YAGI COMPUTER AIDED DESIGN By Paul McMahon VK3DIP Version 5.0 October 2003 Copyright Paul McMahon 1991,1992,2003. This file is in plain text in YC50HLP.txt and may be printed if required. WHAT IS YAGICAD? Yagicad is a fully integrated analysis and design package for yagi aerials. WHAT CAN YAGICAD DO? With Yagicad it is possible to enter a base design from scratch or use one of a number of saved well known designs. This design can then be optimised or scaled to suit particular requirements. Once this has been done a matching unit can be estimated and overall performance characteristics can be calculated and displayed graphically. Also available are radiation patterns and hardcopy print-out of results. Allowances in an analysis can also be made for element cross sections other than simple circular ones, as well as boom mounting techniques. DEFAULTS AND SUGGESTIONS. When ever YAGICAD prompts you for some value it will suggest a value that based on past experience may be suitable. These values can be over written if required. AN EXAMPLE. The operation of YAGICAD can be most easily understood by studying an example. Suppose you are interested in a Yagi for two metres. 1. Get the base design by either typing it in using the Add Element menu item under the Edit menu, or use one of the saved designs and choose a design that is close to your requirements. 2. Say you choose to use the saved NBS 6 element design. So you load (using File, Open) the NBS6.YAG file. 3. If, as in this case, the saved frequency is not what is required then you can quickly scale the design roughly to the new frequency say 147 MHZ using the SCALE function under the Toolbox menu. You could also have changed the element diameters to some new value if desired also using the SCALE function, or even allowed for non circular cross section elements by using an EFFECTIVE DIAMETER. These also are found under the Toolbox menu. CALCULATION options are "L" shaped, rectangular, or folded dipole. You could have also edited the saved data using either the EDIT menu Functions, or just double clicked on the item to be changed orto add or delete an element etc.. 4. You could then do a BASIC calculation (Calculation Menu) to determine the gain, front to back,etc., for this new design. Graphs of pattern and other features are also available under the Calculation menu. 5. The design could now be optimised for MAXIMUM FORWARD GAIN, (FRONT TO BACK, RESONANCE of the DRIVEN ELEMENT, or MINIMISE LARGEST SIDELOBE could also have be chosen). This is done by selecting the appropriate entry from the AUTO area under the CALCULATION menu and selecting the value to be varied,ie. a spacing, or a length. 6. Steps 4 and 5 are repeated as many times as is necessary. When satisfied you could do a SWEPT FREQUENCY RESPONSE to see how the design performs over a range of frequencies. 7. Assuming all is OK you could now proceed to calculate a matching network for your design. You select MATCH from the TOOLBOX and say decide to use a FOLDED DIPOLE. (GAMMA, or TEE matches, or many others could also have been used). Let us say you are going to use a 4 to 1 coaxial balun with 50 ohm coax ie the required input impedance to be matched is 200 ohms. The program gives you the required diameter of the folded part of the element. If you had chosen to use a gamma match then it may have been appropriate to RESONATE the driven element to some negative (Capacitive) reactance. A figure of say -15 to -20 Ohms has been found to give better results than a true resonance ie. 0 Ohms. 8. You could now get an OVERALL RESPONSE of the design including the effect of the matching network, but it would be best to allow for the folded dipole by first CALCULATING its EFFECTIVE DIAMETER and EDITING in the new effective diameter of the driven element. Steps 6,7, and 8 could then be repeated to ensure that nothing untoward has happened. The final resultant plot as with all other graphics can be either printed out (File, Print) or copied to the clipboard for pasting into some other program. The option is also given to save the numerical results of the plot to a .CSV file (File, Export) if you wish to say use a spreadsheets graphing capability instead. 9. You may now wish to allow for a metallic boom. This is done using the BOOM item in the TOOLBOX. The values suggested by YAGICAD allow for the element being mounted though the centre of a circular cross section metal boom with good electrical contact. (Classic option), or insulated through boom using the GE3SEK formula. You can allow for non circular booms with the EFFECTIVE DIAM's tool. 10. It would be a good idea to investigate the effects of varying the parameters and type of the mathematical model used to calculate the Yagi's performance. A large number of variations are available via the Options menu. In general the Classic model is probably the least accurate (but fastest) and increasing the number of segments used on the NEC2 model should produce a better result. Please however note the caveats and warnings below. 11. When satisfied the final results can be printed out using the File, Print function. 