Index

 

G5RV design

G5RV re dipole

matching section

balun special

efficiency

Multiband G5RV antenne

(Published in CQ-QSO, # 9/10, 2006)

 

 

 

Introduction

 

The G5RV antenna was originally published by Louis Varney (G5RV) in Radio Communications in July 1984 and before that in 1966 in an earlier RSGB-Bulletin. As a matter of fact the G5RV is a (not too big) variation on the pre-war Collins multi-band antenna, which in turn is an improvement of the Levy antenna: an open-line fed dipole antenna from the early radio days.

Over time, the concept has been applied by many radio-hams, in many variations and 'improvements' and has been proven to be a useful antenna for many applications. Most improvements, both by amateurs and professionals proved to be minor variations on the theme without affecting basic behavior.

 

Because of its widespread use, the G5RV antenna system sometimes has taken on mythical proportions to some radio-hams, putting the design in the area of 'wonder-antenna's. It therefore seems a good idea to look again into Louis Varney's original design and analyze the antenna with contemporary means and methods.

 

 

The antenna concept

 

According to the original G5RV article, the antenna has been designed as a symmetrical, center fed dipool, with 'long-wire' behavior on the 20 m. radio-amateur band. On this wavelength, the open-line feeder is behaving like a 1 : 1 impedance transformer to connect to 75 Ohm twin-lead or 50 to 80 Ohm coaxial transmission line into the transceiver while ensuring low SWR. On all other HF radio-amateur frequencies the specific length of this open feeder line is acting as a transformer to reduce the sometimes wildly fluctuating antenna impedances.

 

From this description one may already deduce that the G5RV concept will only exhibit (very) low SWR on this 20 m. radio-amateur band and that an antenna tuner will be necessary to keep SWR below 1 : 2 on all other HF amateur frequencies, as required by modern transistor transceivers.

 

In his article Louis Varney presents his design as to be useful on all HF amateur wave lengths, i.e. 80, 40, 30, 20, 17, 15, 12 and 10 m., and also on 160 m. with the two feeder lines connected together, the antenna driven between this point and a good ground. In this situation, the G5RV is radiating as a small, top-loaded vertical antenna instead of a dipole, thereby exhibiting a low radiation resistance and therefore being very susceptible to all sorts of loss mechanisms. This type of application will be regarded as a different design and will not be part of this discussion.

 

According to Varney, antenna height is supposed not to be critical although the original antenna has been designed for 34 feet above ground (10,36 meters) and still exhibit good characteristics at 25 feet (7,6 m.).

 

Louis further tells us that a balun is not required between the open feeder line and a coaxial transmission line to the transceiver. He contradicts this last statement however when he writes:

"Under certain conditions, either due to the inherent "unbalanced-to-balanced" effect caused by the direct connection of a coaxial feeder to the base of the (balanced) matching section, or to pick-up of energy radiated by the antenna, a current may flow on the outside of the coaxial outer conductor. This effect may be considerably reduced, or eliminated, by winding the coaxial cable feeder into a coil of 8 to 10 turns about 6in in diameter immediately below the point of connection of the coaxial cable to the base of the matching section."

This clearly is a description of a 1 : 1 current balun, also known as a 'braiding-choke'.

 

In general it is always a good idea to place a balance-to-unbalance transformer (balun) at every change-over from a symmetrical to an a-symmetrical situation to keep currents from following unwanted paths, so we preferably apply a balun at the end of the symmetrical feeder line of the G5RV system when connecting into and a-symmetric coaxial cable or tuner / transceiver.

 

On the characteristic impedance of the coaxial transmission line to the transceiver Louis prefers this to be between 50 and 80 Ohm, as this line will see a 'rather high SWR' anyway, except for the 'basic' 20 m. band. He expects this high SWR not to result in high cable losses, provided 'this cable to be of good quality and reasonable length', f.i. no longer than 70' (21,3 m.). As has been shown in my article "Where does the power go" , cable losses can be rather high in badly terminated coaxial lines, so this 'extension line' should be as short as possible.

