DIY 4G LTE Yagi Antenna in 10 Steps for $10
by Damon Chandler

|
Copyright © 2012,
Damon Chandler and
EnCoded Communications Group
Last Updated:
September 1, 2012 |
 |
|
Introduction |
|
This article
demonstrates how to build a 14-element Yagi antenna for Verizon 4G LTE
(or for 3G)
in 10 steps for $10.
Background
Many users in smaller towns
and rural locations rely on cellular service as their only
source of broadband internet. Unfortunately, being in a rural
location often also means being quite far from the nearest
cell tower and/or in a weaker signal area due
to terrain, trees, or other obstructions. In these situations,
an external antenna designed specifically for your cell
carrier's frequencies can often make the difference between
having a slow and unreliable connection and having a
consistently fast connection.
The Yagi-Uda antenna--often just called a "Yagi"--is a popular antenna due to its
gain, directionality, and relatively lightweight design (see the
figure to the right). Unlike the compact internal antenna on a USB modem or cell
phone, a Yagi's driven element is large enough to be
fully sensitive to the frequencies of interest. But, unlike an omnidirectional antenna or rabbit ears, a Yagi's passive
elements can "focus" the signal from a particular direction and
reject signals from other directions, thus increasing its
directional gain. Yet, unlike
other directional antennas (e.g., a parabolic grid), a Yagi usually weighs
less than a few pounds.
There are many online
resources that describe how to create homemade Yagis for VHF, UHF, and WiFi, and for
Australia's 3G/NextG networks (see the
Acknowledgements section). A particularly useful
article by Jim Klitzing (W6PQL),
published in 2006 in the
QST
journal, provides very detailed instructions on how to build
a UHF Yagi antenna. Inspired by these references, I decided to
build my own Yagi antenna for the Verizon 4G LTE bands.
This tutorial
Here, I document the
process required to build a 4G LTE Yagi antenna using the
construction techniques described in
W6PQL's article. His techniques are simple and achievable
with relatively common hand tools. I
have actually built three antennas based on his techniques, the first two using a
PVC boom, and the latest--the one described here (and pictured
at the top of this page)--using a metal boom. This tutorial is
essentially a spinoff of
W6PQL's article for 4G LTE rather than for general UHF. My primary goals were
twofold:
- Keep costs at a bare
minimum while still yielding an effective and relatively
rugged antenna.
- Utilize parts that are readily available (e.g., from a
local hardware or electronics store).
The result came to
approximately $10 in parts, 10 steps to construct, and an
afternoon of your time (approximately 4-5 hours). This is not
bad at all considering that the antenna has a theoretical gain
of 13.5 dBd and 15.7 dBi. The same construction
techniques can be used to create a 3G antenna; I've provided
datasheets for both 4G LTE and 3G versions.
It's important to note
that the information provided here is not meant to serve as a
definitive reference on how to build an optimal Yagi antenna.
Remember, I've built only three such antennas. Some sacrifices were made in the interest of cost and
simplicity. Don't hesitate to experiment with alternative parts or
construction techniques based on your own judgment--it's
certainly possible to improve the design (e.g., by using
larger-diameter elements), and I welcome
feedback
on such improvements. Note that if you use different materials,
you'll most likely need to generate a different datasheet; see
the section on The Design Software.
Organization of this
tutorial
The remainder of this
article is organized into the following sections:
-
The Design Software (the software used to design the
antenna)
- Parts for the Antenna (the required
parts)
- Building the Yagi (the construction
steps)
- Quick Testing of the Yagi
(preliminary testing results)
- Notes on Mounting, Cables, and Coax
Connectors (brief tips on mounting and connecting the
antenna)
- Acknowledgements (and links to useful
sites)
- User Comments (provide feedback
here)
|
|
|
The Design Software |
The
design softwareAs
described in
W6PQL's
article, I used VK5DJ's Yagi
Calculator software to design the Yagi. The software provides a datasheet with the required
element lengths and positions, and further specs for the coax
connection (balun).
If you plan to follow this
tutorial as-is, using the same parts described here, click the link below to download
my datasheet for either 4G LTE or 3G (coax balun info included):
For 4G LTE, Verizon
currently uses 747-787 MHz (with a 20 Mhz gap in the middle of
this range). For 3G, Verizon uses either 824-894 MHz or
1850-1990 MHz (again, with gaps in the middles). I did not
provide a datasheet for the 1850-1990 MHz range because the
dimensions of the resulting antenna are much smaller. It's certainly possible to create a Yagi for this range; just use the software to create your own
datasheet and keep in mind that the boom length can decrease
from 5 ft to approximately 2 ft.
The design frequency
Both of the above datasheets have
been designed for the top ends of their respective frequency ranges--787
MHz for 4G LTE, 894 MHz for 3G. I chose the maximum frequency as the
design frequency because a Yagi's sensitivity supposedly drops very rapidly for
frequencies above the design frequency, but drops more slowly for
frequencies below the design frequency. Thus, if your elements' lengths
are a bit off during construction, you should still have a usable
antenna. Although I have not tried it, you might want to experiment with
a design frequency somewhere in the middle of each of the above ranges to achieve
better performance (e.g., using 767 MHz for 4G LTE).
If you need to make a change
The above datasheets
assume that the passive elements make metal-to-metal contact
with the boom. Because such metal-to-metal contact tends to
shift the antenna's sensitivity to a higher frequency range, the
Yagi Calculator software has compensated for this effect by
slightly lengthening the elements. The amount of compensation
