3
$\begingroup$

I tried to make a dual-band (2m/70cm) dipole this morning by following the instructions on this page: http://www.amateurradio.bz/2m-70cm_vertical_dipole_antenna.html, with the rather major significant exception of placing the antenna's elements in a groove I routed out of a 2x4 and I'm using bare #10 copper wire instead of stainless steel.

I plugged a VNA into it and the antenna's lowest SWR was ~1.09 at ~124MHz. I tried it in a couple different environments and orientations and made sure there wasn't anything conductive right near the antenna (other than the wood, which I'm expecting to reduce the effective velocity factor of the wire.

Since I have no shortage of #10 wire, I started trimming the ends in 1/2" increments. I got the lowest SWR to 1.04 at 148MHz... but only while horizontal! Orienting the antenna vertically causes the lowest SWR point to jump to 1.52 at 153MHz.

What's happening here? Why was the tuning so far off? Is it the wood? Is it the copper wire? Why does orientation affect the tuning so much? I know I've got a lot of variables here, but is there one that is obviously the most important difference?

$\endgroup$
  • 1
    $\begingroup$ Can you rule out common-mode current on the feedline as a factor? $\endgroup$ – Phil Frost - W8II Feb 15 '17 at 1:17
  • $\begingroup$ @PhilFrost-W8II: The entire system is: analyzer -> ~10ft of RG-8X -> antenna. Where would common-mode current get into that? Honest question here, because I don't know. $\endgroup$ – William - Rem Feb 15 '17 at 1:49
  • 1
  • $\begingroup$ @PhilFrost-W8II Good point. Without a suitable 1:1 balun (common-mode choke) at the feedpoint, the SWR could be affected by nearby conductive objects (such as his hands) near the coax. Even touching the VNA can affect the reading! And even with a balun, the SWR might be affected by the antenna's proximity to the earth, house wiring, and his body. Measurements would be best done outdoors with the antenna as high as possible. $\endgroup$ – Mike Waters Feb 15 '17 at 19:08
  • 1
    $\begingroup$ @William-Rem VHF baluns sound like a good topic for a separate question $\endgroup$ – Phil Frost - W8II Feb 15 '17 at 20:17
1
$\begingroup$

It was the wood!

It may also be the fact that I'm using copper wire instead of stainless steel rod, but I can't test that as I don't have access to any stainless.

Further experimentation this evening show much more predictable results. While I thought the 2x4 would be relatively transparent to RF, simply being near the 2x4 seems to cause the whole system to seem electrically much longer than it is, thus making the antenna very difficult to tune.

At first, I thought that it was conducting RF to the wood because the wire was bare, but I cut new elements just now out of my solid-core #10 THHN, this time leaving the THHN on everywhere except where the bolts contact the wire and had basically the same results. Bending the wire away from the wood such that the tips of the dipole (where voltage is highest?) caused the match to jump from 118MHz to 137MHz.

My best guess here is that the lumber is not fully dried and whatever water left in it is causing it to be a (poor) conductor, causing some capacitative coupling. I'd still love for anybody with more experience to chime in though, as I'm just guessing and am pretty new to making antennas.

