

40 Meter Yagi Antennas and StacksJay Terleski, WX0BWhen I received my first license as a novice in 1962, I was 12 years old. I could only afford to put up a dipole for 40 meters. This worked out pretty well but I really admired the "big guns" who actually had a beam on 40 meters. There weren’t too many of them, maybe one or two in the Warren, Ohio area. But things have changed. 40meter beams on top of a crankup tower with a tribander have become the norm. Nearly anyone can put up one of these 40 meter "shorty 40s". Companies like Cushcraft, Force 12, and M˛, have done an excellent job of making shorty 40s for us to enjoy. Goal of this articleThis article is aimed at describing the performance of these 40meter beams singly and in a stack as well as comparing them to a full sized 3element 40meter beam. We shall also investigate a new breed of 40meter 2element beam using dual driven and phased elements and compare them to the 3element full sized 40meter beam. In all cases I will use NEC 2 models. Some BackgroundIn modeling a shortened 40meter element, I have taken the safe way out, to make a "generic" model that uses an inductor and a resistor as a load to simulate a shortened 40meter beam. I did this intentionally to not take sides in the linear loading vs. inductive loaded element debate. My element model simulates all of these antennaloading types in its efficiency. The loaded element I used provides an overall efficiency of 72%, which is reasonable for the size of these elements. As a matter of fact, several sources are available in the literature, which will back up this efficiency number for shortened beams. I will just note that several companies are working on improving this efficiency, such as Force 12 with their new linear loading technique, and Dave Padrick, W6ANR with his high Q coil designs with high current connections. In all cases, we will model these antennas over what is considered average ground (.05 S/M and .30 dialectic constant) with highaccuracy, real ground characteristics in NEC 2. The antennas will be modeled at 70 feet and 140 feet above this ground. In other words, about .5 WL and 1 WL above ground. The elements are Aluminum 6061T6; their resistive losses are included in the model. The 3Element 40 Meter BeamPutting these models together is a subject I won’t go into detail here for the sake of saving space, but I will post them on my website for your use in running your own comparisons. The full sized 3element 40meter beam is placed on a 46foot boom. The reflector is 22 feet behind the driven element, and the director is 24 feet in front of it. Figure 1, 3 element 40 meter full size Yagi model Figure 1 has shows the model of a full size 3 element Yagi at 140 feet above ground. Its pattern is shown for three frequencies at of 7.0, 7.15, and 7.3 MHz in Figure 2.
2a 2b 2c Figure 2, 3 element full sized 40 meter Yagi at 140'
As figure 2 demonstrates, the 3element beam has very good gain and front to rear pattern over the 40meter band. In Table 1 we tabulate the source impedance and SWR that can be expected at the feedpoint of this antenna, gain, takeoff angle, and front to rear (F/R) ratio. TABLE 1 Yagi 3 elements full sized
Let’s graph the SWR over the band to get a picture of the bandwidth. Figure 3 shows an excellent SWR bandwidth with this beam. It actually matches well to a 50ohm cable. Figure 3 SWR of a 3element Yagi
The 2Element Shortened 40 Meter BeamThe model of the shorty 40 I built is shown below. It has 22foot separation and a driven element that is 45.4 feet long with loading coils of 17.25 uH and a series resistance of 3.5 ohms in the coils. The coils are located 75% out on each element. The reflector has the same loading coils and is 47.4 feet long. As in all commercially made shortened beams, the useful bandwidth of this design should be narrower than the full sized 40 we just modeled. I optimized this beam to be close to a 1:1 SWR at 7.15 MHz. At this frequency, it shows an impedance of 56 ohms, so we chose to normalize or match this impedance at this frequency. Figure 4 Shorty 40 model We now need to run the plots over the frequencies of the 3element beam to investigate how this antenna is performing. Figure 5 shows the patterns over the range of 77.3 MHz. I ran more data plots to demonstrate the more sensitive patterns of this class of shortened antennas. 5a 5b 5c 5d 5e 5f Figure 5  2 element Shorty 40 @ 140' The patterns of figure 5 are summarized in Table 2. Comparing to the 3element full sized array, we can see from the data that the beams do have a smaller 2:1 bandwidth, and a more sensitive gain and F/R ratios. The gain varies a little less than 1 dB over this 2:1 range. And F/R varies 6 dB over the 200 kHz range. The F/R numbers have been designated as a number / number meaning the upper rear lobe over (/) the lower rear lobe to include more information into the table. Notice that the takeoff angle has remained the same as that of the 3element beam. This indicates that this function is basically a variable of how high the beam is above ground.
Table 2 Summary of patterns of figures 5 af.
