Yes, it can be quite easy...not for precision but for "effectiveness".
Forget the megger...other than an intellectual exercise, it will tell you
nothing about the effectiveness of your radial field for RF.
If you don't have one, borrow an MFJ-259B or 269 antenna analyzer.
With no radials attached, measure the feedpoint impedance at resonance (X=0
or near 0).
Write it down (it could be 50, 60, 70 ohms or so for a 1/4 wave vertical).
YOU MUST MAKE THIS MEASUREMENT AT THE FEEDPOINT, NOT IN THE SHACK!)
Now, start adding radials, two at a time.
Put down two (opposite directions, 180 degrees apart), re-measure the
feedpoint Z at resonance and then keep adding radials two at a time evenly
spread out on opposite sides and remeasuring feedpoint Z. DON'T FORGET TO
ADJUST THE ANTENNA ANALYZER FOR X=0 EACH TIME. (The resonant freq will
likely change slightly each time you add radials, so you have to re-zero for
resoanace (X=0) each time you take a measurement.After you have done 8 or
so, start adding them in 4's so you don't have to take as many measurements.
Here is what you are likely to see (something like this, I just picked 80m
for the example), and assumes approximately 1/4 wave radials (but there is
no resonance in the radials, so don't worry about precise length)
Freq # of Radials Input Z (at resonance)
3720 0 67
3725 2 61
3727 4 53
3728 6 49
3729 8 44
3730 16 41
3730 32 39
3730 50 38
3730 120 37.5
What is this telling you?
1. We already know that the input Z of a 1/4 wave vertical at resonance over
a PERFECT ground is 37 ohms. This is our "target".
2. Any difference between the 37 ohm ideal and your measured Z at resonance
is largely ground loss in the radial field (JUST WHAT YOU WERE TRYING TO
FIND OUT, NO?)
3. Notice that the feedpoint Z drops quickly at the beginning (with very few
radials) and much more slowly at the end (with many radails). This will show
you the "point of diminishing returns)
4. Notice at 3727 KHz, with only 4 radials, you have 53 ohms input Z...WOW!
Flat 1:1 VSWR...ain't it neat! No, it isn't neat at all. The antenna is
contributing 37 ohms, and your losses in the ground system are contributing
16 ohms ....rotten efficiency. In the case of a 1/4 wave ground mounted
vertical, a 1:1 VSWR is a BAD SIGN, not a good sign. Ideally, you should
have a vswr of 1.4:1, which would indicate 0 ground lossess and 37.5 ohms of
If you want a real thrill, measure the 2:1 VSWR bandwidth each time. You
will find the widest (best match over frequencies) 2:1 bandwidth in KHz with
NO RADIALS. The more radials you put in, the NARROWER your 2:1 vswr
bandwidth will become. Why? Because the earth losses by the inefficient
radial field are resistiive and not frequency sensitive. Your ground system
is acting like half a dummy load, swamping out the frequency sensitivity of
your antenna by putting a resistor on it. As you reduce your ground losses,
your bandwidth (vswr) narrows. When you bandwidth no longer narrows, you
have enough radials. (This is similar to: when your input Z at the feedpoint
stops dropping, you have enough radials).
So, the point of this exercise (I've done it many times, it is quite easy)
is to find the point at which adding more radials does not produce a
significant drop in the feedpoint Z at X=0 (with your handy antenna
analyzer). Depending on your ground characteristics, this could happen with
as few as 16 or as many as 64 1/4 wave radials (or, many more shorter
The seminal work by K3LR on optimizing radial fields is ABSOLUTELY ESSENTIAL
READING before putting in any buried or on ground radial field. It will save
you time. It will save you money. It will save your back! What it tells you
is how to get within 0.5 dB of a 120 radial field. It will tell you how many
radials it will take and how long they should be. Another thing it will tell
you is if you have available to you 1500' of wire for radials, how many and
how long should they be to give me the best efficiency that I can get with
this amount of wire.
I have very good soil here in Iowa. With my Inverted L, I am currently using
26 radials 50' long for 80 meters. I am showing efficiency of greater than
80% (this number is an approximation, not a perfect measurement, as there
are pitfalls translating Zin to efficiency directly) with this setup. I have
another 1000' of wire laying around, so I'm going to put in 20 more 50
footers and that will be sufficient.
I can't emphasize enough reading the K3LR study. There is a VERY simple
formula for calculating radial optimization, and it is completely
consistent with the foundational study by Brown, Lewis and Epstein that has
been the bedrock of radials studies since the 1930's when it was done. Use
it. Then, add the radials as I describe above, writing down the Z
measurements and you will know more about your ground system than 99% of the
rest of the hams in the world. (And you didn't need a megger)
Warning: there is a program on the net called RADIAL_3, that purports to
show efficiency of radial systems vs length and number. Do Not Rely on It.
It does not agree with BL&E, it doesn't agree with K3LR, it does not agree
with NEC-4 computer models, and it doesn't agree with W8JI's actual
measurements. It predicts high efficiency with VERY short radials in high
numbers (like 5 meters long on 80m).
If you want high efficiency in terms of loss, money and effort, follow the
If you do what I suggest, please post your measurements on the list. It
would be most instructive. You need to tell us what kind of antenna you are
using, of course. If it isn't a plain 1/4 wave vertical, then the "target" Z
may be different, but the process is identical.
Hope this helps, have fun.
----- Original Message -----
From: "John Cowan" <firstname.lastname@example.org>
Sent: Tuesday, August 08, 2006 5:08 PM
Subject: [TowerTalk] Ground Resistance
I have followed the posts on radials and ground resistance for some time. Is
there any simple way to measure the impedance of the ground radial system as
you install them to know when you have enough of them in the ground?
Talk is cheap. Use Yahoo! Messenger to make PC-to-Phone calls. Great rates
starting at 1¢/min.
TowerTalk mailing list
TowerTalk mailing list