The W8NX Dipole Takes Shape

Talk about not blogging in real time! Raising an HF antenna has been the never-ending story, no doubt a joke to the guys on the DART repeater. And no wonder. I’ve been talking about this since last summer, but only got the antenna up and on the air in March. Why so long? I’ve come to think that I should build my amateur radio station where possible. This was not some conscious decision. Rather, my determination to homebrew sort of settled on my shoulders over the course of the last year. And what better thing to build than an antenna? It’s made of wire or aluminum tubing, is fairly cheap, and is easily tweaked. Well, sort of easily tweaked.

Some easy candidates are a fan dipole for the WARC bands, a Moxon or short Yagi beam for 6 meters, a quarter-wave vertical and the W8NX multiband dipole. This post is about putting together the W8NX “dominant element” 5-band dipole for 80-40-20-15-10 meters that I talked about in my last post.


With a little handwaving, let’s say that the antenna cost about $150. This breaks down roughly as:

150 ft stranded copper wire, 12 AWG            $25
100 ft Belden RG-58/U coax                     $40
(Ted ‘AQM was kind enough to donate some coax, too)
Coax connectors including 1 coax “bullet”  $15
Insulators, PVC pipe, misc hardware            $10
Electrical tape, liquid electrical tape        $10
600 ft black Dacron rope                       $50

This doesn’t count the cordless drill bought at Sears, nor the crimping tool bought at Harbor Freight. There is also no lightning arrestor installed yet.

I got most of the items at the Findlay Hamfest in September, but some parts such as the wire crimps were bought as late as February.

Main Wire Elements

The wire segments were easy. I very carefully cut them to calculated length plus one foot. Made sure the lengths matched for each pair of wires, not always the easiest thing to do when the wire wants to coil up. The leftover wire amounts to about 35 feet.


The stubs were cut from a length of 1/8 inch OD copper fuel line, intended for RC model airplanes, found at a hobby shop. Soft-drawn refrigeration tubing in this diameter would also work. This was one of the tough parts to source. If I did this again, I would pick a common wire gauge to design with and find solid copper wire, or else choose a larger diameter copper tube.

Winding the coax loads

The loads were built on forms cut from nominal 2 inch ID PVC pipe. I cut these forms and the choke balun form with a hacksaw. It’s not too hard to cut PVC but without a jig, it is a little tough to cut a straight line. After cutting, I sanded the ends to remove any sharp slivers of PVC material. I drilled 1/8th inch holes for the cable and pairs of slightly smaller holes at each end for the main wires to pass through. There is no strain relief at the baluns. The holes for the coax were drilled at an angle so there would not be a sharp bend in the cable. The spacing of the coax holes is somewhat critical, since the load uses a fractional number of turns. I drew lines around the circumference of the pipe at the right spacing, drilled a hole at the inner end on the first line, then estimated by eye the angle needed for the last, fractional turn. I drilled the hole at the outer end offset along the second line.

Then came the tricky part. The loads were spec’d out with 0.1 inch spacing between turns, because this sounded good to me at the time. The W8NX software, SINGCOAX.BAS, turned out to be spot on as far as cable lengths. I stripped the cable insulation and separated the braid from the center conductor about half an inch long on each end, then wound the coil on the trap. Nope. When all the slack was taken out of the cable, the insulated part of the cable stuck through the holes at both ends. Stripping the insulation exactly to specified length caused the end of the insulation to line up within millimeters with the holes in the form.

Al Buxton W8NX specifies one of two specific types of coax for the loads, because his software characterizes the L, C and stray capacitance for just those types, and getting it right is pretty critical. So I wound up using Belden 82xx for the loads and much of the antenna cable. The load is built so that the braid at one end connects through the center of the form to the center conductor at the other end. This was not difficult to solder together, though coating the joint with liquid electrical tape got a little messy.

Adding the tap to the cable was hard. W8NX’s instructions say to take a length of coax and wind the shield around the bare shield of the coax coil at the tap point. Fine, but winding two turns of braid around the tap point, then crimping on a copper collet resulted in… a joint that slipped easily. So I tried to solder the joint. Big mistake! Wound up ruining a length of coax, 14 feet!

The secret is to wind a couple of turns of shield around the tap point of the coil coax, then twist the shield to tighten down the turns at the junction, and only then crimp on the collet to hold everything in place.

This done, I coated the tap joint with more liquid electrical tape. When this had dried, regular electrical tape, with some tension, was wrapped around the whole length of the form to protect the coax from moisture and UV. Buxton is adamant that the tap should connect to the outer element on each side, so I drew an arrow on each form to remind me of its orientation.

Making the choke balun

This was easy. I looked at several sources such as this (link) and this (link) before coming to the conclusion that 14 turns of RG-58 close-wound on a length of 4 inch ID PVC pipe would be the best compromise for a healthy reactance blocking shield currents on the most bands. The 10 meter reactance is a little low compared to other bands, as a result. Like the loads, the choke has electrical tape wrapped around its entire length to protect the coax. Also, I added a strain relief at the bottom of the choke so that pulling on the coax would not add strain to the coil. The same length of cable runs from the antenna feedpoint, forms the choke, then runs about two thirds of the way to the shack.

Final assembly

Putting the antenna together needed little more than a warm day. From the overage on the wire length, I allowed half for putting the wires through the insulators, and half for putting the wires through the load forms. The sequence was, put a collet on the wire, then feed through the insulator. Twist the open end several turns around the main length of the wire, then crimp down the collet using the crimping tool. This was a copper to copper joint all the way around. The crimped joints are surprisingly strong. There’s a lot of tension end-to-end with the antenna raised to full height, yet no sign that any joint is weak.

The load connections needed two collets, one for the open end of the wire, and one for connecting the trap lead at each end to the main antenna wire. So I had to be careful about mounting the collets in the right place. It turns out that connecting the load to the antenna wire at the center point of the wire segment through the coil form was a lousy idea. This was because the crimp pliers actually had the crimp form – a C-shaped cutout on one arm and a similar bump on the other – behind the hinge of the pliers. This meant I had to maneuver the pliers inside the form to get the cutout in the right place to make a proper crimp.

The stubs went onto the inner sections of wire about 5 feet from the center feedpoint. I used a copper junction block meant to add taps to an electrical bus wire. For each stub, the tubing was soldered through a hole, then the block was slid onto the main wire and screwed down tight.

Finally, the center insulator was connected. Each side used two collets, one for the main wire end, and the other to connect one coax lead, center conductor or shield, to the main wire.

The choke balun was hung from the center insulator with a length of rope that took any tension off the coax pigtail. I tightened the hanging rope by wrapping another piece of rope around it just below the insulator. As long as the main rope doesn’t move out of the grooves in the insulator, everything will be fine.

The rope wasn’t used until I erected the antenna. I eventually used something more than 200 feet, leaving more than 350 feet, probably, for other projects.


Didn’t. Asked around but could not find someone willing to loan me grid dip meter for tuning the loads, or an antenna analyzer for tweaking the finished antenna. See the “results” post for more. Erecting the antenna was a whole ‘nother story, as told in my next post.


Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s