On Sunday I did an experiment to compare the performance of a small transmitting loop (“magnetic loop”) to an inverted vee dipole. The results are here:
Very nice bit of work. Thank you Richard.
Various people asked for a comparison on 20 metres - which along with 40m is a hugely popular SOTA band. I am now running essentially the same experiment but using 20m antennas. This link shows live results updated every 2 minutes. Once the experiment is over the link will expire.
The dipole has an 11 dB attenuator in line thus the difference between the antennas will be the displayed median + 11dB.
I will write the experiment up fully later so don’t draw too many conclusions at this stage!
This could be very eye-opening for some. In a way, you are writing the book on real-life antenna performance testing right now.
Can’t wait for the infamous duel, vertical vs inverted V end-fed/dipole.
Razvan M0HZH / YO9IRF.
Some people think it’s a fairy tale of course. Here’s a verification test that Frank W3LPL ran yesterday. Two WSPRlites, identical power, into a combiner and a single antenna. Only spots from Europe used.
Accurate power measurement at 5000 km.
To be on the safe side maybe I should do the experiment, provide the data, and leave the interpretation part out?
Well, I have operated the whole winter and spring within WWFF program having 23 summits. I am using FT-817 with inv-V (DX-Wire UL wire), 12m RG-174 without balun and telescope’s hight has been about 9m. Result has been excellent from my point of view. More than 10 different NA stations and several VK station. One P2P qso to VK too. The antenna installation is always a bit different but I have been very satisfied to the inv-V. Together (SOTA + WWFF) I have done about 150 summits so far.
73, Saku OH2NOS
It will always be interesting to see what new insights others gain by analysing our synchronous data in different (and more rigorous) ways. Owen Duffy has been looking at it recently:
People impatient for the horizontal vs vertical tests will have to wait a little longer. As more people are using the method, more is being learned about how best to do comparative experiments. It’s a new technique and exploring the data is proving very interesting. I hope that it will be worth the wait.
Prior to hillside tests, I will be comparing the same antennas on our open-field test site by way of a calibration experiment. Each experiment takes 12 hours at least to do (before data analysis).
Today I am running a slightly more elaborate trial in further preparation for “the big one”. The purpose here is to look at how much data I gather from different ranges to determine the appropriate power level to use for the main experiment. I want to gather enough data for others to slice it and still have a good sample size. I am also interested in seeing how the ground-wave signal differs locally, hence the third system (8km away).
Antenna 1 20m GP ident G31CWI
Antenna 2 20m inverted Vee ident G32CWI
Synchronous systems; 200 mW, TX only.
Antennas 40m apart on flat open field test site.
Antenna 3 40m long doublet ident G3CWI
Random, TX and RX; 2 Watts
8km from the other systems
Live data here - updates every 2 minutes. Use the menu for other reporting options. Link will cease at the end of the day.
Red dots= only inverted vee being heard
Blue dots = only ground plane being heard
Green dots both being heard
…don’t jump to too many conclusions just yet!
Just curious, what is the height of each antenna and what is the orientation of the inverted V?
73 Marc G0AZS
I will write it up with details later.
First look at the data shows a clear win for the groundplane. I will do some more detailed analysis tomorrow.
Write-up and data here:
Thank you for this, Richard. It confirms the vertical is indeed the better choice for portable operations.
There are however a few caveats: your arrangement is close to optimal. We need to keep in mind that elevated radials have a significant impact on the vertical antenna’s efficiency, as opposed to having them laying on the ground. Also, it would be interesting to see if lowering the number of radials to 2 changes the performance much - simulation seems to suggest it doesn’t.
Razvan (M0HZH / YO9IRF)
Thank you for your comments Razvan,
I don’t believe that such a broad statement is warranted by one rather specific experiment that was done once! I would be more cautious in interpreting the results myself.
Regarding other experiments - it’s always interesting to have ideas but I already have plenty myself of course (and limited time).
Quickly thinking about this Richard, your experiment shows the evaluation on transmit. But how about on receive ?
My thoughts with this design of vertical is that they have been poorer receive antennas when compared to a dipole, inverted V. Perhaps it may be interesting to reverse the experiment using receivers instead ?
Some of us, myself included have used separate aerials for this reason.
Just my thoughts !
I don’t think Richard’s experiment proves a vertical is better at all. A quick glance shows very similar performance - more experiments needed to give a better picture.
The terrain (and ground) on a summit is different to a field.
Like all good science, it good to have a baseline, a control if you will, so this first experiment is just the start!
Colin, you need to give it more than a quick glance. If the performance was similar the slope of the line would be 1 and it would pass through (0,0). Instead, the slope is 0.85 and it passes through (0, -4.9). So, the ground plane is being received at 0 dB where the inv-Vee is at -4.9 dB, Clearly the ground plane is superior.
It is worth noting that the ground plane has 5m less RG174 which is worth 0.49 dB in extra transmitted signal. It also has no choke balun so effectively has 5 radials, the 5m feedline forming the extra radial.
Very interesting result. Thanks, Richard!
73, Gerard - VK2IO
With my vertical which is a 9M squid pole and a loading coil midway I found much better results when the radials were elevated.
Made quite a few dx contacts with this vertical on 20/30/40M.
Quite an easy aerial to set up as well which was a another benefit.
There was one set of reports, sure, but the set contained 989 measurements of the signals from each antenna! Somehow I think that means more than one experiment iteration, which would suggest one report on each antenna.
It is far more conclusive than the usual method used, ie. a bloke on the radio saying one of your antennas moved his S meter a bit more than the other one. Using the DSP inherent in the WSPR software to compare two signals received in the same 110 second period has to be more rigorous than the S meter method so often relied upon by the popular e-forum reviews.
When theory predicts the vertical has a lower angle of radiation and therefore at a “dx distance” will have probably had fewer hops and hence lower loss than the signal from another antenna, I think it’s worth considering. I have used that theoretical prediction for the last 50 years to good effect. With these measurements we start to see the practical, empirical measurements confirming the theory.
As stated earlier, I like this experimental approach and hope Richard has time to do more of it. It’s a great application of the tools, combined with the neat WSPRlite signal sources.
73 Andrew VK1DA VK2UH