Hi Andrew,
how are you? I use a current transformer, a diode and a very small LED measurement module. Not larger than 4x2x2cm. I hung it on the antenna and read it at a distance of 1m. I couldn’t notice a change, when I go to a greater distance.
73 Chris
1 Like
Fine idea. I have done my mesurement 6 years ago. Let’s have an other try.
There is an excellent article about feedline length for end fed antennas here: Of end-feds and feed-lines
It’s a good read for anyone thinking of using an end fed antenna, with much food for thought. The short version is that feedline lengths of less than 1/4 wavelength shouldn’t have a significant impact on antenna performance. Longer feedlines are only problematic at certain lengths, and careful choke placement can be effective with enough forethought.
The same author also did an excellent experiment regarding current distribution, with results that are very similar to what’s already been shown in this thread. You can find it here: Electrically isolated end-fed vs. center-fed dipole radiation
1 Like
Very interesting and thanks to all contributors. I have a question:
I learn from your reference, a single device (ferrite core and copper wire) has a loss about 1.2dB, meaning 100W in, 23W will dissipate in heating the toroid.
I rushed to build 2 devices for a back to back test, my dut has an average loss about 1.2dBm
| Frequency | 3.573 MHz | 7.074 MHz | 14.074 MHz | 18.100 MHz |
|---|---|---|---|---|
| Gen Input | -21.6 dBm | -21.6 dBm | -21.6 dBm | -21.6 dBm |
| Back to back | -25.1 dBm | -23.9 dBm | -24.1 dBm | -23.85 dBm |
| Device loss | 1.75dBm | 1.15 dBm | 1.2 dBm | 1.12 dBm |
| QRP 5W output | 3.34 W | 3.84 W | 3.79 W | 3.86 W |
Why bother to use this kind of toroid transformer with only 77% efficiency ?
It could be better but 77% is not too bad a price for a simple multi band antenna. A ZS6BKW doublet ATU system might be no better. And it’s an order better than those dinky tapped whips touted as great SOTA antennas.
Hi Pascal,
If you read some of the posts from Heinz in this thread, you’ll see how you easily achieve more than 90% efficiency (e.g. using different toroid geometry and thigh winding with CuL).
Why using a resonant EFHW antenna? Well, it’s super easy to setup and it’s resonant on several bands, just to name two reasons.
73 Stephan
1 Like
Hi Stephan
I was badly expressed, my question was attributed to the HyEndFed device, not the Heinz’s 1:49 device and more than 90% efficiency with ferrite toroid.
The author mentions about that HyEnddFed device with proof of RF current circulated on pigtail or coax shield. I had a look at the inside picture, there is a coaxial choke (bifilar choke) built in! Wander to know the source of those currents from coaxial shield or pigtail come from?
73, Pascal
1 Like
Hi Ron,
I developed and built a tuner for ZS6BKW by the use of 300 Ω feed line only.
It works on incremental variation of feeder length and a plastic varicap for fine tuning (10, 20, 30, 50cm lengths with 4 DPDT switches).
A back to back test reveals no reflection and less than 0.1 dB loss.
3.5MHz
21MHz
28MHz
1 Like
We should be careful not to produce quick shots. The photo does not show a HyEndFed product, but a realization of DF1BT of an EFHW coupler with integrated common mode choke AND as can be read in the text of his contribution, as a completion of the EFHW antenna, a so-called “pigtail or counterpoise” (not shown on the photo) to be connected in front of the common mode choke (marked in blue on the photo below).
1 Like
Well I constructed three different couplers (transformers) #1 is a 14/2 with a Bifilar primary and a cross-over (similar to those shown in several websites), #2 is the same as #1 except with 21/3 (21 turns secondary and 3 turns primary) and #3 is no crossover, 24 turns secondary and 3 primary with the windings compressed tightly like the photo shown by HB9BCB in his original response.
I did 3 back to back RBN tests of the three different couplers.
They are with 3 different wire sets for the different couplers. The test was done on 80m band within 30 minutes of each other in the evening near 2334 UTC or local time about 8 pm in the evening.
