Low Q Yagis - G4CQM - 2017

Designing a Dual Band Yagi (6/4M) by G4CQM...

CQM6N455 Dual Band Yagi (6/4M) by G4CQM

Ian White GM3SEK in his online VHF/UHF Long Yagi Workshop says: A good Yagi design is as easy to build as a poor one... SO, ALWAYS BUILD A GOOD DESIGN!

The truth is that all yagi designs are a compromise, there is no perfect design, some designs however are better than others!

My Dual Band yagi design detailed here offers a very short boom length, a really compact design. It has five elements on both bands (6/4M) and uses a single feed point making construction very easy. The current MK2 design was formed by interlacing two very low average Q-factor (when in singular deployment) yagi designs, they are 6CQM5UC (Low Band) and WS45065 (High Band). A MK1 version that was tested used a different yagi design for the High Band.

CQM6N455 MK1 abandoned due
to too high average Q-factor
CQM6N455 MK1 abandoned due to too high average Q-factor

There are many lessons to be learned in creating a Dual Band and Single Feed interlaced yagi, it highlights many of the risks and misconceptions involved in designing yagis!

Analysis of the G0KSC KSC649D and YU7EF 5+5 DUO yagi designs gives an insight into the problems and limitations of interlaced yagis, particularly when it comes to making the High Band work effectively.

CQM6N455 MK2 is designed to give virtually no loss of performance on the Low Band. However, all Dual Band yagis will suffer restrictions on the High Band in some form or another!


Derek Hilleard G4CQM


I am extremely fortunate to own K6STI's Antenna Optimizer Professional (AOP) version software and use it as my primary design tool. It has several clear advantages over NEC-2 or expensive NEC-4 based software. NEC has failings when dealing with wires joined at right angles, wires of different diameters when joined along the same axis, and wires that are close together.

AOP as it is known is still probably one of the most accurate and powerful software programs for antenna design with aerials using wires. For example it has bent-wire correction (can be switched on or off as required) built in which is a serious issue with NEC-2/4 based software when looking at complex driven elements like loops and folded dipoles.

AOP is an enhanced version of the original AO and benefits from very special coding that allows it to run lightning fast. It also has a significantly increased number of pulses available for analysis of very large arrays.

AO is an enhanced version of the MININEC antenna-analysis program combined with an automatic optimizer. AO is more powerful than MININEC, more accurate, easier to use, faster, and has many additional features. AO can analyze and optimize almost any antenna made of wire or tubing.

AO uses the Cartesian coordinate system to refer to points in space. X and Y are in the horizontal plane and Z is height.

I have also used the free and powerful 4nec2 by Arie Voors to provide additional assessment of this and my other designs!

Symbolic Dimensions/Expressions...

AO and AOP use the .ANT (Antenna File), either in numerical file format or as symbolic dimensions/expressions. The latter in my view being much better at describing diameter, position and length for each element. This is because it simplifies and reduces the posibility of human error when entering numerical values and then optimizing/tuning a yagi design. See the CQM6N455 antenna file below...

CQM6N455 MK2 (6CQM5UC & WS45065 yagis) 6/4M Single Feed
Free Space
50.150 MHz
10 6063-T832 wires, millimeters
padia = 15.8750
dedia = 15.8750
rLp = 0
rHp = 350
deLp = 692.0000
deHp = 812.5000
d1Lp = 1009.0000
d1Hp = 1267.0000
d2Lp = 1699.0000
d2Hp = 2029.0000
d3Lp = 2806.0000
d3Hp = 3134.0000
rL = 1497.0000
rH = 1059.5000
deL = 1436.0000
deH = 1015.5000
d1L = 1314.5000
d1H = 955.0000
d2L = 1257.5000
d2H = 931.5000
d3L = 1216.0000
d3H = 871.0000
1 rLp -rL 0 rLp rL 0 padia
1 rHp -rH 0 rHp rH 0 padia
1 deLp -deL 0 deLp deL 0 dedia
1 deHp -deH 0 deHp deH 0 dedia
1 d1Lp -d1L 0 d1Lp d1L 0 padia
1 d1Hp -d1H 0 d1Hp d1H 0 padia
1 d2Lp -d2L 0 d2Lp d2L 0 padia
1 d2Hp -d2H 0 d2Hp d2H 0 padia
1 d3Lp -d3L 0 d3Lp d3L 0 padia
1 d3Hp -d3H 0 d3Hp d3H 0 padia
1 source
Wire 3, center

L = Low Band, H = High Band, and p = Position.

