This September (2014) G0KSC released a new set of 144MHz OWL Yagi designs which have proven to set new standards in G/T performance. Is this as good as it gets or does G/T measurement miss parameters for use in the real world?
4 x 9el OWL-G/T installed at K7BV for EME purposes on 144MHz. The OWL-G/T uses a folded dipole to increase feed point impedance from 12.5Ω to 50Ω
So what is G/T and what does it mean? G/T (Gain over Temperature), is a mean opinion score which provides an indication as to an antennas ability to receive weak signals, it is as simple as that. Gain is one thing but it is not all. As I have mentioned in my writing in the past, many over the years have tried and failed to produce a high-gain Yagi which provides good day to day performance, over a range of frequencies that delivers flat-line performance and can hear well, all in one package.
How many times have you been calling that DX station that is Q-5 your side but he just does not hear you? The problem is (in a nut-shell) that when optimising for those last few tenths of a dB in gain, receive performance is often drastically reduced and this can lead to poor receiver performance (your antenna is part of your receiver right?). This next statement I firmly believe, - if you have any kind of TX amplifier, you will always be in a position where someone will hear you (much smaller station) while you cannot hear them. it is for this reason that VHF antennas should be optimised for lower noise in order receiver performance will be better. Basically, antennas are optimised for better receive performance to help balance the bias introduced by the introduction of your TX amplifier.
What is this noise and where does it come from? This is a good question and one where the use and results of the G/T table become a little fuzzy. The VE7BQH list (Lionel VE7BQH produces this independent list which compares performance parameters against one another) compares G/T of a given antenna array when angled at 30 degrees upward (this is despite the fact that performance comparisons could vary greatly at different elevation angles but that in itself is another discussion point!). The noise being measured within the G/T figure is that of warm Earth, the ground below and all around us. A well-designed Yagi will have a tight rear bubble in every angle other than its intended direction of performance (or main lobe) free from spikes and side-lobes and in turn, any pick-up from warm Earth will be low, the lower the noise and higher the gain, the better the G/T figure.
As discussed above it is often the case the designer has pushed for every last drop of gain without consideration for the rearward lobes and thus out right G/T performance is lower. Another by-product of such modelling is narrow bandwidth. Not just from an SWR perspective but in terms of performance too. it is common that a Yagi will have 'ski-slope' performance parameters too with gain and F/B (Front to Back Ratio) moving in opposite directions from one another and thus, any G/T figure measured is unstable and unreliable.
Elevation pattern shape is the most important parameter for suppressing modern-day man-made noise. The above 9 element 144MHz example has a lower G/T figure (and is much shorter) than the OWL G/T published on this site but will fair much better in receiving weak signals when placed in a city environment.
Finally (and this is where I am going to be a little controversial) G/T is a good means of comparison for modern Yagis but often an antenna modelled for a lower G/T figure (by means of a bias in the sky temperature figure with much higher suppression of rear lobes, lower gain) will often fair better for receiving weak signals in suburban areas than a Yagi with a better G/T figure simply because the lower G/T antenna has the ability to suppress man-generated (much higher than that generated by warm Earth) noise as a result of lower/less unwanted side-lobes in all directions (such as those facing directly below the antenna or pointing into neighbours homes) and hence, lower/weaker desired signals can be heard.
Sure, gain will be reduced but let us put this into perspective here. Let us say you are a city-dweller and have a 2m set-up and your noise floor is S-3. From experience I know that for each 10th of a dB forward gain given up, I can increase the suppression of the 'rear bubble' by up to 1.5dB. Let us say we were not able to achieve this number in this instance but managed 1dB extra suppression per 0.1dB gain sacrificed. There are 6 dB in one S point so if we give up 1dB in gain (extreme but for example purposes) this could be 10dB or more additional suppression in all other directions. 10dB is around 1.5 S-points on the meter. In this ideal scenario, our 'city noise' would drop from S-3 to S-1.5 and now the guy that was previously calling us who's signal was S-2, can now be heard rather than being covered by noise.
So why does the G/T table not cover this? Again, the noise levels being measured in the G/T table are very small while those generated by man are very high. An antenna that fairs well on the G/T table may still not be able to hear as well as other antennas, other antennas with lower gain ans higher noise suppression characteristics. For those hams in the country that live in a house that generates no noise, any standard G/T presented data would stand as a good point of reference for their antenna selection.
So why mention this? This new breed of OWL-G/T designs I have released were produced to humour this list, to provide a range of antennas that could be bought or built by any ham and that provide excellent levels of performance when compared to any other Yagi presented on this list. However, I believe it is important to realise that VHF Yagis (and in particular those produced for 144MHz) are not a 'one size fits all' product and that careful consideration should be given to selection of the same.
Bandwidth is an extremely important part of this selection process. Some ham designers that see my designs as benchmarks compare certain aspects of their designs against mine hoping that the reader will believe their designs are better and while I am sure that any one choosing such a design will be very happy with its performance, it is extremely important that you encapsulate all parameters as one package to compare in order you can be assured you are getting the absolute best you can from your antennas.
To this end, I have listed a number of reference points and parameter to look for and ask about before you make your selection. This will ensure you are comparing apples with apples!
Direct 50Ohm Yagis
I have some designs of different impedances on this site for self-build for those who want to experiment with matching methods. However, in the modern-day of computer optimisation and more recent driven element designs, there is absolutely no reason why any Yagi should be anything other than a direct feed, 50Ω design.
In years gone by when designers has to apply 'cut and try' techniques (where the designer builds an antenna and manually alters element spacing and length to achieve there goals), establishing a 50Ω feed impedance was very difficult to achieve so it became accepted that matching the antenna after the design was completed was acceptable.
There are no positives of installing matched Yagis, only negatives
10 or 20 Years ago, the 50Ω Yagi alternative was down on performance of a lower impedance Yagi which meant that any negatives of a matched Yagi were out-weighed by performance. However, the limitations of matching where still there and we will take a look at some of these here.
Matching loss - Regardless of what anyone tells you or writes on a web page the fact remains there is no form or transformer known to man that is 100% efficient. This means the introduction of a matching device will mean lost performance (Watts) from the transmitter and induced noise for the receiver to deal with.
TX Power limitations - Ask yourself this. Why is that matched antenna only rated for a few hundred Watts or perhaps even 1.5KW? Why is the power limitation more mechanical and based around the material that makes up the antenna? Because it is an extension of the above mentioned loss. Basic Physics can help us understand what is going on here and the fact that these inefficiencies in matching lead to the RF power entering the antenna changing form. Rather than the desired RF signal being fully radiated, a percentage changes to heat and so much of this energy is transformed that any more than a few hundred Watts or so and enough heat is produced the the matching arrangement to get hot and potentially fail.
A 50Ω direct fed Yagi (split dipole, folded dipole, LFA loop, Bent driven element or whatever) does not get hot, it is far more efficient. Additionally (and more importantly) the whole of the 50Ω Yagi is modelled in software first, so any associated losses or issues are seen and accounted for in any presented performance figures, nothing is being added after model to reduce that performance or introduce performance limitations. Ask your antenna designer for modelled losses including matching and matching device structure losses.
The above output from 4nec2 by Arie Voors using the latest NEC4.2 calculation engine shows the earlier displayed 9 el OWL with folded dipole fitted. A radiating efficiency of 94.81% can be seen and this is factored in to any final performance (Gain) figures and ultimately G/T). The software predicts 97.95 Watts for every 100 Watts entering the antenna will be radiated.
Structure impact - In addition to the above, another complication we need to factor in is the physical structure of any matching device, the impacts its physical presence will have on the Yagi (assuming this matching device has not been modelled as apart of the Yagi). Gama, 'T' and hairpin all have their vices which vary and all impact how the antenna performs when compared with model. The Gama match for example is not a balanced method of feeding an antenna (and is better suited to vertical antennas) and thus, the Azimuth pattern of any Yagi it is installed on will be distorted, the level of distortion will vary model to model. T-matching can have a similar impact to the Gama but this time, distortion will be in the elevation plane. Again, the level of distortion depends on the antenna design and also, the side of the T-match, how far above or below the antenna the match extends. The hairpin normally extends towards either the first director of the reflector from the driven element and while this can have little structural impact at HF frequencies, at 2m the influence of extending a mechanical structure (not modelled in software) towards parasitic elements can be quite marked.
