Barry, K3EUI, posted some interesting measurements of his MyAntennas EFHW which he described with some useful detail:
I have been experimenting with a “MYANTENNA” 130 foot long “end-fed” with the 49:1 UNUN*
I replaced their 130 ft antenna wire with a heavier gauge #12 stranded insulated wire (I had a few hundred feet).*
This is classified as a ONE-HALF wavelength antenna on 80m, hence the need for the 49:1 UNUN to transform 50 ohm (coax) to a few thousand ohms*
It has resonances on the other ham bands (harmonically related) but I wanted it mostly for 80m.
One leg runs about 60 ft horizontally to a tree and then the next 70 ft makes a 90 degree bend (to fit into my yard) still horizontal.*
At this time I removed a 15 ft “counterpoise” wire on the GND side of the UNUN.
I will try it later this week (after the snow) as a “sloper” or an Inverted V up to a tall fir tree.
It is only about 20 ft above ground now (with 4 inches snow) for NVIS prop, and fed with 70 ft of RG213 coax (50 ohm) with a RF choke on the coax feed line 10 ft from UNUN (the counterpoise?) and another RF choke just as the coax enters the shack.
His VSWR curve is interesting, a minimum at source end of about 1.32 @ 3.66MHz as built and measured.
Minimum VSWR is about 1.32 @ 3.66MHz.
There is no discussion on preservation of the common mode current path during measurement (see Disturbing the thing being measured – coax line.) If this is ignored, it is often because its importance to valid measurement is not understood.
It would have been interesting to see a plot of R,X at the feed point (with with the coax / counterpoise etc in situ).
He also posted a Phase of s11 graph.
A reasonable guess is that both measurements are at the source end of 21.3m of RG213. At 3.66MHz, about 142° at 3.66MHz.
So, can we estimate feed point impedance, ie at the EFHW transformer but of course, with the coax / counterpoise etc in situ.
We can estimate |Τ| from VSWR @ 3.66MHz, it is 0.138, and reading the phase plot, phase of Τ is about -105°, so Τ=0.138∠-105° .
A rough estimate is that the measurement had an additional phase delay of 2*142°, so phase of Τ at the load is should be about -105+2*142=179°. This is very close to 180°, and so Z should be very close to 50/1.32+j0=38+j0Ω. This estimate does not account for the line loss or the fact it’s Zo departs a little from 50+j0Ω.
A better estimate come from a more detailed transmission line model using TLLC.
Above, the input form for the calculation.
The extract from the results above shows the conditions at both ends of the line, and importantly that the estimated Τ at the load is 0.145∠178.7°… close to the rough estimate.
Above is the estimated Zload, pretty close to the rough estimate. Importantly, we can see the estimate of line loss under the mismatched conditions. The line loss under mismatch is a tiny bit less than matched line loss. This is not an error… though lots of hams would insist that VSWR always degrades line loss, the ARRL graph shows so.
Although analysed in detail at only the frequency of minimum VSWR on only one band (80m), the results look good, the measurements seem credible, and they speak well of the MyAntennas EFHW used, the modifications and feed arrangement. Being quite low to the ground, the low feed point impedance is not too surprising, and that will exacerbate ground losses (not captured in the transmission line efficiency figure above).