I will go on an activation next Wednesday to obtain a similar comparison from the field. I have already results from the field from #1 (12/29 W4T/SU-024) and #2 (1/13 W4T/SU-024) and now want to get results from #3 (planning on W4T/SU-062 just 20 miles north of W4T/SU-024). I am not sure how objective this is but Design 2 has 1.5 S unit better on the RST compared to #1 and more answering stations on 60m. The comparison has been made using SMP. I also managed a QSO on 80m with #2 (an S2S) but not with #1 on 12/29/20. We will have to see how #3 does in the field.
Ariel NY4G
1 Like
Ariel,
Such comparisons of RBN spots can actually be misleading if one does not consider the triggers the skimmers use to drop a spot.
- If, for example, a station calls cq on the same frequency with the same callsign and the same speed for a certain time, the number of RBN spots will be not the same as if the same station would have changed the frequency or/and the speed by a few kHz or/and wpm several times, e.g. after changing the test object
- However, even taking this into account, there is still the possibility that interference occurring at the location of a skimmer could impair or make reception impossible in a certain frequency range for a certain period of time
- In order to be able to exclude also any unknown peculiarities of the reporting system, one would probably have to swap the order of the test items in a further test
- In addition, if the condx are fluctuating, a test duration of approx. 30 minutes can definitely influence the results.
From this point of view, your above test documentation is unfortunately incomplete and therefore not necessarily meaningful.
But that someone could come up with the idea of making antenna comparisons based on RBN spots that are 2 weeks apart, as in the case of the comparisons made above at 60 m, I have never seen before, really not.
1 Like
Not really coming to conclusions at this point. Just analyzing data and presenting it. I did change frequency between transmissions from the different couplers. I agree with your assessment. Perhaps a head to head test using WSPR conducted at the same time will be a more objective test.
73 Ariel NY4G
1 Like
Well I did a head-to-head WSPR test using WSPRlite transmitters using two of my call signs. I used VY2AJ (my Canadian Call Sign) to transmit using the high efficiency core which I called “Design 3” versus the 14/2 secondary/primary using the “conventional” bifilar winding with a crossover which I called “Design 1”. I only have two WSPRLite transmitters. Both antennas are configured the same way - inverted vee with 6 meter apex and the same azimuthal orientation from the feed point to the end of the wire. The transmitters were started within a minute of each other and the test was conducted over a 6 hour period from 11:30 AM local time to 5:30 PM local time to get a variation of propagation for both the transmitter and receive stations. The result is as shown by the DX10 Graph which plots the average of the 10 most distant receiving stations for the previous hour. The red trace is the less efficient core and the blue trace is the more efficient core.
During very good propagation, the less efficient core can hit distant receiving stations just as well as the more efficient core. As the sun sets over Europe, the more efficient core transmitter can still hit distant stations and the less efficient one less so.
DXplorer (software used in analysis of the WSPRLite) has a feature called “simultaneous spots” where the receiving stations report the SNR from the two transmitters received at the same time. This of course changes over time but the SNR from the more efficient core shows a trend for better SNR (signal to noise ratio).
Ariel NY4G
4 Likes
Those were activation two weeks apart and not tests per se but an effort to find if any meaningful improvements in signal reports are realizable. I was just curious to find out if such a comparison is meaningful. I will still do the activation since that is what SOTA activators do which is to activate. If the results are not seen to be meaningful then I will not report it and just keep the results to myself.
73 Ariel NY4G
Thanks Ariel for sharing the results of the WSPR test carried out on 20 m.
I would actually have expected that the graphs for the two test objects would more or less shake hands over the entire duration of the test.
Therefore I suspect at the moment that an unknown player could be responsible for the indentation in the red graph. In order to confirm or rule this out, the test would probably have to be carried out twice on exactly the same antenna locations with interchanged test objects.
I would also be very interested in WSPR test results on 80 or 60 m. The test results from the 2 different bands could then on the one hand be checked for agreement with the calculated values and, on the other hand, contribute to a more well-founded assessment of the examined test objects from a point of view of radio practice.
Heinz,
Thanks for your suggestion. I will repeat 20m and interchange transmitters keeping everything else the same including the start/stop times.
With respect to 60m, I will have to build two low pass filters for that band before I can run that comparison.
Ariel NY4G
1 Like
Here are the results of the repeat WSPR head-to-head transmissions on 20 meters. The initial head-to-head test was run yesterday and reported on further up in this thread.