The point being, if it is not modelled, it is not just the additional loss in transformation you will not be seeing, potential mechanical limitations and losses will not be seen either.
Good Yagis do not need matching - We are in an age where software simulation is extremely accurate and optimisers can make a job that at one time could take months, take only a few hours. The time of requiring any form of matching in a Yagi has long past. Recent innovators such as K6STI, DG7YBN, UA9TC and RA3AQ (and G0KSC !) have adopted methods of altering the driven element shape in order to achieve a 50Ω impedance and as a result, any quoted performance figures on their Yagis can be considered accurate and performance will be close to that predicted in software.
A G0KSC designed VHF OP-DES (Opposing Phase Driven Element System). The OP-DES has part of the driven element bent back towards the reflector. These sections act as an impedance controller, altering the feed impedance to whatever the designer needs it to be. native impedance (should a traditional split-dipole be used) is between 30Ω - 40Ω
So in summary, you really do not need to make any complex matching devices to get the very best in terms of antenna performance and make sure when selecting an antenna, you take a good look at all those hidden specifications too that may not be covered by simple 'spot' features. A good example of this is the quoted F/B and gain figures. You will note that on the G0KSC website along with both InnovAntennas and Force 12, a spot frequency is given for both gain and F/B is the same. I have covered the point before that on poorly designed Yagis, all parameters ski-slope in opposite directions from one another (along with the antennas impedance). In these cases, the best gain will be at the bottom of the antennas working range while the best F/B will be at the top of the antennas working range. Therefore, it a given antenna was quoted at working between 144.0MHz and 144.6MHz, the gain figure presented within it's performance stats could be from 144.0MHz while F/B from 144.6Mhz and any performance between these points somewhere in between.
The OWL G/T High Q version still provides exceptionally flat gain and F/B & F/R (Front to Rear) characteristics
Another good point to look for is the bandwidth of two antennas and to make sure these are compared. The wider the bandwidth of the antenna, the more flat-lined the performance parameters are and the more stable the antenna will be (wet weather, snow and ice for example). While the narrower antenna may look to be a better performer, if presented figures were from opposite sides of the band and you live in an environment where you do not have 365 days per year of sunshine, you just might be being a little misled.
As always I am happy to discuss any requirement with anyone and help them get the best from their system be it a self-build or commercially purchased option. Let me know your requirements and what you want to achieve and I will help you decide your best direction.
OWL G/T comparison
To give you an idea of the shape things to come, here is a indication antenna, a 13el OWL HQ (High Q) as presented in the above gain graph. The antenna is 8.814m long (4.24WL) and uses a folded dipole and hence, has a direct 50Ω feed impedance. Below is the TanT output should G/T performance for this antenna.
As can be seen above, this antenna has a G/T figure of -0.73dB G/T when standard DL6WU stacking distances for 4 antenna are applied. If we take a look at the VE7BQH list for comparison purposes, the best G/T figure for any antenna of that length is -.082dB G/T which just so happens to be one of my LFA Yagis. After that (and excluding the complex boxkite mulli-element quad-shaped antenna) anything else is in the -.90+ dB G/T range so this does mean quite a step in traditional Yagi performance has been made.
There are other examples too, the TanT output for a 12 element version of this antenna which has a 7.9m long boom (3.8WL) showing a G/T of -1.16dB G/T is shown. The closest to this figure again is one of my antennas, a 13el OWL which shows -1.30dB G/T then you are up to -1.35dB G/T (for a longer antenna) so again the difference is marked.
As a footnote, remember what I have spoken about with reference to G/T performance and the fact that a better G/T antenna, may not be the best choice for you in your location, it might well be an LFA with a slightly lower G/T figure may prove more successful at suppressing unwanted noise.
Before closing, take a look at the pattern of the 13el OWL G/T-HQ below:
The azimuth plot of the 13el OWL G/T HQ which shows a remarkable 16.86dBi from an 8.8m long boom.
There are models of these OWL's from 7 element to 16 elements long that I will make available but I am not going to make it easy for the 'antenna thieves' both at home and abroad to use my antennas as templates this time :-). So contact me if you are keen to see more info, want to build or want to buy.
until next time!