Mike Gibbings, G3FDW
Issue 58, August 1998
In the long history of beam aerials there has persistently existed the claim that certain
types of aerials perform better than yagis. The classic case of the quad verses the yagi
is typical of this controversy.
In the latter case red herrings such as polarisation have been offered to explain the discrepancy in observed gain. There are similar cases such as log periodics and even the much neglected slot aerial, all of which have been observed to produce better results in respect to gain than the yagis with which they were compared. Could it be that all the observations where wrong, or that all the experimenters where lying in their teeth?
Strangely, we in the Old World have yet to become embroiled in this controversy, but on the left hand side of the pond they have come to blows over the subject and the arguments have raged for the last thirty years on the pros and cons.
Being a simple soul, having used LPs and LPYs with great success for some five years and having watched in amazement my friend G3JYP gain a Five Band DXCC using a slot aerial while beating opposition from six element tribanders around Europe into the dust, I had a sneaking suspicion that either I could not believe the evidence of my own ‘ S’ meter or there must be a simple explanation.
Then an interesting tale came to light of a local amateur who was the proud owner of a three element tribander of a well known make. One day in the depth of winter the beam was seen to be covered with a thick coating of ice, but as the amateur had a sked due he went ahead and used the beam. When he checked the beam half an hour later he was surprised to find that the ice had all melted from the six traps where as the sub zero weather had held the rest of the beam in its icy grip! As he was only running 100 watts he was to say the least surprised, added to which he had noticed no change in the SWR. In the light of this observation was it possible that yagis with traps could lose a considerable amount of the input power in losses in the traps?
At this point a far more important question dawned on me. Just what was the efficiency of any particular beam aerial, or put another way, just how much of the power that was fed into an aerial was actually radiated?
We all knew that a dipole was about 80 to 90% efficient but was a yagi more or less efficient? We could all quote the gain of our own yagis because any computer would work out exactly what it was, but was it correct to assume an efficiency of 100%? That is, was all the gain really realisable?
If our dipole is 90% efficient, and the radiation resistance is 70 ohms, then we know that the loss resistance is 10% of 70 ohms, or 7 ohms. These figures would be approximately correct for a continuous aluminium tube dipole at least half a wavelength above ground.
At this stage it was decided to consult one or two of the standard reference books on the subject which could throw light on the possible losses in traps and yagis in general, as certain aerials with high radiation resistance feed systems, eg slots, LP cells, loops and HB9CVs etc, had very low resistive losses inherent in the design.
Let us consult ‘HF Antennas for all Locations’ by Les Moxon, G6XN (page 109). In this we find confirmation of the ability of traps to melt ice. G6XN calculates that the loss for an input of 100 watts in the driven element traps is 25 watts and that in the parasitic elements to be 18 Watts, a total loss of 3.06 dB for a three ele tribander! So in spite of the manufacturers inflated gain claims the beam may only produce some 3 dB of gain in the real world. But at a cost of £350 plus, it does have a fair back-to-front ratio over a narrow bandwidth.
No wonder the DX fraternity opt for 3 ele monobanders. Are they right? Bear in mind that any aerial with parasitic elements have the element losses in common with yagis. For all practical purposes these losses are common to both aerials and can be ignored for differences in gain calculations.
My baseline for comparison is a five element LP which covers 18, 21 and 24.9 Mhz. This beam, designed with ‘log periodicity’ on the three bands, has a gain of 4.5 to 5 dBd on the three bands, a relative spacing factor of 0.05 and a taper factor of 0.843. The boom length is 10.1 ft and the longest element is 30.3 ft. Feed impedance is 50-70 ohms on all bands. All elements are feed with the usual open wire feeder which crosses over to feed each element 180 degrees out of phase. This feed system has very low loss as radiating elements are in parallel. The losses would at least be comparable to a dipole or even better, ( 80 to 95 % efficiency.).
Taking information from ‘Yagi Antenna Design’ by W2PV, we find that for a computer optimised 21 MHz beam we get the following figures. (pp 8-24): boom length = 13.5 ft, feed impedance = 16 ohms and gain = 6.06 dBd.
This is not a very promising design if we want a short boom and a reasonable feed system. How do we feed it and not incur the extra losses of say a gamma match? If the loss resistance is only 7 ohms ( as calculated above) and that in the matching system only 1 ohm , then the losses are such that only half the input power is radiated. That is, the gain loss is 3 dB. The beam gain is now only 3.06 dBd on both transmit and receive. But G6XN estimates that for a similar super gain yagi design the feed impedance is of the order of only 9 ohms, not 16 ohms as the computer claims. So what is the gain of the beam now in the real world, only 0.5 dBd?
Could this ‘yagi loss’ explain the apparent increase in gain claimed for a simple LP when a single director is added, of 4 dB and the predicted 2.5 dB if a director is added to a 2 element yagi?
The losses in a multi element yagi? Your guess is as good as mine! Have we been measuring beam efficiency rather than gain when we have tried to compare beams in the hard practical world in which we amateurs try and operate? Could this built in loss in a short boom yagi be the reason why the ‘quad’ fraternity also claim greater gain in practice when quads are pitted against a yagi? As I read it, W2PV’s book, or rather computer, proves that a quad has hardly any gain compared to a yagi with the same number of elements and with the same boom length. Ask any quad owner if this is true!
Is it all down to dear old ohms law and the beams efficiency?
It can be concluded that all close spaced yagis for both HF and VHF have a built-in loss factor due to the very low feed impedance of the driven element, whereas aerials with inherent high impedance feed systems such as LPs , quads, slots and HB9CVs etc, when used to feed parasitic elements, are much less lossy, and therefore more efficient as radiators of RF power and give more observable gain than a yagi with the same length boom.
No type of matching gets over this loss created by the basic low feed impedance of a short yagi, as the losses remain the same in the feed dipole.
The moral of this tale is that a prudent amateur does not spend hard cash to put a dummy load on the top of his mast. In practice a two element beam may give more gain, certainly more dBs per dollar, than the three element beam at twice the price. What ever the computer or the adverts say, remember,
DON’T YOU BELIEVE IT!