depends on the boom's diameter (and possibly also the elements'
diameter). For this reason, the above datasheets are specific to
the parts described here.
If you plan to use a different
design frequency, a boom with a different diameter, or elements with a
different diameter, then you'll need to use the software to create your
own datasheet. Similarly, if you plan to use a non-metallic boom, or you
plan to mount your elements insulated from the boom, then you'll also
need to use the software to create your own datasheet. The software is
free and very easy to use (click the image above to download it). |
|
|
Parts for the Antenna
(Total Parts Cost = $10 USD) |
|
Four parts are
needed for the antenna:
|
|
1/2" x 10 ft Electrical Metallic Tube (EMT) Conduit (approx.
$2)
[the link to the left directs to this item
at Lowes.com] |
|
We'll
use EMT conduit
for the antenna's boom.
We need only 5 ft
of EMT conduit for a single Yagi antenna, but you'll pay about
the same price for a 5 ft length vs. a 10 ft length. Plus, you
can save the other 5 ft for another Yagi.
Note that
if you can't fit the
10-ft EMT into your car, just borrow a hacksaw from the Tools
section of Lowes/Home Depot and cut it yourself in the store.
Or, ask an employee to cut it for you.
EMT conduit is a
fraction of the cost of aluminum or copper. Similar to aluminum,
it's lightweight and more corrosion-resistant than copper. But,
unlike aluminum, you can solder EMT as long as you sand-off the
finish at the intended soldering point. (EMT does not accept
solder as readily as copper, so it's still a bit of a challenge; be sure to use plenty of flux.)
|
 |
| |
|
|
Rigid metal rods (12 gauge or thicker) (approx. $2-3)
[the link to the left directs to a steel tomato
cage at Lowes.com] |
We'll need
some form of thick, rigid metal
for the antenna's elements (reflector, driven element, and
directors).
There are many options for this metal, spanning a wide range of
prices. Two of the cheapest options that I've found are a
galvanized steel tomato cage or metal coat hangers.
You can
buy a
galvanized steel tomato cage from your local hardware/garden
store for $2-3. This metal is well-suited for outdoor use.
Alternatively, 7-8 metal coat hangers will suffice. I
purchased an 8-pack from
Walgreens for about $2.50 (shown on the right). These
particular hangers have a plastic coating, which can easily
be stripped off.
Keep in
mind that you really just need metal rods or tubes that can be
soldered. If you
can't find coat hangers or a tomato cage, you can use an old
grill, thick copper wire (6-12 AWG), welding rods,
or some form of metal tubing, perhaps from an old TV antenna. In fact, using larger-diameter
metal is supposed to increase the Yagi's bandwidth,
which may lead to better results.
The total length of metal
that you'll need depends on the number of elements you intend to
put on the Yagi. For the 14-element Yagi shown in this tutorial,
we need about 9 ft (2.7 m) of metal.
|
 |
| |
|
|
Plastic Standoff
(approx. $0.50)
[the link to the left directs to a PVC T-fitting at Lowes.com] |
We'll need one plastic standoff to mount the Yagi's driven element to
the EMT conduit (boom).
The driven element is the only element
that's physically connected to the coax (and thus to the modem).
The driven element needs to be electrically isolated from the boom; all other
elements will be soldered to the boom.
Find something which you
can repurpose for use as the standoff, preferably something
durable. A plastic bottle cap, for example, will work; but, it
might not last long in the sun. A
1/2" PVC T-fitting, an electric fence insulator, or even a 1/2" hose repair coupling are
better
choices if you plan to mount your antenna outdoors.
Here, I will
demonstrate the use of a 1/2" PVC T-fitting as the plastic standoff.
|
 |
| |
|
|
Short Length of
Flexible Coax Cable
(approx.
$4-5)
[the link to the left directs to an RG58 cable at
WeConnex.com] |
We'll need about 1.5 ft of
flexible coax cable to create the balun and pigtail connection for the antenna.
For this
short length of coax, I suggest choosing the type of coax with
the same impedance as your longer coax run (the long run of coax
that connects the modem to the antenna).
-
If you're using
50-Ohm coax such as LMR400, then use RG58 for the antenna's balun/pigtail.
-
If you're using 75-Ohm
coax such as RG6, then use
RG59 or RG6 for the antenna's balun/pigtail.
The
previously described
datasheets contain information regarding the baluns for
RG58, RG59, and RG6.
RG58 is 50-Ohm coax;
RG59 and RG6 are 75-Ohm coax. 50-Ohm coax is theoretically matched to the impedance of the antenna and the modem. But, in practice, the impedance of the antenna varies
within the 4G LTE and 3G frequency ranges,
so there's no way to be perfectly impedance-matched across the
entire band of interest.
The antenna's balun/pigtail
is very easy to construct. There's no harm in trying RG59 or RG6
if you already have some on hand. You can always replace it
later with the RG58 version. (Click
here for more notes on this topic.)
Here, I will
demonstrate using RG59 because I had a remnant piece available (with a
pre-attached F-type connector).
|
 |
|
|
Building the Yagi |
|
The following 10
steps will be used to build the antenna:
- Making the boom
- Making the elements
- Mounting the elements to the boom
- Making and mounting the coax connection
The following tools and supplies are needed to
construct the antenna:
- Hammer
- Drill with 3/32" bit (or
anything large enough to accommodate the elements)
- Hacksaw
- Metal file and coarse
sandpaper
- Center punch (or nail punch,
or awl, or strong nail/screw)
- Wire strippers
- 4-ft or longer straight edge
(ruler or level or other)
- Ruler with cm scale and mm
tick marks
- Soldering iron (40-80 W
preferred); plus solder and flux
- Permanent markers (one
fine-tip marker and one medium-tip marker)
- Epoxy and screws
If you don't have a
particular tool, it's certainly possible to use a substitute. A vise is
also highly recommended.
|
|
Step 1: Cut the EMT conduit in half |
|
Using a hacksaw, cut the 10-ft EMT conduit into two 5-ft sections.
You'll only need one of the halves, so set the other half aside
for another project.

