October 1966 QST
Wax nostalgic about and learn from the history of early electronics. See articles
QST, published December 1915 - present (visit ARRL
for info). All copyrights hereby acknowledged.
Part I of this article appeared in last
month's (September 1966) edition of QST, which explains why Fig. 3 is
the first one in this article. It introduced concepts in antenna
types and siting. This second part talks about cost tradeoffs for various aspects
of a DX setup. Author Paul Rockwell does a nice job of providing graphs of cost
versus performance increases for transmitter power, antenna gain, tower height and
constructions, etc. He uses prices typical of the mid 1960s, but even without knowing
the equivalent modern day equipment prices, the shapes of the curves are good indicators
of where the point of diminishing returns exists.
In 1966, Paul Rockwell wrote a 4-part
series for the ARRL's QST magazine on station design for long distance
communications (DX) that covered antenna selection and siting (Part I), economics
and construction (Part II), Station Configuration and Receiver Topics (Part III), and
Propagation Quirks and Operating Tips (Part IV).
Part II- Economics of Station Design and Construction
By Paul D. Rockwell, W3AFM
Fig. 3 - Transmitter power costs.
In the pursuit of amateur radio, dollar limitations are always present. What
is the most practical allocation of available funds? Let us first illustrate an
analytical - approach to this question from the standpoint of effective DX-radiated
power (DX e.r.p.). Assume 20-meter operation, flat terrain, no voice modulator,
and optimum radiation angle of 1°. As a frame of reference, 0-db. will be taken
for 100 watts c.w. input, 30-foot tower height, and 10-foot Yagi boom-length.
Transmitter power costs run about as shown on Fig. 3. The costs include
driver, but exclude v.f.o. The curve would have to be shaded upward for first-class
features, and downward for some home constructors. Make your own curve, if you prefer.
What is being shown on this and the following curves is a design technique - not
a universally applicable set of data. What is important to note, in this example,
is that the last db. (from 780 to 1000 watts) costs $100.
Now consider antenna costs. Fig. 4 presents these for Yagis. The db. gain
values are relative to a half-wave dipole, same height and foreground. The next
db. beyond 30-foot boom length costs $200. Stacking two beams, which gives 3-db.
gain at the expense of 40-foot additional tower height is attractive beyond about
30-foot boom length. However, this introduces the problem of rotating both beams
without interfering with guys. The Telrex Big Bertha solves this by rotating a self-supporting
tower. Such a tower, 112 feet high, equipped with antennas and accessories, costs
over $15,000 in place.
Now tower height. Fig. 5 presents costs, based on $10 per foot for ordinary
lattice tower, guys, anchors and foundations. Erection costs are added, beyond 40-foot
height, up to $400 for the 150-foot height. No allowance is included for rotator,
indicator, insurance, etc. Gains are related to the assumed ideal of 1° takeoff
angle by use of the image-antenna geometric construction. Analysis by the indicated
technique shows that, to a close approximation, DX e.r.p. at 1° increases as
the square of tower height. That is, each time the tower height is doubled, 6-db.
improvement is appreciated. Because DX signals often arrive (and should be transmitted)
at angles considerably above 1°, this figure must be weighted downward. Fig. 5
has been constructed on the basis of linear relation between e.r.p. and tower height
- 3-db. improvement for each doubling of height. This agrees fairly well with Utlaut's
results for very high effective heights.
Fig. 4 - Antenna costs.
Fig. 5 - Tower Costs.
The concept being developed is: Cost per db. for the last db. of improvement
which can be handled economically. Suppose we can afford $200 for the last db. By
examination of the curves, we see immediately we should run 1-kw. input, for in
this department the last db. costs only $100. We choose from Figure 4a boom length
of 40 feet. Tower height per Figure 5 is 75 feet. Total cost, adding the corresponding
ordinates of Figures 3, 4, and 5 is $1980.00.
Perhaps this cost exceeds our means. Maybe we can afford only $50 for the last
db. in each of the three departments principally affecting DX effective radiated
power (DX e.r.p.). On this basis, we choose 150 watts, 23-foot boom, and 40-foot
tower height. Total cost is $350.00.
Once the concept is understood, curves may be developed to fit the individual
situation, and to take in to account all sorts of other variables: cable losses,
fixed costs for auxiliaries, commercial increments of sizes, nested-rhombics-plus-real-estate
versus Yagis, etc.
An important consideration, so far excluded in order to simplify the discussion,
is the fact that antenna db. work both ways: send and receive. Appraisal of antenna
and tower costs for DX e.r.p. should therefore be weighted, so as to allocate a
share of these costs to the receiving advantage. A reasonable factor is one-half.
