Modelling Performance of Small Antennas for 3999 khz Portable Operation
by Alan Biocca WB6ZQZ 12/3/2002 (updated 1/19/03)
The following article will present and discuss results of modelling various configurations of small antennas for 3999 khz that are, or might be used for Pedestrian Mobile operation (see http://www.hfpack.com for more info on this
popular and growing aspect of Ham Radio).
This article is being posted incrementally, this document will grow as sections are inserted.
I have already analyzed many antennas, but it will take some time to collect the details into
this document from my notes.
If you have any feedback for this please send it to me via wb6zqz at qsl.net.
In this section several antennas will be covered, plus some variations. Many of the basic configurations are very similar in that they are based on a whip vertical with the base at 5 feet above ground, and a drooping counterpoise that begins at this 5 foot height, angles down toward the ground at 45 degrees, and then hovers at 0.1 foot above the ground out horizontally to the 50 foot point. It does not lie on the ground because Eznec would connect it to ground, some spacing is required. The effect of this spacing should be small, think of it as thick insulation on the counterpoise.
Default Modelling Conditions
Unless otherwise specified:
The ground conditions are High Accuracy Medium Ground.
Losses are Copper.
Measurements are in DBi, or decibels over an isotropic antenna.
The gain given for an antenna is the gain of the peak lobe.
Note that most of these antennas have less peak
field strength than an isotropic antenna, so they will be expressed in negative DBi.
The more negative the value, the less effective the antenna.
The antenna loads (if any) were adjusted to resonance
(near zero reactive component at feedpoint).
Load coil default Q is 500.
Tuner coil Q default is 200.
Tuner capacitor Q default is 1000.
The Fishpoles are www.cabelas.com 14 foot telescoping crappie poles.
For the dipole, a 4' tube is assumed to slide the pole handles into, making the assembly 31 feet.
For the verticals the rod handle is fixed to the packframe making the length
available for the antenna a bit less - I used 12 feet.
The antennas are:
- A1) Full quarterwave whip (vertical monopole) with 52' counterpoise
- A2) 12 foot tall whip fed by antenna tuner
- A3) 12 foot whip with base loading coil
- A4) 12 foot whip with loading coil 4 feet up
- A5) 18 foot whip with coil 6 feet up
- A5b) 31 foot whip with coil 6 feet up
- A6) horizontal fishing-pole loaded dipole 31' long
- A6b) Fishing pole vertical loaded dipole on 16' mast
- A7) small loop 4' square
The software used to perform these calculations will be discussed at the end of the article.
Just to remind us, here is a DB scale refresher:
- 1db - just discernible change
- 2db - min useful improvement. small change
- 3db - power doubling. 5w to 10w. small improvement. half an s unit
- 6db - quadruple power. 1w to 4w. big change. one s unit
- 10db - 10x power. 100w to 1kw. major improvement. s9 to s9+10.
- 20db - 100x power. 1w to 100w. huge. s9 to s9+20.
- 30db - 1000x power. 1w to 1kw. really big change. s4 to s9
Radiation Pattern Overview
The vertical whips have directivity favoring the direction of the counterpoise.
There is a null in the opposite direction.
The depth of the null varies from 10 to 25db down from the peak lobe,
depending on the vertical angle.
The favored signal is in the direction of the countepoise when over real ground.
Over more conductive surfaces the directionality changes (see details below).
All the vertical patterns are substantially similar, the most variation is with the full quarterwave.
The dipole (and loop) patterns are more symmetrical, with the best gain straight up (Near Vertical Incident Skywave style).
Gain Summary Table
||Fullsize Quarterwave Vertical
||Whip and tuner, not including tuner loss (1-15db) or lineloss (47.32 - J 1330)(at 10w 600v is required)
||using 2' rg58 tuner to antenna (swr 426) 0.023db line loss but tunerloss rises to 2+ db
||using 100' rg58 tuner to antenna 18.5db line loss
||Base loaded whip. Q500 Coil at the base, tuned to resonance.
||Loaded whip, load up 4 ft (Q500, top #12,#20 cp 52' #14 at 0.1')
||Loaded whip in Free Space (Z 14.5 -j 153) peaks at 4 deg above horizon, main lobe perpendicular to wires
||Loaded whip over Perfect Ground (Z 7.9 -j89) peak gain at horizon
||Loaded whip over Saltwater (Z 10.6 - J17.4) peak gain at 10 deg
||Loaded whip raised counterpoise to 0.3'
||Loaded whip raised counterpoise to 5' above the ground over the whole length
||Loaded whip with Lighter wire counterpoise #14 to #20
||Loaded whip with Heavy top wire - #20 changed to #14
||Loaded whip with Lower Q coil 200
||Loaded whip with even lower Lower Q coil 100
||Taller vertical 18' (sd-20 lashed to packframe) Q500 loading coil up 6'
||Taller yet vertical 31' (mfj-33 lashed to packframe) Q500 loading coil up 6'
||Vertical Dipole 17' hi with inverted T counterpoise (no drag wire)
(includes load loss of 3db)
||Fullsize Halfwave Dipole up 8' (peak lobe straight up)
||Halfwave dipole up 16'
||Halfwave dipole up 25'
||Fishpole loaded Dipole 31' long up 16' (lobe straight up, -7dbi at 45 degrees)
||Fishpole loaded dipole lowered to 8'
||Fishpole loaded dipole raised to 25'
||Fishpole vertical loaded dipole (135 deg) on 16' mast peak at 30 deg
||Fishpole vertical loaded dipole on 25' mast peak at 30 deg
||Fishpole dipole 45 degrees Sloper up 16' peak at 34 deg
||Smaller Loaded Dipole, 18 foot, big Q500 coils up 8'
||Small Loop, 1 turn #14 4'x4' base 12' high (0.5 + j 160)
||Small 1 turn Loop, change wire from #14 to #12
||loop, 2 turns #12 (1.1 + j 410)
||loop, 2 turns 1" aluminum (0.18 + j270)(2kv 7.5a 10w) (plus -5.3db loss in default tuner)
Observations and Conclusions
- Still Forming. Mostly these are well-known effects.