Please note that 4nec2 by Arie Voors can also work with Symbolic Dimensions/Expressions!

Average Q-factor...

As my research continues I'm more convinced than ever that Q-factor is probably the most important consideration with any yagi design. It defines what yagi type you are using and how it is likely to perform!

The Q of a resonant circuit is a valuation of its efficiency, or sharpness of resonance!

Calculation: Q-factor is calculated from R - in (real) and X - in (imag) against frequency, it gives a real handle on where 'Q' sits relevant to VSWR and they don't always follow one another particularly so at the HF end where trouble awaits!

WAXX10 R-in (real)WAXX10 X-in (imag)

If X>0, the reactance is said to be inductive.

If X<0, the reactance is said to be capacitive.

High Q designs require careful and accurate construction otherwise boom effect and other such variables become overwhelming, giving rise to poor performance because they are off tune. This might also be wrongly perceived as a quiet antenna! Where there is a steep exponential curve at the HF end then high Q is very likely to result in instability in bad weather.

My research has revealed that when it comes to Q-factor there are three types of yagi, see table examples 1-3 below. Don't confuse true resonance with matching, the point of lowest Q and minimun VSWR are not always on the same frequency! A good match on a spot frequency can be effected no matter where Q is, but the available (useable) VSWR bandwidth around that spot frequency reduces as average Q-factor rises.

It has become clear to me that those yagis with resonance LF of the band are most at risk of compromise due to dielectric loading effects from wet weather etc. This risk becomes real when there is a steep exponential curve (Q-factor) at the HF end, regrettably this concerns a large number of yagi designs out there!

However, yagi designs that have a low average Q-factor and shallow gradient like those by Gunter Hoch DL6WU offer a high degree of stability. Better still those yagis with resonance in-band or HF of the band and very low average Q-factor will be the most resilient!

Highly stable DL6WU 10 ele yagi
Highly stable DL6WU 10 ele yagi

At 70MHz the G0KSC KSC649D and YU7EF 5+5 DUO fall into the first category where resonance is actually LF of the band. Q rises dramatically in both of these examples on the High Band at the HF end and this will likely give rise to instability in wet weather. Average Q-factor on 70MHz is also relatively high in both, so boom correction will be required when using Stauff type clamps on the High Band yagi. In this case could be as much +15mm to each overall element length!

Resonance LF of the band
(YU7EF 5+5 DUO)
Resonance LF of the band (YU7EF 5+5 DUO)

CQM6N455 MK2 actually has resonance in-band at around 70.250MHz, and at the HF end Q has not risen by much at all! However, because average Q-factor is relatively high then boom correction will be required on the High Band when using Stauff clamps.

When it comes to Q-factor there are three types of yagi...

1. Resonance LF of the band
2. Resonance in-band
(CQM6N455 MK2)
3. Resonance HF of the band
1. Resonance LF of the band (KSC649D)2. Resonance in-band (CQM6N455 MK2)3. Resonance HF of the band (6CQM5UC)

Note that on the Low Band CQM6N455 MK2 average Q-factor is very low, and so is highly stable!

Driven Element and the feed point...

The idea behind this type of interlaced yagi is to use the Low Band driven element to also excite the High Band driven element. Mutual coupling occurs due their close proximity and immersion in the surrounding intense electromagnetic field. The original lengths and positions of the two driven elements will change as a result, not the same as in singular deployment. Finding these new positions and lengths is in fact the tuning process. Numerical file format is frankly a nightmare when trying to work through this process. Symbolic Dimensions on the other hand makes easy work of getting the best match on both bands as speculative changes are implemented!

CQM6N455 Low Band VSWRCQM6N455 High Band VSWR
CQM6N455 Low Band VSWRCQM6N455 High Band VSWR

A downside of feeding the Low Band driven element and relying on mutual coupling to excite the High Band is that average Q-factor increases noticeably on the High Band. It is a significant challenge to bring this down or indeed create true in-band resonance in a bid to improve stability.

Pattern Assessment...

CQM6N455 MK2 50MHz @ 10M AGLCQM6N455 MK2 70MHz @ 10M AGL
CQM6N455 MK2 50MHz @ 10M AGLCQM6N455 MK2 70MHz @ 10M AGL

One serious problem affecting the High Band that I have discovered when interlacing yagis is pattern distortion in the H-Plane, resulting in an apparent reduction in forward gain! It is possible to overcome this by developing unique/individual yagi integration and not combining off the shelf singular solutions. I have struck a compromise with CQM6N455 on the High Band by careful positioning.