It helps to use a vise to
hold the conduit while cutting (see
Figure 1). If you don't have a vise, you
can use clamps or your foot.
After you've cut the
conduit, briefly sand or file the
cut ends to remove any of shards of metal.
Again, one of the 5-ft sections of
conduit will serve as the antenna's boom. |

Figure 1: Cut the
10-ft EMT conduit into two 5-ft sections. |
| |
|
|
Step 2: Draw guide lines along the
length of the boom |
|
The next step is to draw two lines along the length of the boom,
one line on the top, and one line on the bottom. These lines
will serve as guides to help ensure that the elements are aligned
(see Figure 2).

To draw the top line,
place your straight edge along the length of the boom, clamp the
boom to the straight edge (or step on the boom to prevent
movement), and then follow the straight edge with your
marker, drawing a line on the boom.
To determine the location
of the bottom line, use your ruler to measure 18 mm down from
the top line (the diameter of 1/2" EMT conduit is 18 mm).
Mark this location. Then,
flip the boom over and use the straight edge and marker to draw
the bottom line--starting at your mark--in the same fashion used for the top line. |

Figure 2: Draw top
and bottom guidelines along the length of the boom. |
| |
|
|
Step 3: Mark the positions of the
elements |
|
The next step is to mark the position of each element on the
boom.
Using your ruler and marker, measure along one of the guide
lines and mark the positions of the elements according to the Yagi
Calculator datasheet. Repeat
this process for the other guide line. Thus, you should have
marks for the elements on both sides of the boom (i.e., along
both guide lines).
Note: The idea is
to have marks...
As suggested in
W6PQL's article, the idea is to have marks
on both sides of the boom so that you can drill halfway through
from each side of the boom rather than drilling all the way
through in one shot. Trying to drill all the way through in one
shot is a recipe for misalignment. By drilling from each side,
you have a better chance of having aligned holes (though, slight
misalignment is inevitable and can be corrected later by bending
the elements slightly).
Figure 3 shows a screenshot of the datasheet for 4G LTE with the
element positions highlighted in yellow.
- The reflector is to
be located 30 mm from the beginning of the boom.
- The radiator (driven
element) is to be spaced 76 mm from the reflector, which is
106 mm from the beginning of the boom.
- For the directors,
the "Spaced" column indicates the spacing from the previous
element. The "Boom position" column indicates the position
relative to the beginning of the boom.
Figure 4 shows an illustration of the relative position
("Spaced") and absolute positions ("Boom position") of the first
four elements.
Note: Unless you
happen to have a metric ruler...
Unless you happen to have a
metric ruler that's 5 ft long, you'll inevitably need to use the
numbers in the "Spaced" column to measure the spacing from the
previous element rather than the position from the beginning of
the boom. The disadvantage of using the "Spaced" column is that
it can lead to error-propagation. But, if you take your time,
the errors can be negligible (sub-millimeter).
You can round the datasheet's numbers to the nearest millimeter
without affecting performance. But, when rounding, you must be
consistent for both guide lines; otherwise, you'll introduce
misalignment.
Marking the element
positions along each guide line is a tedious and time-consuming
process. But, it's important that you take your time and measure
correctly. The element positions are not as crucial as the
element lengths (see Step 5). The most important thing is to be
consistent for both guide lines in order to maximize the
alignment of the holes on both sides.
Figure 5 shows the boom with the marked positions. The "FD"
stands for "folded dipole;" this is the driven element
(radiator). |

Figure 3:
Screenshot of 4G LTE Yagi datasheet with positions highlighted.

Figure 4: Diagram
of the relative and absolute positions of the first four
elements based on the 4G LTE Yagi datasheet.

Figure 5: Mark the
position of each element along the guide lines on both sides of
the boom. |
| |
|
|
Step 4: Drill the holes for the
reflector and directors |
|
The next step is to drill holes through the boom for the
reflector and directors. (No hole is needed for the driven
element because it won't be mounted through the boom).
To keep the drill-bit from
drifting during drilling, use a center punch to make an
indentation at each marked position (except for the driven
element); see Figure 6. If you don't
have a center punch, a nail punch, awl, or even a strong nail or
screw will work. (As you can see in the background of the left
image of Figure 7, I had an actual
nail punch, but I couldn't find it when I needed it; so I
resorted to using a plain nail.)

Next, drill the holes
halfway through from each side of the boom (see
Figure 7, left). If you don't
have a drill bit that matches the size of your elements, it's
better to go slightly smaller, since you want a tight fit during
soldering.

As you drill the
opposite-side hole for each element, run the drill bit briefly
though both holes to remove metal shards (Figure
7, right). If your drill bit is too small, you can
rock the bit in a circular motion to slightly enlarge the hole
(but, don't overdo it--remember, you want a tight fit so that
the elements don't move during
soldering).
|

Figure 6: Make an
indentation at each marked position.

Figure 7: Drill the
holes halfway through from each side of the boom. |
| |
|
|
Step 5: Measure and cut the reflector
and directors |
|
The next step is to measure and cut the reflector and
directors according to the Yagi Calculator
datasheet.
Figure 8 shows a screenshot of the datasheet for 4G LTE with the
reflector and director lengths highlighted in yellow.
- The reflector is
197.6 mm long (round to 198 mm).
- For the directors,
the "Length" column indicates the length of each director
(again, you can round these to the nearest millimeter).
The datasheet states that
the elements lengths must be within one millimeter of the stated
lengths.

Use a pair of metal-cutting dikes or a hacksaw
to cut your elements. You can remove bends by gently pounding
them out with a hammer.
Be sure to cut each element a bit long, and then file the element down to the
correct length. This filing also helps remove jagged edges from
the cut ends.
Figure 9 shows one of the elements, stripped, straightened,
and then cut/filed to length. The element shown is Director 1,
which has a length of 178 mm. Figure 10 shows the reflector and all 12 directors cut to
length.
After all of the elements
are cut to the proper lengths, mark the midpoint of each element.
Also mark half the boom diameter (9 mm for 1/2" EMT) away from both sides of
the midpoint; see Figure 11. These
marks will later serve as guides when mounting the
elements to the boom. The particular element shown in
Figure 11 is the reflector, which has
a length of 198 mm, a mark at the midpoint of 99 mm, and
flanking marks at 90 mm and 108 mm.
Note: It's wise to label the elements...
It's wise to label the elements
as you cut them. The lengths are so close to each other that
it's very easy to get them confused. As shown in
Figure 10, I simply poked the
elements through a labeled piece of paper to keep them in order.
As a final sanity-check,
you can line-up the elements against a straight edge to ensure
that they gradually decrease in length from the reflector to the
last director. As you can see in
Figure 10, this is indeed the case. Also observe that the
lengths are extremely close to each other, so only pull them out
from the paper one at a time, on an as-needed basis.
|

Figure 8:
Screenshot of 4G LTE Yagi datasheet with lengths highlighted.

Figure 9: Cut each
element a bit long, and then file it down to the correct length.

Figure 10: All 12
elements cut to length and organized.

Figure 11: Mark the
midpoint and two 9-mm-away-from-midpoint points. |
| |
|
|
Step 6: Measure and cut the driven
element |
|
Now it's time to tackle the folded dipole that will be used as
the driven element.
The driven element is arguably the most crucial piece of the
entire antenna. It's the only element that's directly connected
to the coax feed; all other elements serve to focus the signal
onto the driven element. The driven element's length will
affect the antenna's frequency tuning, and its shape and
placement will affect the antenna's impedance.
The Yagi Calculator datasheet lists the specs of
the folded dipole, a screenshot of which is shown in
Figure 12 (for 4G LTE) with the most important parts
highlighted. As stated in the
datasheet, the distances should be measured from the insides of
the bends (inner edge on one side to inner edge on the other
side).
For the 3G version of the antenna, the lengths will be shorter
because the design frequency is higher--refer to the actual
datasheet PDF.
To create the folded
dipole, cut a length of your metal slightly longer than the
datasheet's "Total rod length" (403 mm). Then, mark the
distances HI and GF (70 mm). Next, bend the coat hanger around
some round object to create the bends (e.g., a can, bottle, jar,
etc.).
Figure 13 illustrates the basic structure
for which you should aim.
Note: Unless you
have specialized bending tools...
Unless you have specialized
bending tools, you'll probably find it nearly impossible to meet
all of the distances between each of the labeled points in the
datasheet. However, the most important distance is the
folded dipole's length (183 mm in the datasheet; the distance
AE). As long as you achieve that length, and come reasonably
close on the bend diameter length (BI = DF = 37 mm) and the gap
(HG = 5 mm or smaller), the tuning and impedance will be fine.
Figure
14 shows the resulting folded dipole, which I created by
bending the coat hanger around the spindle of my vise. It's not
perfect shape-wise, but it's extremely close to 183 mm length-wise. Also, note that I used a brass-looking coat hanger for
this folded dipole only because I already had a spare cut to a
total length of 410 mm. You can use the same metal as
used for the other elements.
Note: The bend
diameter and gap...
The bend diameter and gap
are free parameters that can be adjusted in the Yagi
Calculator. I chose 37 mm for the bend diameter because that
happens to be diameter of the spindle on my vise. If you have
something else like a can, bottle, or jar that you plan to use
to create the bend, that's fine. Just be sure to maintain the
folded dipole's length (AE = 183 mm). In fact, I've created
folded dipoles in the shape of a rectangle (and even in the
shape of sunglasses), and they have worked fine.
|

Figure
12:
Screenshot of 4G LTE Yagi datasheet
with most important folded dipole dimensions highlighted.

Figure 13:
Illustration of folded dipole shape and important dimensions.

Figure 14: Actual
folded dipole created from a coat hanger. |
| |
|
|
Step 7: Mount the directors to the boom |
|
Now that all of the elements have been created, the next step is
to mount only the directors to the boom. (The reflector and
driven element will be mounted later.)

Because we'll be soldering
the directors to the boom, we need to prep the surfaces to
ensure that they accept the solder. To do this, sand around each
hole on the boom (see Figure 15).
Now, for each director:
- Place the element
through its hole in the boom.
- Ensure that the
element is centered by using the 9-mm-away-from-midpoint
marks drawn previously in Step 5 (see
Figure 16). You can also measure the amount by which the
element protrudes from each side of the boom. It's centered
when these protrusion amounts are equal.
-
Apply a generous amount of
flux around the joints.
- Solder the element to the boom. To do this, first touch
the iron to the joint, wait several seconds for the joint to
heat-up, and then feed the solder onto the joint.
Warning: The
elements will get hot...
The elements will get hot while
soldering, so wear gloves and don't grab the elements near the solder joints. The boom also
transfers heat quite rapidly, so avoid touching the boom near
the solder joint until after it has cooled off.
Figure 17 shows a close-up of
one of the solder joints.
My
soldering job in Figure 17 is poor.
There is way more solder than is needed. However, I had a weak
25 W soldering iron that would not stay hot for more than a few
seconds after contact with the boom/elements. I had to let the
iron reheat for 15-30 seconds after each application, and thus
the messy soldering job. It helps to have a more powerful iron
or a propane torch.
After you've soldered all
of the directors, sight down the boom to check that
the directors are aligned (see Figure 18).
Bend the directors as needed to correct for inevitable
misalignments. |

Figure 15: Sand
around each hole down to the bare metal to ensure that the
solder will bond.

Figure 16: Place
each element through its hole, and ensure that it's centered
across the boom by using the 9-mm-away-from-midpoint marks drawn
previously in Step 5.

