Optimising a G5RV with hybrid feed
(Varney 1958) described his G5RV antenna in two forms, one with tuned feeders, and the more popular form with hybrid feed consisting of a so-called matching section of open wire line and then an arbitrary length of lower Zo coax or twin to the transmitter.
(Duffy 2005) showed that the hybrid feed configuration is susceptible to high losses in the low Zo line as it is often longish, is relatively high loss line and operates with standing waves. Varney did offer two options for the low Zo line:
any length 72Ω twin or coax.
A real world example
Lets look at measurement of a real antenna at VK2OMD, broadly typical of the G5RV. The antenna measured is a G5RV of 2mm HDC rigged in Inverted V form, 11m height at the apex and around 8m at the ends. The feed line is 2mm diameter copper spaced 50mm with occasional plastic insulators.
The measurements are to some extent dependent on the environment, and whilst there will be variation from one implementation to another, the measurements provide a basis for exposing the configuration challenge.
Above is a plot of VSWR50 looking into 0.5m of RG142 and a Guanella current balun, then an open wire section that is similar to that used in the hybrid feed configuration. This is the character of the impedance presented at the lower end of the ‘matching' section.
Let us take two spot frequencies in this range and calculate the loss in some different types of low Zo line of 15m length.
|Loss 15m Belden 8222 (#20 twin 72Ω) (dB)||1.1||0.98|
|Loss 15m Belden 8210 (#13 twin 72Ω) (dB)||0.52||0.53|
|Loss 15m Belden 9212 (75Ω RG-11/U) (dB)||0.30||0.28|
|Loss 15m Belden 8259 (50Ω RG-58/AU) (dB)||0.53||0.96|
|Loss 15m Belden 9258 (50Ω RG-8/X) (dB)||0.38||0.40|
|Loss 15m Belden 8267 (50Ω RG-213) (dB)||0.20||0.37|
The above table gives loss in several ‘low Zo' feed line options, at the frequency of minimum VSWR50 (3.5MHz), and just 200kHz higher at 3.7MHz.
We can observe that the feed line loss away from minimum VSWR50 is worse for the lossier feed line options, the data demonstrates the advantage in using lower loss feed line, AND tuning the system so VSWR50 is minimum at the desired part of the band.
Popular discussion on online fora suggests that the ‘original' 72Ω twin which Varney mentions and which is no longer readily available is much better than coax. The table shows otherwise and that the recommendation for Unobtainium is just uninformed ham nonsense… the best option at minimum VSWR50 from those listed is RG-213, RG-8/X is nearly as good.
It can be seen from the plot above that the low VSWR region is narrower on 80m than on the higher bands, and since losses on 80m may be quite high, tuning the system so VSWR50 is minimum at the desired part of the band should be a priority.
(Duffy 2006) described a method for optimising a G5RV using very simple equipment, and gave the rationale for tuning.
Essentially it involved two steps:
- tuning the matching section to a half wavelength (electrically) at 14.15MHz;
- adjusting the dipole length to set the minimum VSWR50 region on the low Zo feed line to the preferred part of the 80m band.
Both steps can be performed with a modern antenna analyser.
Note that although Varney specified the length of the matching section as 34′, he later explains that it is a half wavelength on 20m, and it is the electrical length that is key to its proper operation. Hams commonly pay attention to the 34′ length than the more important half wavelength at 14.15MHz, and the relationship between physical length and electrical length (known as the velocity factor) varies..
A scanning analyser makes for a convenient overall check and fine tuning.
At the completion of Step 2, make a scan from 2 to 30MHz and check that the other bands ‘drop in' as desired. If necessary, some minor tweaks can be made by complementary adjustments to dipole length and matching section length to move the higher bands around relative to the 80m sweet spot.
- Duffy, O. 2001. RF Transmission Line Loss Calculator (TLLC). VK1OD.net (offline).
- ———. Oct 2005. Feeding a G5RV. VK1OD.net (offline).
- ———. May 2006. Optimising a G5RV. VK1OD.net (offline).
- ———. 2008. A model of a practical Guanella 1:1 balun. https://owenduffy.net/files/GuanellaBalun01.pdf.
- ———. 2010. Additional loss due to VSWR. VK1OD.net (offline).
- ———. 2011. Estimating parameters of two wire transmission lines
- Varney, Louis. July 1958. An effective multi-band aerial of simple construction In RSGB Bulletin July 1958.