- Loading coils are better than tuners, but a good tuner may not be too bad.
Experience indicates the commercial autotuners are not so good at high mismatches (such as these are).
Experience with commercial auto-tuners indicate perhaps a two S-unit loss compared to a loading coil, so choose the tuner carefully. It may be necessary to build the tuner to achieve low loss. While I did not exhaustively try all tuner configurations on all impedances, it was apparent that the pi configuration produced higher loss than the lowpass L-C or C-L configurations. Since harmonic suppression is a good idea, and since it appeared to have low loss I concentrated on that configuration.
- Moving the coil up the vertical helped, but not a great deal (less than 1db). Note that raising the coil can have other benefits. The section above the coil is much more sensitive to changes in capacitance, so you to increase the stability of the system tuning it may be important to have the coil up higher. The voltages above the coil are also much higher than below, a possible safety issue.
- Increasing the height of the vertical portion from 12 to 18 feet helped 1db.
Adding 6 feet to a 64 foot system (including counterpoise) is a small change.
- Using smaller (lossier) counterpoise wire helped the gain (0.27 db).
- Raising antennas to 25 feet did seem to help a lot for the horizontals (3-6db).
It did not help much for the verticals (1db).
- Saltwater was worth 6db peak lobe gain for the verticals.
Note that at low angles it can be a lot more.
- Shortening the dipole from 31 feet to 18 feet cost 5db.
An 18 foot dipole would be
similar to adding 80m coils to a commercial Buddipole.
At this frequency the short dipole
does not have enough radiation resistance and more power is lost into the loads -
even though they are low-loss Q500 coils.
The 31 foot dipole I envisioned is a pair of 14 foot telescoping fishing poles
(6 ounces each), and a center section of a pair of 2' aluminum tubes and a
fiberglass insulating tube in the center. The idea is to make a center section
that will collapse into 2' lengths. The tubes are selected to fit the fishing rod
handles (slide them in 6") and become the center portion of the dipole. The
fiberglass tube fits inside and forms the center insulator. (Note that I have
not yet selected the tubes or constructed this antenna).
- Ultimately, depending on the circumstance, it may be desirable to have the
selecting between horizontal, sloping, or vertical antenna configurations.
The horizontal optimizes the gain
straight up, good for local communications.
The Vertical optimizes low angle and great
distance contacts. Note however that low angle absorption is significant on 80
meters, and longer distance contacts are often the result of multi-hop higher angle
paths rather than the very low angles (different than 20 meters and up).
The sloper splits the difference, favoring one direction.
A flexible kit that allows these choices in the field might be the best approach.
A pair of 14 foot fishing poles (6 ounces each) plus some wires and
low loss coils, and a flexible mounting/supporting system for the poles would make
a good system (this was suggested to me by Oliver KB6BA a few months back).
- A tall whip can be made by using military antenna parts or using telescoping
fishing poles from Cabelas, MFJ, WorldRadio and others. A conductor will be needed with
the fishing poles. The assumption here was that the first couple of feet of the
pole were lost to the mount on a packframe, the rest was used for the vertical.
- Thanks to many who provided feedback and comments have been helpful in the
preparation of this note including that from Bonnie KQ6XA, the organizer of hfpack;
Cortland KA5S with loop antenna comments; and others. The 31 foot vertical was
inspired by Harry W6DXO's MFJ pole and loading coil on pipe insulation. The
18 foot horizontal dipole was inspired by Budd W3FF's Buddipole antenna.
Software Tools and Techniques
The antenna modelling was performed using Eznec 3.0.
The transmission line and tuner analyses was done with Transmission Lines for Windows,
included with the ARRL Antenna book.
About the Author
Alan Biocca has been a licensed Amateur Radio operator since 1967. He currently holds
the Extra class license and has been modelling antennas for several years to gain
understand of the real ones he has been building and using for 35 years. In his
work life he holds BSEE and MSCS degrees from UC Berkeley and leads the
Control Systems Group for the Advanced Light Source at Lawrence Berkeley National Laboratory.
Search the Net for further info
WB6ZQZ Home Page