That is, the dollar values of ordinates of Figs. 4 and 5 can be cut in half, for
economic optimization of design with respect to DX e.r.p. only. On this basis, Fig. 6
shows optimum combinations as a function of funds available. The figure is constructed
by assuming various dollars-per-last-db. values, and connecting the resultant values
by curves. Of course the optimization differs somewhat from the $50/db. and 200/db.
examples above, because the receiving components of costs have been broken out separately.
For example, suppose $500 are available for the relevant parts of the station.
Refer to "Total Cost" on Fig. 6 at $500. Draw a line straight up. Parameters
are: Power, 150 watts; Boom, 27 feet; Tower, 50 feet. Gain relative to the reference
installation is 5.3 db.
If $2000 are available, parameters are: Power, 1 kw.; Boom, 40 feet; Tower, 75
feet. Gain relative to the reference installation is 17.2 db.
Fig. 7 presents compatible equipment complements with regard only for DX
e.r.p. - no allowance for concurrent receiving advantages. This represents a more
conventionally accepted approach. In effect, transmitter power is given greater
initial emphasis. These db. are cheap and more convenient than antenna/tower db.,
but do not bring corresponding receiving advantages. After the legal power limit
is reached, optimization proceeds much as on Figure 6. For $500, read off: Power,
275 watt; Boom, 26 feet; Tower, 44 feet: Relative gain, 7.3 db. For $2000: Power,
1 kw.: Boom, 40 feet; Tower, 75 feet; Relative gain, 17.2 db.
Economically, c.w. telegraphy gives by far the most DX per dollar. Not only is
this true because more DX stations are available by c.w., but also because of greater
efficiency, expressed in db. as follows 17:
C.w. ............................................ 0 db.
D.s.b. a.m., order-wire
quality... +17 db. required
S.s.b. order-wire quality ........... + 14 db. required
S.s.b. DXers will nearly all aver that the table above should be corrected to
read "11 db." instead of "14 db" for s.s.b.
After reading this, it is fair to ask: "What does a db. in DX e.r.p. really buy,
after all, in terms of DX capability'?" The answer is that, other things being equal,
it buys a lot. Six db. buy, competitively, a decisive advantage.
So far, system-design trade-offs have been discussed. The matter of constructional
alternatives is also important in station economics. The remainder of this month's
text is on miscellaneous antenna-construction comments. Antenna mounts are frequently
the major item of home built equipment.
Fig. 6 - Compatible equipment complements (with allowance
for receiver advantage).
First, re antenna towers. There are fine products on the market. These firms
also sell the numerous necessary and desirable accessories: brackets, clamps, clips,
anchor, winches, guys, and even gin poles. Only a small percentage of DXers use
these products, because the majority (a) can't afford them, and/or (b) home and
neighborhood consideration won't permit them. Speaking in generalized terms, short
of all-out optimum performance, a practical and almost universally applicable construction
is to use telescoping pipe sizes, side-supported to the house, with a hand-winch
for running the antenna up and down. This is what is done at W3AFM. Some particulars
The cheapest and most universally available mast structural element is water
pipe. It comes in 21-foot lengths. It should be ordered black, unthreaded. Local
suppliers usually deliver. Prices run about $10.00 a length, depending on weight.
Figure 12¢ to 20¢ a pound, depending on discounts, location etc. Sizes
are confusing, because they are based on nominal i.d. of the standard weight. "Extra
strong" and "double extra strong" are of the same material, but smaller i.d. (same
o.d., to match fittings) for greater wall thickness, Some examples are given in
the table below, in which "XXH" means "double extra heavy":
Many Yagis are made to mount on 1 1/2-inch
pipe. Speaking in generalities, and depending on prevalent winds, antenna, etc.,
16 feet of unsupported height (i.e., 16 feet above guy attachment or last bracket)
can be good design, whereas 20 feet can be risky. It is wise, if using water pipe,
to telescope section, in such a way that the top 10 feet are single-wall, next 10
feet double-wall, next 1 foot triple-wall, etc.
Steel much better than water-pipe iron exists. Chrome-molybdenum electroweld
or seamless AISI 4142, heat-treated to 180 k.p.s.i. looks great - but costs ten
times as much per pound and seems almost impossible to get in less than mill lot.
A popular mast in the Northeast is Diamond "E" (1020 cold-drawn steel) 2 inches
o.d. by 0.25-inch wall X 20 feet long, selling for about $60.00. So far as known,
one of these has never folded.
Fig. 7 - Compatible equipment complements (DX e.r.p, optimization).
Aluminum alloys have a modulus one-third that of steel. This can make them very
willowy, unless kept short, and thick-walled.
