Ground Systems for HF Verticals some experimental comparisons to NEC'

1
Ground Systems for HF Verticals some
experimental comparisons to NEC.

• Rudy Severns N6LF
• antennasbyn6lf.com

2
Some typical questions on verticals

• How much of ground system is it worth putting
  down?
• What will I gain (in dB!) by adding more
  radials?
• Does it matter if I lay the radials on the ground
  surface?
• Are a few long radials useful?
• Are four elevated radials really as good as lots
  of buried radials?
• How well do gullwing elevated radials work?

3

• We can use modeling or calculations to answer
  these questions but most people dont have a lot
  confidence in mathematical exercises.
• High quality field measurements on real antennas
  are more likely to be believed.
• Over the past year I have done a series of
  experiments on HF verticals with different ground
  systems.
• That is the subject of todays talk.

4
Comment

• Todays talk is a snapshot of experimental work.
• The talk will only cover the highlights.
• A detailed summary of the test range and
  instrumentation along with reports on each
  experiment can be found on my web page
  antennasbyn6lf.com .
• A copy of this PowerPoint presentation will also
  be on the web site.
• You may also see other interesting information on
  the web page.

5

• Whats the purpose of the ground system?
• Its there to reduce the power absorbed by the
  soil close to the antenna (within a ¼-wave or
  so).
• The ground system increases your signal by
  reducing the power dissipated in the soil and
  maximizing the radiated power.
• Any practical ground system will not affect the
  radiation angle or far-field pattern!

6
Power transmission
antenna 1
antenna 2
antenna equivalent circuit
7
Measurement schemes

• The classical technique is to excite the test
  antenna with a known power and measure the
  resulting signal strength at some point in the
  far field (gt2.5 wavelengths for 1/4-wave
  vertical).
• This approach takes great care and good equipment
  to make accurate measurements.

8
S21

• The modern alternative is to use a vector network
  analyzer (VNA) in the transmission mode.
• This approach is capable of reliable measurements
  to lt0.1 dB.
• The VNA will also give you the input impedance of
  the antenna at the feed-point.

rx antenna
test antenna
9
Some experimental results
10

• The first experiment was a 160 m, ¼-wave wire
  vertical with two ground stakes and 4 to 64
  radials.
• Measurements were made with a spectrum analyzer
  as the receiver.

11
Test Results
delta gain 2.4 dB
12
A new antenna test range on 40 m
13
Antenna under test
14
Test antenna with sliding height base
15
Adding radials to the base
16
Elevated radials
17
Elevated radials close-up
18
Loop receiving antenna
19
Receiving antenna at 40
N7MQ holding up the mast!
20
Network analyzers
note, automatic, organic, heating system
Homebrew N2PK
HP3577A with S-box
21
Inside the N2PK VNA
22
Test antennas

• A 1/4-wave 40m tubing vertical.
• An 1/8-wave 40m tubing vertical with top loading.
  
• An 1/8-wave 40m tubing vertical resonated with a
  base inductor.
• A 40 m Hamstick mobile whip.
• 40m SteppIR vertical

23
1/8-wave, top-loaded, 40 m vertical
24
What about a few elevated radials versus a large
number of surface radials?
25
NEC modeling prediction
26
NEC predictions

• There will be a very rapid change in peak gain as
  we raise the base of the antenna and the radials
  above ground.
• Lifting the radials only a few inches makes a
  substantial difference.
• When the base of the antenna and the radials have
  been elevated several feet, the peak signal will
  be very close to that for a large number of
  buried radials.

27
Experiment 3

• I began with sixty four 33 wire radials lying on
  the ground surface.
• The length of the vertical was adjusted to be
  resonant at 7.2 MHz.
• I removed the radials in the sequence 64, 32, 16,
  8, 4, measuring S21 as I went.
• With only 4 radials left I then raised the
  radials and the base of the antenna above ground
  incrementally measuring S21 at each height.
• There were no ground stakes and the feedline was
  isolated with a choke.

