3
$\begingroup$

I've build a trap dipole for 20/40m bands. Traps are made of a coil wound on an insulator of RG213 coax and 3 kV NP0 capacitors. The impedance measured with EU1KY antenna analyzer is following:

enter image description here

This is a single trap, before and after putting a heatshrink. The final version of this one is for 14.080 MHz. The second one turned out a bit different, the resonant frequency is 14.150 MHz.

Here is an SWR plot of a final antenna:

enter image description here

The mast is non-conductive, 10 meter high. The balun is 7 turns of RG58 on FT140-43 core. The active impedance is > 1000 Ohm from 80 to 15 meters. The antenna is used in inverted-V configuration.

The antenna was tested using 100W CW. It works OK. However I was surprised to get such a narrow band and SWR >= 1.5 on 20 meters. What brings a question - is it an expected result for trap dipoles? I always thought that it supposed to be more narrowband on lowest band, because traps work as loading coils there. Can it indicate that something is wrong with the antenna?

Improve this question
$\endgroup$
8
  • 1
    $\begingroup$ What is the "reference plane" for the SWR measurement? Do the graphs show the SWR at the antenna feedpoint, or do they show the SWR at the transmitter end of the feedline? $\endgroup$
    – Brian K1LI
    Jun 30, 2020 at 11:21
  • 1
    $\begingroup$ @BrianK1LI the measurement was done over 10 meters of RG58 cable and then re-checked using several other lengths. The plot doesn't change much with different lengths of the coax. I think it's more or less accurate. I used this method before for single band inverted-V and a fan dipole under same conditions, which gave results that are close to the theory. $\endgroup$
    – Aleksander Alekseev - R2AUK
    Jun 30, 2020 at 17:12
  • 1
    $\begingroup$ @BrianK1LI Still this is a good point. I will re-check the plot using λ/2 RG213 coax, but currently I'm out of RG213 and the delivery to my QTH will take some time. $\endgroup$
    – Aleksander Alekseev - R2AUK
    Jun 30, 2020 at 17:18
  • 1
    $\begingroup$ You should be able to calibrate the analyzer to eliminate the effect of the feedline in the measurement. This is done by placing the short / open / reference load at the end of the feedline during the calibration procedure. Then, measurements made at the input to the feedline indicate the values at the antenna. $\endgroup$
    – Brian K1LI
    Jul 1, 2020 at 18:06
  • 1
    $\begingroup$ @brian at HF it shouldn't matter too much to measure through the coax, if we're only interested in SWR. And ultimately the SWR seen by the rig is what we're interested in anyway. $\endgroup$
    – tomnexus
    Jul 2, 2020 at 1:26

3 Answers 3

Reset to default
3
$\begingroup$

I don't know how much of an impact it might make, but you might try moving the trap resonance to be outside the 20m band.

Trap losses are highest at the trap's resonant frequency because at this frequency the current circulating between the inductance and capacitance is at a maximum. Thus, ohmic losses within the trap are maximized.

At resonance is also where the rate of change of impedance with respect to frequency is highest. Designing the trap to be resonant within the 20m band means the trap impedance will vary rapidly over the band, which could explain the narrow SWR bandwidth.

To work as a trap you don't need the impedance to be infinite: only high enough to sufficiently isolate the following length of the antenna. Moving the resonant frequency of the trap may lower its impedance, but you can compensate for that by adjusting the length of the antenna. On the up side, losses may be lower and SWR bandwidth may be improved.

Further reading:

Improve this answer
$\endgroup$
1
$\begingroup$

My quick opinion is that this looks reasonable. A trapped dipole will likely have a smaller bandwidth than a regular dipole. A short simulation is what is required to be sure.

You're getting 5% bandwidth at 40 m (the traps don't load it very much at 40 m) which is normal. And to be generous 1% at 20 m. The traps are working on the high impedance tips of the dipole, the frequency range where they're high enough impedance is quite small.

Do check that the traps are tuned to the same frequency, it won't help if they aren't.

The impedance never goes through 50 Ohms, so there's a further reduction in usable bandwidth there. With a VNA, calibrated to the end of the cable, you could try to figure out whether it's high or low, and put in the appropriate parallel cap or series inductor (and re-trim the antenna).

Improve this answer
$\endgroup$
1
  • $\begingroup$ "Do check that the traps are tuned to the same frequency" - as I mentioned, they are tuned to close frequencies, 14.080 MHz and 14.150 MHz. I don't know if 70 kHz difference matters. Anyway I'm going to build second antenna with traps tunet to exactly the same frequency and compare results. $\endgroup$
    – Aleksander Alekseev - R2AUK
    Jul 2, 2020 at 12:25
0
$\begingroup$

OK, I've checked 5 or 6 types of traps made of different materials, tuned to the same or slightly different frequencies, tuned right on the band, below the band and above the band. Most of them were not better than the first pair. Tuning above the band doesn't work. Traps made of RG58 coax work best. You get SWR 1:1 on both bands, 320 kHz bandwidth on 20m and 225 kHz bandwidth on 40m with SWR <= 2. According to WSPR report the antenna is no less effective than my previous fan dipole was. The article An Attic Coaxial-Cable Trap Dipole for 10, 15, 20, 30, 40, and 80 Meters by John DeGood, NU3E explains how to make such traps in great detail. Turned out these traps are tuned to the slightly different frequencies below the band.

Answering my own question: NO, this is not an expected result. Although such antenna kind of works, you can make a much better one.

Improve this answer
$\endgroup$

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .