BACK TO ANTENNA OVERVIEW

Construction of a dual band yagi for 4 and 6 meters

I need to replace my 6m antenna because of tear and wear. At the same time it would be nice to get a 4m antenna up the big mast. A dual band yagi is an obvious solution.

The first design to examine was the IW0FFK design. It did not exactly fit the bill because it is a bit too long for me, and the dual feed points are unpractical.

Next option were the YU7EF designs. They suffer from the fact, that priority was given to 6m performance. Ideally I would prefer more emphasis on 4m, as this is more of a Tropo band, whereby the gain is needed. I am also less concerned about side lobes than Pop. On the other hand the YU7EF designs are great because they have a common 50Ω feedpoint, something that is very convenient in a dualband design, where a (frequency dependent) transmission line impedance match is not an option. Instead a couple of meters of coiled 50Ω cable can serve as a balun.

So although it is new to me, I had a go of doing a design myself. I am using MMANA-GAL (MININEC engine) to do the optimisation, and verifiying in 4NEC2 (NEC2 engine). I use NEC-2 for MMANA to convert  .maa files to .nec files (cut and paste from the NEC-2 input window. Delete the second line that begins with "RP"). The MMANA-GAL optimisation is very easy to set up, and allows for simultanous optimization at multiple frequencies.

Note on NEC2: Aim for a value of around 30 - 40 segments per half wave in this frequency range. Fewer segments cost precision, and too many invalidates the NEC-2 model, as the segments stop being wires (length/radius ratio should be at least 8). If for some reason you need more segments per half wave you need to invoke the EK card (Extended thin-wire Kernel). Use the Average Gain test and the Convergence test to verify your model.

Dual band antennas will often give segmentation check errors because two elements of different length are close to each other. This will lead to non-aligned segments in the two elements. The trick to avoid this is to chop up the longest element into 3 wires (GW). A central wire being the same length as the shorter element, and two wires at each end to bring the element back to the original length.


 

The starting point was the GW3YDX 4/5ele yagi because it has a good balance between 6m and 4m gain. I started by adding one extra element per band, dimensions extrapolated from the existing directors. After optimizing on this design I transformed it to a single feedpoint (parasitic drive element on 4m, also called open sleeve dipole) à la YU7EF, because I am not too fond of the parallel feeding solution, that is more suited for shortwave yagis.

Used optimization parameters: Weight: Gain, SWR, F/B (bit less emphasis than the other two). Goals: F/B = 20dB SWR = 1:1.2 Frequencies: 50.2 & 70.3 MHz (because the frequency of lowest SWR appears to drop a bit in the NEC2 simulation)

YU7EF, Pop helped me refine the design.

I am quite happy with the simulation results of the design. Biggest concern is the sharp resonance on 4m, and that will require attention during the construction. I'm curious how the design will reproduce in real life.The yagi is 5.4m long and uses elements with a diameter of 12mm.

Dimensions can be seen in the  simulation file

Boom (m)	Half element length (m)
0		1.4615
0.875		1.0265
1.336		1.4185	(feed)
1.376		1.011
1.749		0.990
2.311		1.294
2.758		0.970
3.505		1.338
4.425		0.9795
4.71		1.303
5.4		0.952

50 MHz

70 MHz

I will go for a light construction using NUXCOM components and a 20x20mm boom. The distance between the two feeding elements have been fixed during the optimization to 40mm to fit the distance between the mounting holes in the standard electrical equipment boxes normally used for dipoles.

Given the fact, that the antenna is very narrowband on 70 MHz, you are encouraged to leave the 4m dipole and the first 4m director at least 6mm longer (of the total element length) than the values given in the simulation file. This is equivalent to a downwards frequency shift of about 200 kHz..

Using the dimensions of the simulation and using plastic end caps the antenna showed minimum SWR just below 70.150 when in the test position.

How to tune the antenna

The sharp reasonance at 4m requires the antenna to be tuned in-situ, which can be a little cumbersome, but that is the price to pay for the high gain on both bands. You should follow these steps:

  1. First make sure the antenna is working as expected with an SWR dip on both bands. The 6m dip should be better than SWR 1:1.2. The 4m dip should be 70.200 or lower.
     
  2. If needed cut the 6m dipole for best SWR at the proper frequency (the 6m response is more or less independent of the 4m elements). Around 3 mm/100 kHz of each end of the dipole. A pipe cutter is perfect for this job, but finish off with a file to make the cut straight.
     
  3. Now cut the 4m dipole and the first 4m director together for reasonance at the proper frequency. Around 1.5 mm/100 kHz of of each end of the elements.
     
  4. If needed move the boom position of the first 4m director to make the 4m dip deeper (use two crossed tie raps to hold the element).
     
  5. Steps 3. & 4 should be done in an iterative manner.
     
  6. Be aware of things influencing the antenna. My antenna changes response as I turn it around. If you cut the 4m SWR dip a bit too far it might be kind of an advantage because rain will move it downwards. Remember that plastic end caps will move the dip a bit down in frequency as well.

The antenna in the mast, and the SWR curves at the antenna (compensated for 20m of RG213 = 1 dB of cable loss). The 4m response is sensitive to objects near the antenna, rain or ice.

50 MHz

70 MHz

BACK TO ANTENNA OVERVIEW