In figure 1 we see a diagram of the original G5RV antenna system.

 

 

 

Figure 1: Original G5RV antenna

 

 

On the total antenna width of 102' or 31,1 m. Louis Varney comments that on each end 10' or 3 m. should not necessarily be in a horizontal position, presumably because most of the radiation comes from the middle 2/3 rd of the antenna when resonant at half wavelength. This may be true for these half-waves, but no more in case of the G5RV since this will accommodate more than half a wave length on multiple amateur bands, all contributing to the total radiation pattern.

 

Antenna characteristics

 

Fortunately we are currently living in interesting times, with many antenna calculation programs at our disposal and computers to perform calculations for us. For this analysis I have been using the antenna-modeling program EZNEC and transmission-line modeling program TLW. All further calculations have been performed using elementary math's, with EXCEL to perform the tiring repetitions.

 

To obtain a fair impression, I will compare the G5RV behavior with dipole antenna's, each cut for resonance on the particular HF amateur band. Antenna's are modeled at 10 m. above 'average ground' (5 mS, ε = 13). Since each dipole is situated at a fixed antenna height (a variable number of wavelength per amateur band), the antenna impedance will vary resulting in a variable SWR. Lets look at the first table (table 1).

 

 

 

 

Dipole

 

 

G5RV

 

 

 

f

gain

elevation

SWR

gain

elevation

SWR

SWR

MHz

dBi

degrees

re 50

dBi

degrees

re 50

re 300

3,65

7,15

90

1,3

6,87

90

108,6

32,6

7,05

6,46

68

1,4

7,24

68

53,6

9,5

10,15

6,32

42

1,6

8,66

42

83,5

14,1

14,175

7,3

29,5

1,4

6,56

29,5

2,7

3,3

18,11

7,95

23

1,2

9,76

23

60,4

10,1

21,225

7,56

19,5

1,4

10,93

19,5

70,0

12,9

24,95

7,14

17

1,6

10,65

15

11,4

3,4

28,75

7,43

14,5

1,5

10,52

15

53,5

8,9

 

Tabel 1: Maximum gain (at elevation angle) re 'standard dipole'

 

 

In table 1 some interesting features may be noticed.

Firstly, it appears that a 'standard' resonant dipole antenna is not bad at all: with a maximum gain of around 7 dBi at each HF amateur frequency band, and a low enough SWR to allow direct connection to a transistorized transceiver, it is a simple, cheap and reliable antenna to start-off with. A number of those standard dipoles may be connected in parallel on the same balun, as only the resonating dipole will exhibit a low connection impedance and will carry most of the current (and therefore perform most of the radiation). Such a dipole assembly is called a cob-web or cats-whiskers antenna system.

 

Secondly, when comparing maximum antenna gain of the dipole and G5RV antenna, one will notice a comparable gain below 14 MHz. and more gain above this frequency for the G5RV, because more wavelengths fit and contribute.

 

A third observations may be made when comparing elevation angles; these are remarkably similar for both antenna's. This is no coincidences but stems from the fact that both are situated at the same height above the same ground; as the antenna radiation pattern is the vector summation of both direct and (ground-)reflected radiation, which is the same for both antenna's, the elevation angels are also the same. The azimuth patterns will be somewhat different though, again for the multi-wave fitting at the G5RV. This will result in a multi-lobe pattern, with deep nulls in between.

 

Looking at SWR re 50 Ohm in the 4th and 7th column, one will notice big differences, except for 14,175 MHz. the original design frequency for the G5RV. Although not exactly very low, the SWR figures with respect to 300 Ohm (last column) are already looking somewhat better. This tells us that a non-resonating dipole antenna in general may be better connected through high impedance (open) transmission line, resulting in lower SWR (and so lower cable loss) through mismatch, on top of the already low losses of these cables when characteristically terminated.