$\endgroup$
  • 1
    $\begingroup$ I think you're right with the wood. Even fully dry wood has some moisture content from ambient humidity. Water is a lossy dielectric, and wood isn't a perfect insulator. The antenna can couple to it inductively and capacitively, even if it's insulated from direct contact. Copper vs stainless steel probably has negligible effect. $\endgroup$ – Phil Frost - W8II Feb 15 '17 at 2:12
  • $\begingroup$ Thanks, @PhilFrost-W8II, for the confirmation. I'll print a bracket out of ABS soon and that should solve that. $\endgroup$ – William - Rem Feb 15 '17 at 5:02
  • 1
    $\begingroup$ Attaching the wire directly to the wood will not only add a little capacitive loading, it will also effectively decrease the velocity factor just as insulation will. In pure bare copper in air it can range from .90 to .97. The wood could easily be dropping it to .70 or even lower, resulting in a ~20% increase in electrical length even ignoring any loading. $\endgroup$ – Hamsterdave Feb 15 '17 at 13:49
  • $\begingroup$ "Bending the wire away from the wood such that the tips of the dipole (where voltage is highest?) caused the match to jump from 118MHz to 137MHz." This is not clear to me, and I don't see how any dry untreated wood would affect it, seeing how the feedpoint Z of a dipole is so low and only the feedpoint is touching the wood (right?). When bending the wire away from the wood, did you also move the dipoles farther apart from each other, and/or farther away from the mast? What is your mast material? Can you post a photo of your antenna here? $\endgroup$ – Mike Waters Feb 15 '17 at 16:56
  • 1
    $\begingroup$ @Mike Waters, the rf loss tangent for most cellulose based materials I can find good numbers for (including wood, paper, and phenolics) range from about 0.025 to 0.05. This is approximately 100 times more loss than typical plate glass, and >1000 times the loss of low loss dielectrics like PTFE and HDPE. Dielectric losses would also be max near the ends of the dipole where the e-field is strongest. Make the wood just a bit wet, which pine nearly always is, and it is reasonable that losses, capacitive loading, and velocity factor could result in the 20% change in resonant freq OP observed. $\endgroup$ – Hamsterdave Feb 15 '17 at 21:04
1
$\begingroup$

I suppose you were standing on the ground when you tested your antenna so rotating the elements in a vertical position made them closer to the ground effectively detuning your antenna.

To avoid that, the antenna elements must be above the ground at least 1/4 of the wavelength - in your case higher than 2m / 4 = 50cm at all times.

Another factor could be your feed line. Try coiling some coax near the feed point (simple RF choke) and see if your tuning changes - http://www.k3dav.com/rfchokecoil.htm

$\endgroup$
  • $\begingroup$ I wasn't handholding the antenna at the time, but had set it on a non-conductive surface. When I did final tuning though, after figuring out not to put it inside of a 2x4, the bottom tip of the antenna (attached to the shield) was only about 20cm from the ground. $\endgroup$ – William - Rem Feb 22 '17 at 1:08
  • $\begingroup$ Willam - you want it higher than that above the ground to avoid detuning. $\endgroup$ – Paul Shiryaev Feb 25 '17 at 0:15
1
$\begingroup$

This is an interesting antenna with but with no information included about behavior of the unique part at 70cm.

For RF insight into the design, a horizontal 2 meter dipole for 144 MHz was modeled using #8 copper wire at 2 meters elevation. The 70cm portion was formed by wires from the feedpoint connections going down then to the sides beneath the 2 meter arms. This is an approximation of the "U" shaped elements described in the article.

Tuning for 144 MHz was easy. It worked out to be 19.055 inches for the long arms with a feedpoint separation = 1.05 inches. Total Width=39.16 inches.

Horizontally: SWR=1.5, Z=69 Ohms, Efficiency=77%, Gain=7.5 dBi.

Vertically: SWR=1.4, Z=68 Ohms, Efficiency=51%, Gain=2.6 dBi.

However, tuning for 446 MHz was critical and required radically different shape and dimensions.

Spacing between arms was much wider: 3.89 inches down and the length of the arms much shorter: 2.12 inches... with total width=5.29 inches.

Horizontally, the "L" shape still radiates like a dipole with a very minor vertical component.

Horizontally: SWR=1.05, Z=50 Ohms, Efficiency=73%, Gain=9.4 dBi.

Vertically: SWR=1.02, Z=51, Efficiency=60%, Gain=6.2 dBi

Note: The Z impedance of a vertical is not effected by feedpoint height. Not so horizontally. Z goes up and down around 72 Ohms (1.4 SWR) so tuning changes with height.

$\endgroup$
  • $\begingroup$ Thanks for modeling this. I ended up with a kind of L shape and I'd been wondering about what kind of nonsensical radiation pattern it might have. $\endgroup$ – William - Rem Mar 12 '17 at 20:00

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.