Figure 17: Closeup
of solder joints (they're not perfect, but they're quite solid).

Figure 18: Sight
down the boom to ensure alignment of the elements; bend the elements as
needed to correct.
|
| |
|
|
Step 8: Mount the folded dipole to
the boom |
|
The next step is to mount the folded dipole to the boom. The folded dipole needs to
be insulated from the boom. This means no metal-to-metal contact
and a vertical separation of at least half a boom diameter away
from the boom (as a rule of thumb).
Thus, to mount the folded dipole, we need some form of
insulated standoff.
Figure 19 illustrates how the folded dipole will be
positioned around the boom. Because our folded-dipole's bend
diameter is 37 mm, and because the boom diameter is 18 mm, the required standoff height is 9.5
mm. A slight deviation from this height is OK (e.g., 10-11 mm
instead of 9.5 mm); it will simply result in the folded dipole
not being perfectly centered. I have done this in the past with
no measurable effect on performance.
To use the PVC T-fitting
as the standoff, do the following:
-
Slide the T-fitting onto the boom, measure
9.5 mm up from the top of the boom, and then mark and drill a hole from each side
(see Figure 20).
-
Trim the ends of the "T" so that the fitting can
slide into place without hitting the first driven element. (You
really only need to trim one end of the "T"--the end that will
face toward the first director.)
-
Slide the folded
dipole through the drilled hole as shown in
Figure 21 (you may need to
temporarily unfold the dipole to get it around the curves).
-
Drill two holes for mounting screws, and then mount the fitting
+ folded dipole onto the boom using these two screws (see
Figure 22). Ensure that the folded
dipole lines up with your previous position mark for the driven
element (you may need to rotate the T-fitting to see this mark).
-
Bend the folded dipole as needed to be as
straight as possible, and then apply epoxy to the hole through
which the folded dipole goes through the fitting (see
Figure 22).
As you can imagine, there
are numerous way to secure the folded dipole to the boom in an
insulated fashion. Experiment on your own to determine the
technique that you find best, given the parts that you have
available.
A better alternative to
the T-fitting would
be to use some form of plastic box such as a plastic conduit
body. This way, you can potentially enclose your coax
connections inside the box.
|

Figure 19:
Illustration of the general idea behind using a plastic
standoff.

Figure 20:
Prepare the PVC T-fitting by trimming the ends and drilling
two opposite-side holes through which the folded dipole will
fit.

Figure 21: Slip the
folded dipole through the holes (you might need to undo some
bends temporarily).

Figure 22:
Attach the assembly to the boom.
|
| |
|
|
Step 9: Mount the reflector to the
boom |
|
The next step is to mount the reflector to the boom.
Following the same procedure
described in Step 7, mount and solder the reflector to the boom.
Figure 23 shows the result. (Again,
the soldering job is not great.)
After mounting the
reflector, sight down the boom to ensure it's aligned with the
directors, and bend as needed to maximize alignment.
Note: The only
reason the reflector...
The only reason the reflector
must be mounted after mounting the other elements is to
ensure that the PVC T-fitting can slide onto the boom from one end of the boom.
If you mount the reflector prior to mounting
the driven element, the reflector will prevent you from sliding
the T-fitting onto the boom.
Of course, if you're using an
insulated standoff that doesn't need to slide onto the boom,
then you can certainly mount the reflector at the same time as you
mount the directors during Step 7.
|

Figure 23: Solder
the reflector to the boom. |
| |
|
|
Step 10: Create and attach the balun
and coax connection |
|
The final step is to create and attach the balun and coax
pigtail to the antenna.
The Yagi Calculator software creates a picture of the balun
and pigtail, which is shown in Figure 24
for RG59 coax with polyethylene insulation. Here, I will demonstrate using RG59 because
I had a remnant piece with a pre-attached F-type connector.