In the best installations, the mast or top tubular-section of the antenna-mount,
projects only a few feet above the main steel-lattice tower. The rotator is then
a few feet below the top of this tower. Sometimes, for reasons previously mentioned,
a lattice tower is not practical. In such cases, the pipe mast is extended down
to the ground, and the rotator mounted near the ground. Such an antenna support
is commonly clamped, loosely so the antenna can be turned, to the side of a house.
When this is done, it is important to spread the stress on the house structure.
This is done by angle-iron, channels, or wooden members, coupled typically by 1/2
inch threaded bolts all the way through, for example, the attic walls. At W3AFM,
a vertical 2 inch X 6 inch X 12 foot plank is bolted to the side of the house, with
2 inch X 4 inch X 6 foot lateral stress-spreaders horizontally inside the attic
wall. The strongest wood commonly stocked is oak. Clear white oak, unfinished, and
suitably stained, is used. The vertical plank is attached by four 1/2 inch bolts,
and projects 4 feet above the peak of the roof. Three husky electrical clamps attach
the mast to this plank. The second 21 foot pipe section up from the ground is slotted
to fit over a 1/2 inch dowel in the lowermost 21 foot pipe section; so the antenna
may be lowered to a height reachable from the roof by first raising it a few inches,
then lowering the disengaged part to the ground. Such a load requires work advantage.
Boat winches, available from Sears or Ward's at about $25.00, are well suited to
this purpose. Half-inch polyethylene boat rope is a good value.
It can be dangerous and expensive to economize on small hardware fittings: eyebolts,
U-bolts, clamps and the like. Items stocked at Sears, Ward's and neighborhood urban
hardware stores generally go only to 3/8 inch sizes, and are cheaply made. For example,
a 3/8 inch eyebolt, of the kind having a formed, unwelded eye, if used at the top
of the vertical plank mentioned above, could easily unwind and drop the load on
the ham below at the boat winch. Where does one get better hardware? Try industrial
suppliers, such as McMaster-Carr Supply Co., 2828 North Paulina Street, P.O. Box
4355, Chicago, Illinois 60680.
An item not as widely known among amateurs as it should be is the screw anchor.
This is a long rod with eye at one end and an auger plate at the other, by which
it is screwed in to the ground. A common size is the Hubbard 7526 or Chance 6346,
66 inches long, 3/4 inch rod, 6 inch blade, which sells for about $5.00. Fully screwed
into good soil, these withstand 4500-pound pull. Many varieties are made: swamp
anchors with blades 15 inches in diameter; rock anchors, etc. A well-stocked supplier
is Graybar, with warehouses in most U. S. cities. Graybar also has excellent ground-rods
- not as cheap as you see in radio stores - but better. Typical sizes are Hubbard
9348, 5/8 inch in diameter X 8 feet, or Hubbard 9,450, 3/4 inch X 10 feet. They
even make one (No. 9697) 1 inch in diameter X 40 feet long. Graybar also stocks
clips, clamps, thimbles, arming bolts, eyebolts, shackles, etc.
If the terms used above, and other such as: "gin pole," "tag line," and" come-along"
are unfamiliar, some preliminary reading or talking with persons having experience
in rigging, is desirable prior to undertaking a major antenna project. Alternatively,
there are people who, for a fee, will take the problems off your hands.
Rescue squads or fire departments which accept public contributions, can be helpful
in raising antennas to the top of, say 60, foot towers or poles. A local amateur
made a $25.00 contribution (tax deductible?) and found willing and effective cooperation.
Raising a tower with a crane can be dangerous, though it is common procedure
commercially. A 130 foot, 24 inch tower was once being raised with a 60-foot crane
by wiring the tower base to its concrete foundation, and picking up the tower by
attaching the crane hook below the tower center. The base temporary wires failed.
The tower base whipped sidewise and killed a rigger instantly.
Locally, the most experienced ham riggers, W3MSK and W3GRF, prefer to assemble
towers such as the AB-105 vertically in place by carrying up pieces, bolting them
on, climbing up to the next level, etc., using a light gin pole and ground helpers
to pull up materials as required.
The reason why self-supported high towers are rare compared to guyed towers,
is that they cost several times as much. (The next installment will appear in an
17 "Median Signal Power Required for Reception of Radio Transmissions
in the Presence of Noise,"
Technical Report 5, U.S.
Army Radio Propagation Agency . .June, 1961.
* Costs in parentheses are used for apportionment purposes in construction of
Total costs in all cases are taken from the left ordinate.
Posted Posted July 15, 2021
(updated from original post on 2/28/2014)