The setup is as shown in the image below. The two antennas are arranged as inverted Vee in the same azimuthal orientation with the feedpoint on the south side denoted by the blue boxes. The image shows 10 foot topographic contours and also shows the global imagery to show both the presence of roads and vegetation. The two antennas are separated by about 75 feet of space. The NY4G coupler has the 2 turns primary with the crossover and bifilar winding of the primary. The VY2AJ coupler has the 3 turns primary with a tap instead of the bifilar winding and the turns are wound tightly together with no crossover. Both transmitters are identical 200 milliwatts WSPRlite programmed with their respective call signs. The NY4G and VY2AJ positions were swapped from yesterday. The image shows the positions for today’s comparison. The only variables are the couplers and the relative position of the antennas with each other. They both see the same propagation conditions.
The results are as follows following the 6 hour run time. The chart shows the average of the 10 most distant stations for the preceding hour.
The blue trace being VY2AJ and the red trace NY4G. Similar to yesterday, VY2AJ outdistanced NY4G. Today by quite a bit more than yesterday - the Y axis being a logarithmic scale on average by almost a factor of 2:1. The position of the VY2AJ appears to have some sort of additional advantage (in addition to the more efficient coupler/transformer) - perhaps better ground conductivity or some other reason. The advantage of VY2AJ is perhaps the composite from yesterday and today.
The simultaneous spots report shows the number of spots and computes the difference between VY2AJ and NY4G in terms of signal to noise ratio.
73 Ariel NY4G
3 Likes
As only vk2cow reported both signals, almost no comparison is possible between these data sets. I notice vk2cow managed to be about 42 km further away in one data set. How is that possible? I know Dimitris VK2COW as he lives about 20 km away from my home location. I doubt if he was operating two different receiving stations.
So what do these two data sets mean? I’m confused.
73 Andrew VK1DA/VK2UH
Is following method viable for efficiency comparison?
A. Creation of BASE Level transformer
- I created two identical 49:1 cores using FT140-43 and 21:3 turns 1.0mm magnet wire
- Then I measured losses of both cores using nanoVNA and S21 method.
- I established that each core has following losses
80m - 0.75dB
40m - 0.70dB
20m - 0.72dB
15m - 0.81dB
10m - 1.05dB
B. Measuring another transformer
- connect measured transformer to BASE Level transformer with known losses per band
- measure cobnined losses for each band
- substract known losses from measured losses to calculate actual losses of measured transformer.
- Prerequisites for measured transformer is the same impedance ratio 49:1, use of the same equipment and connecting wires.
73 de Marek
Thanks Ariel for running the WSPR test again with the couplers swapped and for the now detailed test documentation, I appreciate that. The only thing I may have overlooked is the antenna length used (1 or 2 half waves?).
When looking at the 2 recorded graphs, the discontinuous course is again noticeable. So the unknown player still seems to be fooling us and even after the 2nd test we cannot see that the difference of approximately 1 dB in the transmit power delivered to the antenna is clearly confirmed by the RBN spots.
As mentioned in the email from yesterday, a possible player could also be the effect of the antenna setup, especially with antennas that are opposite/parallel to each other.
With the distance of approx. 75 feet (about 1 lambda) specified in the test documentation, the following 2 effects are conceivable:
- Antenna diversity gain (spatial diversity in our case), possibly a little more pronounced due to the sloping terrain. However, that would hardly explain the discontinuous course
- Additional gain by antenna separation at a certain distance to use the neighboring antenna radiator as a reflector. That could be very well an explanation for the discontinuous course.
When looking at the skimmers listed for the two antennas, it is noticeable that some of them come from very different regions.
This would support the above thesis that the two antennas could radiate in the opposite direction due to the influence of a reflector. As a result, one antenna would point in the direction of sunrise and the other in the direction of sunset. Therefore it is conceivable that e.g. VK2COW could have received from one antenna on the long path and from the other on the short path.
The simplest solution for creating clarity in this coupler test using WSPR would be to carry out the test again with identical couplers (no matter which variant) for both antennas and possibly with exactly the same antenna setup.
Sorry Ariel, that would also make it clear that the key to solving this phenomenon (unfortunately) can only lie with you.