Note: You can
choose the coax type...
You can choose the coax type in
the Yagi Calculator software. The only thing that
changes when using different coax is the length of the balun's
"U" loop. Different types of coax have different velocity
factors, and it's the velocity factor and design frequency (787
MHz in the present example) that determine the length of the balun's "U" loop.
K7MEM's site has an excellent description and online
calculator for the balun.
Regardless of which balun calculator you use, you'll need to
know what type of dielectric (inner insulation) your particular
coax uses. RG58 commonly uses polyethylene (PE), whereas RG6
commonly uses foam PE. As a rule of thumb, if the dielectric is
translucent enough that you can see through to the center
conductor, then it's most likely PE. If the dielectric is nearly
opaque, then it's most likely foam PE.
For the RG59 used here, look closely at the dielectric in
Figure 27. Notice that it's
translucent and is thus PE.
To construct the balun and
coax pigtail, first cut and strip the pieces as shown in
Figure 25. The top piece will be used
to create the balun, and the lower piece will be used for the
pigtail.
Note: The length of the balun's
loop...
The length of the balun's
loop (126 mm in the current example) should be measured over the
unshielded portion. For this reason, you should initially cut
the coax for the balun to be longer than the loop's length so
that you have sufficiently long center conductors and
sufficiently long twisted outer conductors on both sides.
In Figure 25, the total length of the
top piece measured from the ends of the center conductors is
approximately 180 mm.
Next, as shown in
Figure 26, simply solder the
connections to match Figure 24's diagram.
If you're using RG6, you won't be able to solder the outer braid
because it's most likely made of aluminum. In this case, just
twist the outer braids together and secure them with a crimp
fitting.
Finally, as shown in
Figure 27, solder the center
conductors of the balun to the folded dipole and use tie-wraps to
secure the coax. If you plan to mount the antenna outdoors, be
sure to cover the stripped portions of the coax with a thick coating of
epoxy to prevent water ingress.
The construction of the
antenna is now complete! Figure
28
shows a photo of the completed antenna. |

Figure 24: Diagram
of balun and coax pigtail for 4G LTE Yagi with
polyethylene-insulated RG59.

Figure 25: Cut and
strip the pieces for the balun (top) and coax pigtail (bottom).

Figure 26: Solder
the connections to match the diagram.

Figure 27: Attach
the balun and pigtail to the folded dipole.

Figure 28:
Completed antenna.
|
| |
|
|
|
Quick Testing of the Yagi |
|
Due to weather
constraints (and an intimidating nest of hornets), I have not yet had a chance to test this
particular antenna on my roof; I've tested the
antenna only in my attic. However, I have tested a
10-element version of this antenna made with a PVC boom,
and it yielded impressive speed boosts placed on a mast
on my roof. I plan to provide more thorough test
results in the near future.
For
now, here are some quick test results of the antenna
in my attic, on a tripod, and using the UML290 modem.
Some brief details of
the testing:
-
Modem: Pantech
UML290
connected directly to laptop (.230 firmware). The antenna was plugged into the
modem's normal antenna port, and not the MIMO/diversity
port.
-
Software:
Verizon Access Manager version 2.6.3.4. This older
version of VZAM was used because, unlike the later
versions, it displays the RSSI and SINR without
going to the diagnostic screen. However, this older
version seems to display fewer bars for the same
RSSI and SINR as compared to the newer versions.
-
Location:
Mid-level in my attic of a one-story home in
Stillwater, OK. I would estimate that my attic is
about 15 ft above the ground, and the tripod adds
another 4 ft.
-
Terrain: In a
shallow valley. Many trees around, with leaves, but
the trees are short (20-40 ft is my guess).
-
Coverage: I'm
in an Extended 4G zone according to VZW coverage
maps. But, I'm within a mile of a regular 4G zone.
The LTE tower is 7-8 miles away.
-
Aiming: I knew
the general direction in which to aim based on my
other antenna adventures. I pointed this antenna in
that same direction.
|
|
Results
with no external antenna |
|
- RSSI: -90 to -86 dBm.
- SINR: -3 to 2 dB.
- Speeds: 2-5 Mbps DL, less
than 0.3 Mbps UL.
 |

|
|
Results
with the DIY antenna |
|
- RSSI: -77 to -72 dBm.
- SINR: 12 to 17 dB.
- Speeds: 15-23 Mbps DL, 5-8
Mbps UL.
 |

