DK7ZB attempts to undermine G0KSC continued
11 Sep 2012
DK7ZB attempts to undermine G0KSC - continued 10/9/2012
For the original post you will need to look at the top tool bar link. Our friend Martin, DK7ZB has really had his noise put out of joint now but why has he reacted this way and to what? He has set about trying to discredit me and the LFA Yagi but for what reason? Is it the fact that he feels I have undermined his work with my designs being so much better than his? There was no intent in this regard, it just happened that way.
The fact is, with YO or (Yagi Optimiser, used by DK7ZB) antennas can be produced within a few minutes. Within this software you can select the number of elements you wish to use, an approximate length of boom and press go (you need very little knowledge about Yagis and adjustment parameters at all to use it). A few minutes later, you have a Yagi. Not a good one, but you have a Yagi. As with many amateur designers, Martin makes the 'gain, gain, gain' mistake and seems to have simply optimised for maximum gain without regard for pattern, side lobes (in particular the elevation plane) or bandwidth. Any of his antennas that have some kind of bandwidth appear to have been 'lucky hits' and those with seemingly good F/B have HUGE, uncontrolled elevation lobes as a byproduct.
Some tips for Martin:
Optimise in the elevation plan only. It is clear to see you optimise in the azimuth plane as almost always your elevation plots are 'blown'. Optimising in the elevation plane will result in the azimuth plane looking after itself, this dose not work in reverse. Optimising in the elevation plane will also result is flatter performance curves in all areas and will help control bandwidth.
Use 1 degree steps, using 5 degree spacing allows for odd lobes to appear between the steps, especially at UHF.
Fix your first and last element positions in order the optimiser has to search for the best results between those points.
If with fixed first and last elements the F/B is not good enough, shorten the boom slightly and optimise again. It takes longer than 5 minutes in YO but you get far superior results.
These points are free, mail me and I may give you some more hints to make your antennas better :-)
Joking aside, YU7EF antennas have always been far superior to the DK7ZB YO designs. However, while Pop has a good command of the English language, he has perhaps not been best placed to communicate what makes a good Yagi good. I know from speaking with him that he has chosen not to write on the subject too much. The importance of flat-line impedance curves, flat bandwidth and flat performance (performance meaning gain and F/B) when hams have searched for a Yagi design, are overlooked. They look at number of elements versus gain and decide upon an antenna that way, that's it. Despite the fact they cannot use it in rain and snow and get huge static noise is realised after they have built the Yagi they selected. The fact is, per foot or per metre of boom, the gain figures on the YU7EF antenna are very similar and yet they offer so many more benefits including the ability to hear much better. In general, YO optimised Yagis have far too much space between the elements to maintain stability. The irony here is people find the DK7ZB antenna has less elements for a given boom length and select this one as being better than an antenna with a higher concentration of elements.
The difference between Pop and me being that I have been able communicate what makes my antennas different more easily and applied lateral thinking to my designs. Martin points out in his latest rant that the LFA is nothing more than a 12.5 Ohm antenna with a loop on it. Congratulations, well done. This is exactly the communication I have made in my articles, the fact that if an antenna is designed right, it need not be anything other than 50 Ohm and lower impedance antennas are best when the driven element is designed to present a 50 Ohm impedance. The OWL is a 12.5 Ohm antenna with a folded dipole, this makes for a 50 Ohm direct feed Yagi. The LFA can be anything between 6 and 20 Ohm (if a split dipole is used instead of the loop) and the LFA loop is adjusted (in both X and Y planes) to create the impedance transformation to 50Ohm. The OP-DES Yagi uses bent back tips at the ends of the driven element to raise the impedance for 50Ohm (first used by UA9TC on his reflectors), this indeed was the target.
To Martin's untrained eye, he sees 'nothing more than a 12.5 Ohm Yagi'. The fact is, this 12.5 Ohm Yagi as he calls it, has a driven element which is high impedance and requires no matching arrangement at all. unlike his rain-shifting, narrow band 12.5 Ohm Yagis, the LFA is stable and provides performance over wide bandwidths. It also has a closed loop providing all the benefits that brings in terms of noise. The fact that his 12.5 Ohm Yagis are extremely narrow band demonstrates again the lack of understanding of his tools and the products which it produces. He has allowed his optimiser to take the easy route to gain which is extending boom length to a point where the elements are too far apart to provide any form of stability.
In one of my earlier DUBUS articles, I discussed the fact that if designed right, a low impedance Yagi can provide much wider band performance than an OWA Yagi. I discussed an experiment where I optimised an OWA (50 Ohm dipole fed Yagi) for maximum bandwidth with at least 20dB F/B. When I re-optimised with a 40 Ohm feed point impedance, I could achieve wider bandwidth with more gain and better F/B, proving the lower the impedance, the better it could be in addition to having the potential to provide performance over much wider bandwidth than a comparable OWA. However, until now no one had taken advantage of the far superior performance that could be achieved from lower impedance antennas. Other than me that is.
The issue for me was the matching arrangement; placement of a matching device in the antenna can have detrimental effects on pattern and performance and therefore, the objective of producing a driven element that could provide a Yagi that would otherwise be low impedance if a split dipole feed were used.
In Martin's original post he made these comments for the reason an 'LFA' was being used in the WWII 'Due the problems with the allied blockades of the system they needed a broadband yagi with high directivity for frequency changing'
This demonstrates a lack of understanding of indepth Yagi design (along with confirming the reliance on the simplicity provided by YO) as the loop does not make the LFA broadband. Careful optimisation and an understanding of the subject in hand along with the correct placement of elements is what produces a broad band Yagi antenna. Allowing an optimiser to run away with boom length, ever increasing the spacing between elements is why in general (although there are a few exceptions) the DK7ZB antennas are the way they are (12.5 Ohm in particular).
From his comments, it is not clear if Martin really has an understanding of the difference between a 12.5 Ohm antenna and a 50 Ohm antenna so for you (and him) I will explain through pictures and text.
If you have a low impedance dipole fed Yagi, and install some form of matching device, the driven element (at the other side of the matching device) is still low impedance and hence, high current. There are issues with low impedance driven elements such as higher noise pick up and high current across joins and connections which can lead to problems in QRO enviroments.
OK let us look at a 12.5 Ohm OWL antenna below
The above is a 12.5 Ohm dipole fed Yagi of my own design. The green lines are the elements, the pink curved lines represent current distribution throughout the elements. We can tell this is a low impedance antenna as the current distribution is high within the driven element. Now let us take a look at the same antenna fitted with a folded dipole.
All that has changed in this example is a folded dipole has replaced the split dipole. Notice the current curve? It is much lower (in the folded dipole) as the radiating element is now HIGH IMPEDANCE, This is different to a low impedance radiator fitted with a matching device as in this scenario, the low impedance radiator the other side of the matching device, is still low impedance. The Above Yagi IS now a 50 Ohm Yagi and this carries with it all associated benefits. It is simple isn't it? And yet it seems so hard to grasp for Martin. The fact that the driven element is also its own matching device means the entire antenna (as it will be built) can be modelled without having a nasty low impedance split dipole the other side of any matching device. The LFA is the same and so too is the OP-DES. Ultimately, if an antenna is modelled as it will be built with nothing added thereafter, a closer replication of model can be achieved in the real world.
As an example, below is a correctly modelled 12.5 Ohm Yagi using a split dipole. This is a 14el 144MHz of my own design:
Note how very close the first few elements are together? This is something you would simply not achieve with an optimiser that was allowed to extend its boom size. The lower the impedance a Yagi is, the smaller this 'driver cell' (first 3 elements) has to be in order to achieve stability. Now let us look at SWR:
Note the 'Ski Slope' Impedance curve and V shaped SWR that a 12.5 Ohm Yagi has become known is not there. This antenna provides a broad, flat impedance curve across a wide range and as a result, maintains sky temperature and G/T across a wide range too. This is not just SWR that is flat across a wide bandwidth, the performance is so too. When a Yagi has far too much space between elements, performance parameters ski slope too, F/B will be best at the bottom of the band, gain will be best at the top end of the band. If the antenna is modelled correctly and of the correct boom length, Gain and Front to Rear (F/R) will flat-line too. Now, lets take a look at how the plots of a correctly optimised 12.5 Ohm Yagi should look:
This is not an Azimuth plot, this is an elevation plot. Note the near-perfect symmetry at the back of the antenna and the huge reduction in lobes after the 45 degree line and back? This is how a Yagi should look, not with huge side lobes straight up and down which pick up large amounts of noise from your surroundings. Now the Azimuth plot:
Good levels of suppression to the sides, virtually no side lobes to speak of and excellent F/B but this is at 144.100MHz right? How does it look at 144.300MHz?