For the purposes of this paper, we will not adjust the shortened antenna for the CW or SSB side of the band. Just bear in mind that the data in Table 2 can be shifted up or down depending on the side of the band that is more important to the user. The commercial antenna manufactures supply the adjustment information to do just this for their shorty 40s. Figure 6 shows the SWR over the band for the shorty 40 model. Figure 6 Shorty 40 SWR Curve This SWR curve is characteristic of what we see with commercially available shortened parasitic Yagi antennas on the market. It has a little more than 200 kHz bandwidth from the 2:1 SWR points. Typically, the antenna is optimized for the low end of the band or the high end of the band. I set my antennas’ resonant frequency for 7.1 MHz so they have a 2:1 SWR at 7.0 MHz. The SWR is higher in the SSB portion of the band but this can be compensated with the pinetwork output of a tube amp. I find my tube amps still trip out easily, so I have built an L network to put inline with the feedline for my shorty 40m beams for SSB contests. It gives the amps a perfect 1:1 SWR higher in the band, and everyone is happier.
Stacking a pair of shortened 40 meter antennasThis next section will investigate what happens when a pair of these shorty 40s is stacked 1/2 WL apart in a vertical stack. Figure 7 Shorty 40 Stack model Our model simply copies all the information to create a second beam that is 70 feet high. We now have a stack of two 2element 40s, 140 feet over 70 feet. See Figure 7. I connected the stack together with two 3/4 wave 75ohm transmission lines and feed them at the center of these two lines. This will give us a good indication of how the two antennas will react to each other, and will give us a better feel for the feed impedance of the array as a stack.8a 8b 8c 8d 8e 8f Figure 8 Shorty 40 Stack Patterns
Figure 9 SWR – Shorty 40 Stack In figure 8, Notice the difference in the patterns compared to a single antenna. The upper lobes have all but disappeared and the overall gain has been increased in the main lobe. This is one huge benefit of stacking antennas, the undesired lobes are suppressed due to the interference patterns of a properly spaced stack. The SWR at the center of our transmission lines is plotted in figure 9. We are seeing the combined impedance of nearly 35 ohms at 7.15 MHz. I normalized the SWR curve for 35 ohms for this plot. If we were to tune these antennas 50 kHz lower we could cover the whole band with a very good SWR.
Table 3 summarizes the characteristics of this stack. Notice the very peaky F/R data, which changes from a high of 20dB to a low of 5 dB.
Table 3 Shorty 40 stack 140/70 feet
The Dual Driven 2Element 40 Meter BeamThe next antenna we shall investigate is the next step in the evolution of the 2element beam. The model I used is the exact tapered model of the real beam developed by CalAv Labs, of Tucson, Arizona. It basically is a 16foot boom, with two full size elements at the ends. Both elements are fed with a phase difference but with equal magnitudes of current. Both have adjustable hairpins; one hairpin contains an integrated, highpower wideband balun. Figure 10 shows the basic model of this antenna. The transmission line, the line running between the two elements, creates the phase difference between the two elements. The antenna is fed at the right hand element in this figure.
Figure 10 CalAv 2d40 model
11a
11b 11c 11d 11e 11f Figure 11af, CalAv 2d40 at 70 feet Figure 11 shows that the Gain and the F/R ratio stays very smooth across the whole band. This is typical of a dual driven element design, and offers quite an advantage over the parasitic arrays due to the nonpeakiness of the gain and F/R curves. Table 4 summarizes the plots of Figure 11. Table 4 2d40 CalAv 2element dual driven beam @ 70 feet
SWR is very good as well Figure 12 is a plot of SWR. The higher end of the SWR curve tapers off slower than the lower end. I plotted the SWR up to 7.4 MHz to see how it operated at this frequency. Instead of setting this beam up for 7.15 MHz we could set it up at 7.1 MHz and almost cover the whole band with an SWR of 2:1 or less. Figure 12 SWR of Dual Driven 40m beam
Moving this antenna up to 140 feet and repeating the plots over frequency, we have Figure 13 af. 13a 13b 13c 13d 13e 13f Figure 13 a – f, CalAv 2d40 at 140 feet
Table 5 CalAv, 2d40 at 140 feet
Table 5 displays the results of the plots of figure 13. Again the single antenna has the characteristics of the double lobe, like all of the previous antennas at 140 feet, but notice the great pattern. Also we see that the SWR curve has moved up slightly due to taking the antenna to 140 feet. Figure 14 is a plot of the SWR of this antenna at 140 feet. It appears that we can adjust this antenna for a 7.1 MHz resonance and, almost cover the whole band with an SWR of less than 2:1.
Figure 14 SWR at 140 feet.
The Stacked 2Element 40 meter Dual Driven Array ModelNow comes the fun part. Lets stack them 140 over 70 feet! Figure 15 CalAv Stack 140 over 70 feet As seen in Figure 15, we have two of these antennas fed at the center with a 3/4 wavelength of 75ohm cable. We can now run the same plots (figure 16) over the frequency range we have been using, and put the data into table 6.
16a 16b 16c16d 16e 16f Figure 16 af, CalAv 2d40 stack 140/70 feet Table 6 Stacked 2D40s 140/70 feet
The results listed in Table 6 show that the stacked array has maintained its smooth gain and F/R ratios. The gain is increased more than 2 dB over a single antenna and, surprisingly, the SWR curve has narrowed by 50 kHz over the band as compared to the single antennas curves. Figure 17 examines the SWR from 7.0 to 7.4 MHz. Notice that the resonance point has dropped 50 kHz as well. Because of mutual coupling, the SWR plot seems to have changed by 50 kHz lower. But this may be a function of this models transmission line feeder. If this really happens, it will be interesting to document. Figure 17 SWR of stacked 2d40s Let’s do one last test on this antenna stack and see what happens to the pattern when we intentionally drive the antennas 180 degrees out of phase or the BOP mode. This is a very good technique for raising considerably the takeoff angle of the main lobe. This is a very desirable capability on the 40meter band.