28
4-64 radials lying on ground surface
5.8 dB
29
4 radials raised above ground
5.9 dB
30

• NEC modeling predicts that four elevated radials
  will perform as well as 64 radials lying on the
  ground.
• In this example, measurements show no significant
  difference in signal strength between 64 radials
  lying on the ground and 4 radials at 4!

31
Some more elevated radial experiments
32
Gullwing radials a la N6BV
33
Variations in elevated radials
34
comment on four elevated radials

• From these experiments and NEC predictions it
  would seem that four elevated radials are all you
  need.
• Thats deceiving! Antennas with only a few
  elevated radials suffer from a number of
  problems
• hi-Q, radials tune the vertical
• asymmetric currents in the radials leading to
  pattern asymmetry.
• tuning and current symmetry are very sensitive to
  ground and mechanical variations as well as
  nearby conductors.

35
More on elevated radials

• Use more than 4 elevated radials
• the Q and radial current asymmetries decrease.
• tuning is less sensitive
• the reactive part of the feed-point impedance
  changes more slowly as you add radials so you
  have a better SWR bandwidth.
• however, the ground loss does not improve much.

36
Some experiments with radials lying on the ground
surface
37
Measured improvement over a single ground stake
f7.2 MHz
38
Caution!

• Your mileage may vary!
• My soil is pretty good but for poorer soils
  expect more improvement with more radials.
• The degree of improvement will also depend on the
  frequency
• soil characteristics change with frequency,
• at a given distance in wavelengths the field
  intensity increases with frequency.

39
Measured base impedances
40
Antenna resonance versus radial number
41
Radial current for different heights
42
A current sensor
43
Radial current measurements
44
Measured current distribution on a radial
45
Radial current distribution
46
NEC modeling prediction
47

• Lets do an experiment
• isolate the base of the antenna with a common
  mode choke (a balun).
• lay out sixty four 33 radials and adjust the
  vertical height to resonance (reference height).
• remove all but four of the radials
• Measure S21 with the reference height.
• Measure S21 with the vertical shortened to
  re-resonate.
• Measure S21 with the reference height as we
  shorten the radials.

48
Effect of shorting radials, constant height
49
Radial current distribution
50
The lesson here!

• When you have only a few radials lying on the
  ground you can have much higher losses than
  expected!
• These losses can be reduced by shortening the
  radial lengths, i.e. less copper less loss.

51
Practical example Field day scenario

• You want a 40 m vertical for field day.
• ¼-wave 33. So you start with about 33 of
  aluminum tubing for the radiator and four 33
  wire radials.
• You erect this, with the radials lying on the
  ground and its resonant well below the band!
• What to do?
• Nothing, use a tuner and move on,
• Shorten vertical until its resonant,
• add more radials
• or, shorten the radials until the antenna is
  resonant.
• Which is best?

52
Direct measurement of several options

• Do nothing G 0 dB
• Shorten height G-0.8 dB
• Shorten radials G3.5 dB
• Use 16 radials G4 dB
• Use 64 radials G5.9 dB

53
Another experiment
54
An observation

• When you have only four radials the test results
  are always a bit squirrelly
• small variations in radial layout,
• coupling to other conductors,
• like the feed-line,
• all effect the measurements making close
  repeatability difficult between experiments.
• The whole system is very sensitive!
• This nonsense goes away as the number of radials
  increases!

55
Summary

• Sparse radial screens (less than 16 radials) can
  have a number of problems
• increased loss with longer radials
• unequal current distributions between radials.
• system resonance shifts.
• A few long radials can be worse than shorter
  ones.
• screen resonances can alter the radiation pattern
  as the radials begin to radiate substantially.

56
Summary continued

• Try to use at least 8 radials but 16 is better.
• The more radials you use, the longer they can be.
• A number of 1/8-wave radials will be better than
  half that number of ¼-wave radials. At least
  until you have 32 or more radials.
• In elevated systems
• try to use at least 8 radials
• you can use radials shorter than ¼-wave and
  either re-resonate with a small L or make the
  vertical taller or add some top loading.
• the gullwing geometry can work.