 

 

The transformer section

 

Let's take a better look at the second characteristic element of the G5RV system. This has been modeled as TV-line with a characteristic impedance of 300 Ohm, 8,5 m. long, with a velocity factor 0,82. Cable loss when terminated into its characteristic impedance is comfortably low at 0,6 dB / 100 m @ 3,5 MHz. when compared to good quality coaxial cable RG213 (1,2 dB / 100 m @ 3,5 MHz.). It is interesting to notice the effects on cable loss and SWR of this transforming section over frequency with this antenna. Lets look at table 2.

 

 

 

G5RV antenna

 

After 8,5 m. 300 Ohm

 

f

SWR

SWR

loss

SWR

SWR

SWR

MHz

re 300

re 50

dB

re 300

re 50

re 70

3,65

32,6

108,6

1,12

25,2

4,2

5,9

7,05

9,5

53,6

0,23

9,1

5,4

4,7

10,15

14,1

83,5

0,71

12,0

38,2

28,0

14,175

3,3

2,7

0,19

3,2

2,6

2,0

18,11

10,1

60,4

0,5

9,1

26,9

19,7

21,225

12,9

70,0

0,86

10,6

8,8

7,3

24,95

3,4

11,4

0,24

3,2

1,9

1,4

28,75

8,9

53,5

0,63

7,8

45,2

32,3

 

Table 2: The transformer section

 

 

When we start regarding the fourth column, we find that in spite of the low loss cable the transformer section is showing some loss. This loss is in closer examination related to SWR at cable input, as in the second column. The SWR related loss mechanism is explained in more details in the article "Where does the power go". Highest loss (1,12 dB) is at the 80 m. amateur band with highest SWR (108,6). Although this is still an acceptable loss figure, it still means that at an input of 100 Watt, 23 Watt will be dissipated along the line and be radiated as heat.

When comparing 5th and 2nd column we see that SWR re 300 Ohm is hardly changing from the beginning of the line to the end. This is a demonstration that SWR is constant along a lossless transmission line, regardless of the impedance transformation. We do see however small SWR differences between the two columns as these are related to line losses; the higher the losses, the higher the SWR differences.

 

The operation of the transformer section is showing very nicely when comparing SWR re 50 Ohm at the beginning vs. at the end of the transmission line (3 rd vs. 6 th column). The transformer section performed a good job that dramatically shows at 3,65 MHz.: SWR has changed from 1 : 108,6 tot 1 : 4,2. Also on other amateur frequencies this improvement is evident although to a somewhat lesser extent.

When comparing the last two columns, we see that indeed we may connect 50 or 70 Ohm transmission lines, as differences are small.

 

The balun section

 

As stated above, it is good practice to install a balun at the intersection of the transforming feeder line and connecting coax or to an a-symmetric tuner or transceiver. In a different article, baluns have been described in more detail, see "Baluns". In design examples this component usually is presented for a 50 Ohm environment. In case of the G5RV situation however, this is impedance is variable and depending on the selected amateur frequency. Let's see how this works out in table 3.

 

 

 

f

SWR

Rs

Xs

Zt

MHz

re 50

Ohm

Ohm

Ohm

3,65

4,2

11,9

-4,6

12,8

7,05

5,4

35,4

-77,8

85,5

10,15

38,2

51,4

305,3

309,6

14,175

2,6

96,45

-52,1

109,6

18,11

26,9

62,1

-278,0

284,9

21,225

8,8

31,7

102,7

107,5

24,95

1,9

92,8

16,8

94,3

28,75

45,2

856

1096,0

1390,7

 

Tabel 3: Impedances at the intersection.

 

 

In table 3 we find the SWR figures from table 2, this time decomposed into its real and imaginary constituents and total impedance. Looking at this last column one immediately finds that the balun design should differ considerably from 'standard' 50 Ohm values. As stated in the mentioned article, the parallel impedance of the balun should be at least four times the system impedance. Considering table 3 we find that when this balun impedance is around 1200 Ohm @ 10,15 MHz., we cover most situations although at 28 MHz. the situation may be somewhat critical.