I have reversed the feeding of the lower antenna by 180 degrees in Figure 18. Figure 18 CalAv 2d40 stack fed outofphase 180 degrees I will not run through all the frequencies in this article, but I will plot this pattern at 7.05 MHz. We see that the SWR is very good because the drive impedance is 42.2 + J 11.98 ohms at the center of our matching transmission lines. This would be an SWR (50 ohm system) = 1.363. As can be seen in the plot of Figure 18 we have a takeoff angle of 41 degrees and a gain of 12.57 dBi. Notice he F/R ratio is excellent and the undesired lobes are substantially suppressed. ConclusionsWe have now investigated three types of 40 meter Yagi antenna systems that are being used by Contesters and Dxers. The full sized 3element Yagi, shorty 40s, and we have introduced a dualdriven array that is now commercially available. Table 7 is a summary of the pattern characteristics, which we have observed with this study.
Table 7 Main Characteristics of single antennas at 140 feet and stacks 140/70 feet
The 3element full sized Yagi is a large undertaking for any ham. It requires a substantial investment and is very heavy, reaching 300 pounds (136 kg). It will require a large tower and rotor to keep it up and turning, but it will deliver excellent performance over the whole 40meter band. Although not modeled, a stack of two of these antennas is an awesome array to behold and command. The shorty 40s commercially available have reasonable performance as a single antenna and, as expected, have less gain compared to the larger 3element Yagi. A stack of shorty 40s is very interesting in that the gain is very much the same or more than a full sized 3 element Yagi over the whole band. SWR suffers some as a single antenna, but flattens out in a stack. F/R patterns are very peaky care must be taken to tune them up. But a consumer would rate this type of stacked array on 40 meters a "Best Buy". I will point out that there can be significant issues with shorty 40s, depending on how the loading is accomplished. With linear loading techniques, the loading wires can radiate in undesired directions, which the modeling software cannot predict due to issues with closely spaced wires. Using inductors to load the antennas will also produce losses and undesired radiation, which the simulation software cannot predict. This is because the software treats an inductor as a lumped constant. In reality, the coils radiate, spoiling the pattern to some extent. I tried to at least model the losses with my 3.5ohm resistors at each load. This is a value that has been measured by several sources as the real resistive RF loss in one commercially made antenna’s inductors. Finally, we introduced a new design using two driven elements. This type of array has been around for some time, and is commonly called an HB9CV. Until now, none of these arrays was commercially available. This antenna uses two full size elements on a 16foot boom. It weighs 150 pounds (68 kg), which is half the weight of the full size 3element Yagi. It is more viable and manageable on a tower, and the short boom makes it attractive to owners of smaller towers and rotors. As a single 2element antenna, its performance closely approaches that of 3element Yagi. It is only .5 dB down and has a F/R ratio that is higher. These antennas should be very quiet because interference and noise from the rear will be suppressed. What is really nice about these antennas is that the F/R and Gain remain constant throughout the band. This is an improvement over the peaky performance of the shorty 40s. As a tradeoff to great F/R and Gain, SWR bandwidth seems to want to remain at 200 kHz due to the close spacing of the elements. Stacking them doesn’t seem to widen the SWR response as it does with the shorty 40s. But this may be only a function in the software model, more on this later. It certainly can be compensated for with an L network if needed. And as a tradeoff, SWR is one I find more attractive than sacrificing gain or F/R pattern. A stack of these antennas is going to be 1.31.5 dB higher in gain than a full size 3 element Yagi, and have a F/R ratio very similar to the big Yagi. The ease of maintenance and rotating them make them attractive to owners of midsized stations. These antennas also should appeal to the single 40meter antenna market, because even alone, they will operate extremely well at 70 feet. Another factor that makes them attractive is that both driven elements can be tuned on the tower. This unique tuning system is possible because the tuning adjustments for each element can be reached from the tower with a tuning rod that is supplied with the antenna. This is another benefit of having close spaced elements. I expect to see a new generation of shortboom, dual driven element antennas in the near future, perhaps with shortened elements on 80/75 meters. As an aside, I shall be putting up two of these antennas in the next month or two, and shall be taking down the two M˛ 2element 40s that have served me so well. I will let you know if the software predictions of this article prove to be true. I do expect that they will. Particularly since the models of the stack shown in this study coincide very well with the observed characteristics of my current shorty 40meter stack. I wish to thank Ken, K6HPX of CALAV Labs, and Eric, N7CL the designer of the 2d40 antenna, for providing the detailed NEC model of the CalAv Labs 2d40 40meter beam. Also thanks to Tim, K3LR for allowing me space in CQ Contest. I hope this article has proved interesting to the readers of this column. I can be reached at 972 203 2008, or email to wx0b@arraysolutions.com March 2001 