Virtually exactly the same at 144.300MHz as it was at 144.100MHz this means Sky temperature and G/T are maintained. Now 144.600MHz?
Starting to see a small drop in F/B but generally, F/R is maintained at 144.600MHz and gain, side lobes, and side suppression all maintained. So in summary, building this antenna would ensure you get model performance, even if you were a little out when constructing the antenna because all aspects of performance are delivered over a wide range.
You simply cannot design every antenna of different impedances for every band in the same way, many considerations have to be taken into account. Designing VHF and UHF Yagis in the same way as you would HF Yagis does not yield good results, Martin without realising it has proved that. The above example is how a 144MHz Yagi should look if designed correctly. Additionally, IF DESIGNED CORRECTLY, A YAGI NEVER NEED TO HAVE ANYTHING OTHER THAN A 50 OHM FEED POINT. Add a folded dipole to this antenna, a small amount of reoptimisation and you are good to go with the best antenna available at this boom length, in terms of G/T, period. And yes, it has a direct 50 Ohm feed point.
Now go have a look at Martin's site and some of the plots, which would you rather have on your tower?
As much as I know it must pain Martin so much, the LFA is not a folded dipole laid flat on the boom. Below is another plot from EZNEC showing current distribution, this time on an LFA.
This is a side view of the LFA loop. It is not a folded dipole, it is a rectangular, flat sided loop. The loop sections are numbered above. You have 2 and 5 which are the longer front and back sections of the loop and 4 and 3 which are the side sections. Rather than just show current distribution in this plot, EZNEC is enabled to show current phase angles too. Therefore, the angle that the pink lines move away from the element section is representative of the phase angle. Note that the two side sections of the loop (which are flat, not rounded like a folded dipole) have pink lines running in opposite directions from one another. This is because these two side sections are 180 degrees out of phase. This results in a cancellation effect (as you would see on balanced line feeders) and therefore, if careful optimisation takes place, this can be used to increase suppression off the sides of the antenna and increase front to back ratio too.
Martin's attempts of undermining this time extend to the use of other sites to where the authors have spent little time investigating and comparing antennas other that at the highest level. see an example here:
The author reverse engineers an 8 element element LFA to produce a spit dipole antenna of 12.5 Ohm which does not have the same suppression in the rear lobes as the original so while he does not mention it, he had proven the loop concept is better. He also fails to mention the issues and losses associated with maintaining a 12.5 Ohm dipole and any associated matching devices there may need to be added. Finally the author displays a number of TV antennas displaying FOLDED DIPOLE driven elements.
Finally, I have until now only produced and presented a few 28 Ohm Yagi antennas. One in particular, a 17 element 144MHz antenna which I fitted with a folded dipole was presented in DUBUS magazine. Although the original dipole was 28 Ohm and not 12.5 Ohm there is reason for fitting a folded dipole. The original DK7ZB coaxial dipole connection involves joining in parralell 2 x 75 Ohm pieces of coax. This is messy at best and hard to achieve without additional unwanted losses. However, with the new 112 Ohm folded dipole, a single piece of 75Ohm coax (1/4 wave length) can be used with a simple N type connection on the opposite end for connection to any 50 Ohm feed line. This antenna can be found on the VE7BQH list:
http://www.dxmaps.com/VE7BQH.html
If you are not a subscriber to DUBUS magazine, you would have missed this antenna but scroll down the list at the above link to '+G0KSC 17OWL-FD'
Now take a look at the performance. Note, it has more gain, much lower sky temperature and a much bigger G/T figure than not just Martin's antenna equivlant, but of anything else of that length on the list. This is in addition to having an SWR of better that 1.1:1 across at least 1MHz of bandwidth and therefore, very stable. Note in particular the G/T figure. This is a mean opinion score between antenna temperature and antenna gain which demonstrates the antennas ability to receive weak signals. The more positive the number, the better and my 28 Ohm antenna has a G/T of +0.33dB. Take a minute to scan up and down the list and note how far away from everything else this antenna is. I guess when you see the space I have opened up, it is kind of understandable why he has such a cobb on right? :-)
How was this achieved? Well, the 28 Ohm feed impedance is a great point of balance for gaining best performance in all areas but until now, no one with any modelling ability has taken advantage of antennas using a split dipole feed of this impedance.