Starting at this 1200 Ohm @ 10,15 MHz., the balun should have an inductance of around 19 μH. When using a popular 36 mm. toroide of 4C65 type ferrite (Ferroxcube) or '61' (FairRite, Amidon), having a core factor of 170 nH/n2, we need n = sqrt(19 / 0,17) = 11 turns of coaxial cable for an adequate current balun. This will be difficult when using RG58-type of coax, so we need a somewhat bigger toroide. As an alternative one could use a stack of two such 36 mm. cores, effectively doubling the core factor to arrive at 8 turns of RG58, which presents no problem anymore. Using these two cores, also the allowable self-dissipation is doubled to 8 Watts for a maximum core temperature-rise of 30 C.

 

System efficiency

 

Total antenna system efficiency is related to various factors as described earlier. Let's now calculate loss factors and antenna gain to get an overall impression of the G5RV system performance as related to our 'standard dipoles'-set in table 4. This time I calculated loss factors in heat (Watts) related to a system input of 100 Watt.

 

 

 

 

 

transformer

 

 

 

result

 

dipole

G5RV

section

tuner

balun

power

re tuned

f

max. gain

max. gain

loss

loss

loss

in antenna

dipole

MHz

dBi

dBi

W

W

W

W

S-punt

3,65

7,15

6,87

22,5

1,14

0,0

76,4

-0,2

7,05

6,46

7,24

5,1

1,37

0,3

93,3

0,1

10,15

6,32

8,66

13,9

4,28

3,6

78,2

0,2

14,175

7,3

6,56

4,2

0,69

0,4

94,7

-0,1

18,11

7,95

9,76

10,2

2,50

4,0

83,3

0,2

21,225

7,56

10,93

17,3

1,83

2,0

78,9

0,4

24,95

7,14

10,65

5,3

0,69

0,8

93,2

0,6

28,75

7,43

10,52

10,2

3,84

20,8

65,2

0,2

 

Table 4: Total G5RV efficiency re 'standard dipole'.

 

 

Let's start off looking into 'safety factors' by regarding column six, power lost in the balun. As we selected the two-ring, 36 mm. version, the component is allowed to dissipate up to 8 Watt, which will not be exceeded except for 28,75 MHz. Despite the transforming section, the total system impedance is still too high on this frequency, resulting in a too high voltage and therefore too much power dissipation in this balun, already at 100 Watt total input power. Keep in mind that this calculation being performed for continuous power, as in frequency modulation. In CW and SSB, average power is around three tot six times lower so the 100 Watt input power could still be allowed, although with some care.

 

In the fourth and fifth column we find power dissipated in the 8,5 m., 300 Ohm TV-line section and tuner, the latter being selected as a low-pass, L-type, with good quality inductor (Q = 200). In the seventh column we find the power that will finally be radiated from the (lossless) antenna.

How this will end-up at the receiving site, depends very much on 'conditions' and the direction of the receiving station with regards to the antenna position. As a first and rough comparison, I have related G5RV efficiency to our standard dipole, in terms of perceived signal strength (S-points), both in the direction of maximum gain (different angels for both antenna's). In this comparison (last column) all system losses and antenna gains have been taken into account except for those in the final coaxial cable to the TRX. The losses of the latter depend on the type of cable (unknown) and cable length (unknown) and may not entirely be neglected anymore when more than a few meters long.

 

From this comparison it is clear the G5RV antenna is a compact and efficient 'general purpose' antenna on most HF amateur frequencies.

 

In the process of calculations for this article I also performed some modeling on G5RV-variations, for instance the ZR1DQ proposal. All these variations yield about the same results as the basic G5RV and/or made trade-off's to favor one radio-amateur band over the other.

 

Bob J. van Donselaar

mailto:ON9CVD@AMSAT.org