|
|
|
The results are
promising. I expect further performance gains by
mounting the antenna higher and outdoors. Again, this
older version of VZAM seems to display fewer bars for
the same RSSI and SINR as compared to the newer
versions. I always run a speed test to be sure. During these
tests, the antenna was plugged into the modem's normal
antenna port (the one near the SIM card), and not the
MIMO/diversity port.
3Gstore.com
calls the normal antenna port the "3G port" and the MIMO/diversity
port the "4G port." If you plug the antenna into the
MIMO port, the modem should theoretically use the
internal and external antennas in a MIMO configuration.
However, this will work only if your internal antenna
has a decent signal to begin with. In my setting, as you
can see from the above results with no antenna, the
signal was quite poor to begin with, so plugging the
antenna into the MIMO port gave me about the same DL speeds, but it didn't help the UL speeds.
With that in mind,
I have experienced even better results with using two of
these DIY antennas in a MIMO configuration (i.e., using
both ports of the VL600 or UML290). I hope to post some
MIMO results in the near future. |
|
|
|
Notes on Mounting,
Cables, and Coax Connectors |
There are always hidden
costs to every DIY project, and this project is no exception.
Here, the culprit is connectors, cables, and mounting hardware.
The topic of mounting this antenna could benefit from its own
DIY article. In regards to cables and coax connectors, there are
many good online resources for this topic. Here, I provide some
brief notes on these topics.
|
Notes on mounting |
-
At the present, I have
not yet explored ideal mounting options. This topic deserves
its own DIY article.
-
Although EMT conduit is a
fairly lightweight, I recommend supporting the antenna from
its midpoint or at two equidistant points away from the
midpoint.
-
One of the advantages of
using EMT conduit for the boom is that you can readily find
many clamps, elbows, and connectors designed specifically
for EMT at your local hardware store; and, they're cheap.
Muffler clamps are also cheap or can possibly be salvaged
locally. You'll have to experiment and be creative if you
plan to pursue DIY mounting hardware.
-
Of course, the quickest
and easiest solution is to purchase an antenna mounting kit,
e.g., from your local RadioShack. If you don't already have
a mast or other mount on your roof (e.g., for a TV antenna),
then you'll need some form of mounting kit to get started.
-
If you don't have a metal roof, consider mounting the
antenna in your attic. You won't get the best signal
possible, but at the same time, you won't have to worry
about rust, lightening, or wind.
|
|
Notes
on the long run of coax |
-
If you have RG6 already in
your walls, or you have a
long run of spare RG6 available, then by all means, use it.
Yes, there will be an impedance mismatch. However, in my
experience with both the UML290 and VL600 modems, the
difference in performance is inconsequential (I have not
tested other modems). I've tried both RG6 and LMR400 at 50
ft lengths, and I've come to the conclusion that there's
more fluctuation due to weather or tower traffic than there
is due to the slight impedance mismatch. LMR400 has a bit
lower loss than RG6 (see
here), but the cost and labor needed to run a new length
of LMR400 is something to consider.
-
If you have RG59 already
in your walls, then I'd suggest replacing it with either RG6
or LMR400. If you don't already have a long run of spare RG6
available, then go with LMR400. Although I've never tried a
long run of RG59, it's reported to have a much greater loss
than the other two types in the LTE frequency range (see
here).
-
If you don't already have
cables in your walls, and you don't already have a
long run of spare RG6
available, then go with LMR400. If you have to buy
cable, and you have to spend the time to fish it through
your walls, then you might as well use LMR400.
|
|
Notes on connecting the antenna to
the long run of coax |
-
How you connect the antenna to
your long run of coax depends on what you used for the
antenna's pigtail cable.
-
If you used RG58 for the
antenna's pigtail, then it probably has a BNC male connector.
Assuming your LMR400 cable uses N-type male connectors,
search online for a
BNC-female to N-female adapter. These are readily
available online for around $7 with shipping. Or, better
yet, buy your RG58 with an N-female connector.
-
If you used RG59 or RG6 for
the antenna's pigtail, then you can easily find an
F-type barrel connector at your local Walmart, Lowes, or
other store. A 2-pack at Walmart runs for $3.
|
|
Notes on connecting the long run of
coax to your modem |
-
To attach the other end of the
long run of coax to your modem, you will need to purchase a
short adapter cable designed for your particular modem.
There's no easy way to avoid this purchase. Both
3Gstore.com and Verizon Wireless sell these adapter
cables for around $10-15, and both of these places ship
quickly. Your local Verizon Wireless shop may also have some
in stock.
-
These short adapter cables
almost always have an FME male connector to connect to the
long run of coax. You will need to find an FME-female-to-?
adapter, where ? depends on your long run of coax.
-
If you used LMR400 for
your long run of coax, and assuming it uses N-type male
connectors, then you will need to search for an
FME-female to N-female
adapter. These are readily available online for around $10
with shipping.
-
If you used RG59 or RG6 for
the antenna's pigtail, then you will need to search for an
FME-female to F-female adapter. Maxmost.com sells an an
FME-female to F-male adapter
here
and on eBay
here for around $10 shipped from NY. You can then use an
F-type barrel connector to change the gender. Of course,
you can get creative and use intermediate adapters; below is
what I used during my tests (I'm actually surprised it works
as well as it does).

|
|
|
|
Acknowledgements |
| A special thanks to Jim
Klitzing (W6PQL) for
providing an excellent
QST
journal article from which much of the material on this page
is based, and thanks to John Drew (VK5DJ)
for the Yagi Calculator software. I am also grateful to the
following resources for providing much guidance on antennas and
related mobile-broadband topics:
|
|
|
User Comments |
|
|
|
|
|
|