Therefore, while I am not a fan of the twin line 75 Ohm coaxial matching transformer and much prefer to have a self matching driven element, I will now produce and publish a range of 28 Ohm antennas that are modelled as they should be. They will provide a very wide and stable performance in all areas including Gain, F/B and SWR. The small selection currently linked on G0KSC are early, 'concept' designs. The new line of 28 Ohm Yagis will aim to set new standards as all my antennas do.
So, Watch this space, the best in 28 Ohm Yagi performance is about to arrive!
As a foot note DK0TU the Technical University in Berlin asked me to design their 144MHz X-POL EME array along with some HF antennas which I have now completed. What this space for photos upon completion!
So What started this?
Martin DK7ZB contacted me asking for information on my designs for presentation in an article within the German magazine 'FUNKAMATEUR'. Martin first wrote: 'Hi Justin, with great interest I follow your new Yagi developments. For a short report in the German Magazine "FUNKAMATEUR" about new Yagis I need a high resolution picture of an OWL-Yagi. It is no problem to send me a photo of the commercial version of InnoVAntennas. The remark of the source would be made and I think it would be a good reputation for you, hi. The best would be the 13-El.-OWL with 8.1-m-boom. 73 de Martin, DK7ZB'
Martin suggests he will be complimentary in his writing and presents praise in his words...
Following this he writes: 'Hi Justin, mni tnx for the indeed useful informations! The 7-El-OP-DES is a very interesting model, bandwidth and pattern is astonishing indeed. I have used the picture of the 11-El.-OWL to show how close the elements are spaced in the radiation centre. The article is in the print (I have written it 6 months ago), I have only mentioned the OP-DES technology without details. The focus was on the UA9TC, the G4CQM and your OWL-Yagis.
I see about buying EZNEC Pro and the NEC4-engine, but I am no commercial developer and the price is real expensive. But it is good that younger experts are doing the job, hi. 73 de Martin, DK7ZB'
Martins later comments about NEC4 were on the basis I had to send him screen shots of the performance figures for the OP-DES and OWL Yagis due to the fact NEC2 based products (such as those he uses) are not accurate enough to present the designs correctly.
The published article presents a biased view that the DK7ZB designs are better than anything else. Martin is paid for his work in this magazine so obviously there is a commercial interest in his work being seen to be better than others as this helps justify his position as a writer, this is understable to a point. This is way he would not mention other, less biased articles such as the DUBUS article where 4 x DK7ZB 70cms antennas were compared with just one YU7EF antenna of a similar length, as the single YU7EF was able to hear weak signals as good or better than the 4 x DK7ZB when they were all pointed to the moon (download the article HERE). For most hams with any antenna design experience the result was no surprise as the YU7EF designs are far superior and the choice preferred by serious, knowledgable hams.
Remember:
If an antenna is designed correctly, it does not need to be matched to 50 Ohm, it will be 50 Ohm already.
Straight (driven) element Yagis are a thing of the past, next generation Yagi performance only comes from curved or bent element Yagis. However, you need to have the right tools (computer design) for the job! YU7EF, DG7YBN and UA9TC have all experimented with curved or bent element Yagi designs and shown some incredible, ground breaking results. While some designers seem to spend their time in monthly columns trying to convince their readership that antenna designs of 15 years ago are still current and 'best of breed' today, others are breaking boudarys and setting new standards.
The designs of YU7EF, DG7YBN and UA9TC are excellent (conventional as well as the new bent element versions) and what I use as a benchmark when designing antennas. Considering we now know the limitations of antennas with any form of matching devices, Nothing other than modern-day 50 Ohm direct feed Yagis are worth considering if and when you are looking at new antennas for the modern Ham Radio Station.
Two Faced? You decide