grid (reducing cost (and improve performance))
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Mesh grid model:
d
Infinite soil box
(A,B,C)
Phase center
GP/mesh grid
Soil C (infinite size)
Mesh size:
Pitch: 300 mm
Dia: 2.5 mm
Size of grid: 10m x 10m
Media: copper
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Noise Temperature (Tmg+soil) VS pitch size:
50 100 150 200 250 300 350 400 4500
50
100
150
200
250
Tan
t_gn
d [o K]
Freq(MHz)
Meshgrid + Soil C (d=0.0 m)
300 mm
600 mm
900 mm
1200 mm
spiral LPDA
50 100 150 200 250 300 350 400 4500
50
100
150
200
250Mesh grid+Soil A (d=0.1 m):
300 mm
600 mm
900 mm
1200 mm
Tan
t_gn
d [o K]
Freq (MHz)
50 100 150 200 250 300 350 400 4500
50
100
150
200
250
Vivaldi (meshgrid+Soil A): (d=0.1 m)
300 mm
600 mm
900 mm
1200 mm
T ant_
gnd [o K
]
Freq (MHz)
vivaldi
~100 MHz
~150 MHz
~150 MHz
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Sensitivity with/without mesh grid
LPDA
50 100 150 200 250 300 350 400 450-160
-120
-80
-40
0
20
Vivaldi (mesh only): d=0.1 m
300 mm
600 mm
900 mm
1200 mm
T
ant_
gnd[o K
]
Freq (MHz)
50 100 150 200 250 300 350 400 450-100
-80
-60
-40
-20
0
10
T
ant_
gnd[o K
]
Freq (MHz)
Meshgrid only:(d=0.1 m)
300 mm
600 mm
900 mm
1200 mm
50 100 150 200 250 300 350 400 450-160
-140
-120
-100
-80
-60
-40
-20
0
LPD (meshgrid only: d=0.1 m):
300 mm
600 mm
900 mm
1200 mm
T
ant_
gnd[o K
]
Freq (MHz)
spiral
vivaldi
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Comparison of Tant_gnd (Soil-A)
Freq. Spiral Antenna Vivaldi Antenna LPD Antenna**
ASR-V1.4 FEKO ASR-V1.4 FEKO ASR-V1.4 FEKO
70 MHz 134.0 153.1 209.8 223.5 127.5 198.7
150 MHz 2.12 8.53 65.2 85.0 12.7 27.0
300 MHz 4.00 8.92 40.6 54.5 na 8.1
450 MHz 1.2 7.5 29.2 42.3 0.8 8.25
**: compared with d=0.16 cm data
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Do we need a ground plane? Sensitivity of isolated sheet spiral at SKA site – infinite extent back surfaces
Spiral and ground plane 11
• Spiral with a metallic ground plane has higher sensitivity than SKA-low goal
• Spiral over the soil does not meet goal below 110 MHz, but is close enough to warrant detailed performance and cost trade-
off
• Highly-directive SKA-low antennas (like the conical spiral) may not need a metallic ground plane
• Or use only very wide mesh
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Antenna metrology:
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Conductor patterns:
- 2 identical spiral racks made from 13 sections - Water-jet cut from brass mesh (or copper foil) - low-temperature solder joints, epoxy to cone
SKA Technical Progress Meeting: October 22-23, 2012, Bologna, Italy
Conclusions
― Electrical and chemical parameters of Soil are characterized in
details.
― Debye model is used to fit the measured data. The comparison
between the two are reasonable close.
― Ground plane effects are characterized for 3 antennas using MoM
solver.
― Mesh grid might be a better option and cost effective compared to
metallic ground plane. Tant is sensitive to lower end only .
― And for high requency end (> 150MHz) Tant is insensitive to mesh
grid/GP.
― With Meshgrid Tant is reduced to ~100 K (@70 MHz) for all three
antennas.