A couple of observations , that caught my eye .
But a disclaimer , I am also learning and have mounted antenna's off chain link fence's .
The ground plane should be attached at the base of antenna , or vertical radiating element , the center conductor of coax to that and the braid / shield to the ground plane , as short as possible , maybe a couple of inches , use a panel mount 239/259 connector at this point to ease of testing & feedline .
The vertical part should be on a insulated mount , not just stuck into ground .
It's a lot of extra walking back and forth , but to start , you should connect the analyzer with a short jumper to the panelmount connector , then use optics [ binoculars etc. ] after adjusting / setting up analyzer .
This way you may get some baseline numbers that do not have you in the readings .
I am sure there are lots of my suggestions that can be picked appart , but I think that in general there should be some use .
Now to see what happens
Ok. So my vertical base will be 6 feet off the ground. My radials are not elevated. So do I bond them to the bottom of the fence under the feed point, to the pole. Or should I run them up the fence and connect them to the feed point? Another option might be to lower the vertical to ground level(snow level) but that leaves 4 feet of the vertical element below the fence its mounted to. I can not elevate the radials.
I have eight radials that are 34 feet long and the rest are 17 feet long(to answer another question).
A fence is a poor radial system and the best it will do is look like two radials. If it was good then there would be no reason for the ground radials, however it's likely that the ground radials are the best ground for the antenna. In that case, if you attach the radials 6 ft below the feedpoint that 6 ft becomes part of the vertical section and you have effectively a 40 ft vertical fed 15% above the antenna ground. Unfortunately you have the fence attached to the section of the antenna which is part of the antenna vertical section. That's another reason for the weird resonance points and feedpoint impedances. An off center feed can work but you should have a choke at the antenna feedpoint because the shield will not be attached at the low voltage point on the antenna. That can lead to common mode currents on the feedline, which will change the resonant point, and cause loss because the feedline radiates mostly into the ground, and RFI. Also don't attach it to the fence because that fence will not be at ground potential either (unless you just want to accept the associated loss). With the radials at ground level, there is no way of getting around the fact that 6 ft of the vertical will be radiating directly into the fence. In lieu of modeling it, you will have to experimentally determine the resonant points and feedpoint impedance. It won't be similar to other ground mounted verticals.
Originally Posted by AL3K
When you insulate it from the fence, it's unlikely that you will be able to have resonances on both 40 and 20 meters, or even close. The fact that you now have a feedpoint impedance of about 200 ohms on 20 meters is probably due to feeding RF down the fence and feedline. It should be pretty lossy.
So if I drop the antenna down to put the feed point 1 foot above the ground will I have better results? That would mean about 5 feet of the vertical section will be a few inches from the post and fence.
Anyway I now have the antenna at 6 feet.
At 6.68 MHZ R=67 X=0
7.54 MHz R=168 X=89
8.5 MHz R=278 x=113
9.5 MHz R=96? x=221
14.14 R=11 X=98
Its -20 F so its hard to stay out to long!
Dan/K6MHE wrote an article showing how he made a 33' radiator work on 20 & 40m w/o switching: http://k6mhe.com/opuscules.html. It requires λ/4 radials for 40m.
Moving the antenna to the ground is one option. That would eliminate the need for the choke and would remove most of the current going down the fence and feedline. You do have a fence a few inches away from the vertical but there is no way to get around that unless you keep the antenna where it is and slope the radials away from the top of the fence.
Originally Posted by AL3K
-20 F? Ouch.
I never modeled before. I think my MFJ 259b can not read resistance that high. So what numbers should I start with to make a L match at 14.15MHz?
Resistances how high? What configuration are you talking about?
Originally Posted by AL3K
If you are referring to matching a normal 40 meter quarter wave vertical on 20 meters, WA4PRC gave a good reference for that. The values used in that reference should work for a vertical made from about 1 inch pipe. The actual values required will vary depending on the surge impedance of the vertical and that depends a lot on the diameter of the pipe. For a 1 inch diameter pipe about 2.8 uH and about 40 pf should be very close. A few inches of wire in the matching network box can vary the needed values a small amount.
To determine experimentally what you need, get a variable cap from the junk box that will give about 100 pf or so, plus a coil that you can tap. These don't have to be rated for power. Sit down at the base of the antenna with your analyzer and vary the parts until you get a good match. Then measure the value of the parts. Once you have the values required, then you can obtain suitable size components rated for the power level used and build the matching network. For 100 watts the cap should see about 520 volts peak. The network will have to be contained in a box that will insulate it from the weather. If you use a relay, the contacts will have to be rated for the same voltage as the cap. The coil should have about 1.5 amps RMS at 100 watts, so it should be low loss. Should you decide to run 1500 watts into this antenna the voltage goes to about 2KV peak and the coil current goes to about 5.8 amps RMS.
Since you now have the approximate values needed, you could skip the experimental process (and keep from freezing your butt in -20 deg F weather) and build a network that had some adjustment around the required values. If you don't have a normal 40 meter vertical and the fence gets involved, you will have to determine it experimentally.
If you get carried away, you could add more than one matching network and use it on other bands. The pattern should be pretty good from 40 meters thru 15 meters.
Grounding the vertical;
My first 20M VERTICAL was mounted in the center of a garage roof. The radials started about an inch from the base of the vertical element, ran down the slope of the roof, and were directly bonded to the coax shield. This antenna WORKED,
I got my first VK QSO with this ant and 100 W.
This is maybe the simplest and best way to connect a vertical. (except no lightning ground, don't do this in FL).
Skipping ahead 40 years, now my multiband vertical is mounted a foot above ground on a piece of pipe driven into the ground,
My many radials are buried and are on a gentle slope from the base of the vertical into the ground a few feet from the common point. This antenna also WORKS !
Running a jumper cable from the base of an elevated feedpoint to a radial lfield in the ground is not a good practice and could decouple the RF so it is not actually grounded at the coax connection on some frequencies !
FISTS #3853,cc 455
SKCC # 1395,tribune #12
Official US Taxpayer
Ok a 40pF cap did the trick! I have 20meters. So now I'm building a loading coil for 80 and 160(I know it will not be to efficient) . Some say that the loading coil should be DC grounded? I'm assuming that it means grounding(radial system) one side of the coil and tapping it a few turns up with the input? Should I do this? How many turns up?
Thanks for everybody's help. Learning much!