12.Finally the new design can be saved to disk for possible future use using the File, Save option. In fact if you come up with some particularly good design why not send the data file to a friend using packet, or put it on the web. You could now build the yagi feeling reasonably sure that it will behave at least close to the calculations. HOW YAGICAD WORKS. CLASSIC YAGICAD. (Versions 4.2 and less) The early DOS versions of Yagicad used a simple form of what is known as the 'Method of Moments'. This method is similar to that used in programs like MININEC, however Classic YAGICAD used a small number of piecewise sinusoids to represent the currents flowing in the elements rather than the pulse functions found in MININEC. This does not necessarily make YAGICAD any better or worse than MININEC based programs, however it is different, and because it is different, and uses different assumptions the results obtained will also be slightly different. In fact there are a large number of different ways to model antennas, and an even larger number of possible assumptions and simplifications that can be made. Another variant was that used by Lawson in his works for Ham Radio Magazine. This is known as the induced EMF version. All of these versions give slightly different results, and all are good in some situations and bad in others. Generally there is a basic agreement however on how a particular antenna will behave. In other words don't expect the numerical answers given by Classic YAGICAD to be exactly the same as given by some other program, however general trends should be common. In fact YAGICAD has been specifically made to be conservative with its answers and in this mode for example will usually be somewhere up to about one dB lower than the gain claimed by the originators of some designs. Note Classic YAGICAD gives gain in dBd ie. Gain relative to a dipole, many antennas specs give gain in dBi ie gain relative to isotropic. If you wish to convert between the two then dBd = dBi + 2.15 YAGICAD FOR WINDOWS (Versions 5.0 +) With this version being produced over ten years after the last DOS one the power available on the desktop to run such programs has increased drastically. As well as the classic mathematical model this extra grunt has allowed a version of the very powerful NEC2 engine to be realistically incorporated in Yagicad. NEC2 is probably the best-respected widely available antenna model available Today. The NEC2 version used by YAGICAD is a slightly hacked version of the Standard Fortran code (to make it interface better with YAGICAD) compiled with the Watcom Open Source Fortran Compiler. The remainder of the system has been produced in MS Visual Basic Version 5 WHAT YAGICAD WILL NOT DO. THIS PROGRAM IS BASED ON A THEORETICAL MODEL OF A YAGI ANTENNA, MANY ASSUMPTIONS AND SIMPLIFICATIONS HAVE BEEN MADE. THE RESULTS OBTAINED WITH THIS PROGRAM MUST THEREFORE BE TAKEN WITH SOME CAUTION. IN OTHER WORDS IF YOU BUILD AN ANTENNA THAT HAS BEEN DESIGNED WITH YAGICAD DON'T EXPECT TO GET IT WORKING EXACTLY AS CALCULATED WITHOUT AT LEAST SOME EXPERIMENTAL ITERATIONS. IE. YAGICAD IS NO SUBSTITUTE FOR A VSWR METER ETC., AND AT LEAST SOME TRIAL AND ERROR. ANYONE WHO DOESN'T AT LEAST USE A VARIABLE CAPACITOR OR HAVE SOME MEANS OF VARYING A GAMMA ARM LENGTH IS EXPECTING TOO MUCH FROM THIS OR ANY OTHER ANTENNA PROGRAM. Some common sense should also be exercised when optimising designs once you are in the area where YAGICAD is varying lengths or spacings by less than a millimetre. Sure extra gain is being obtained but you will have to fine tune your final antenna with a nailfile, so be practical and you will get results that are close to calculations. YAGICAD is limited to Yagis with less than 24 elements. EXPERIMENTAL VALIDATION Not withstanding the very important caution above, YAGICAD has been around in various versions for some 14 or so years. In this time I (VK3DIP) alone have built dozens of different yagis as designed by the various versions of this program. Others have built many more. In all cases known to me the antennas did work more or less as designed. Adding NEC2 to the program can only make this agreement closer and give more indication of what is likely to happen. My original motivation for the program was to be able to design antennas that would be good for Two Metre Fox Hunting, as opposed to the maximal gain type usually detailed in books etc. So I have also had quite a bit of testing of patterns and there is no doubt that the Yagis designed with this program have patterns much like those calculated. Similarly Yagis designed with this program have done well when gain was required. For example the VK3DIP 6 element design given in the standard designs has won at least two antenna gain measuring contests, and has been used extensively on field days etc. It is a classic example of what can be done with this program. The design criteria were that it would be for Two Metres, (down in the SSB ) part of the band, would have a boom of as close as possible to two metres in length. (this was the maximum length boom allowed in one of the gain contests) and it would have a largish gap at about the centre of gravity to facilitate mounting. I was able to achieve this design with a few hours work using YAGICAD rather than many days of trial and error. TESTING THE MODEL WITHOUT BUILDING IT One way of getting more confidence in the results given by YAGICAD is to check for variation of those results as you vary the Model Options, or even just the element lengths. If for example there is a wide difference between the result given by the Classic Model verses NEC2, then while NEC2 is probably much closer to the truth (assuming you have sufficient segments selected) the chances are that the design is going to be more difficult to realise in practice, and is probably not for the novice builder. In a similar way as mentioned before, designs where fractions of millimetres in lengths or spacings make dB's of difference, will not perform as predicted in a real world with things like weather, birds, and common materials and construction techniques. If however your design does not have a great sensitivity to small changes then you can probably build it with more confidence that it will come out as expected. DISCLAIMER In no event or way will the Author be responsible or liable for any damages, including any lost profits, lost savings or other incidental or consequential damages arising out of the use of or inability to use these programs, even if the Author has been advised of the possibility of such damages, or for any claim by any other party. YAGICAD 5.0 IS FREE Unlike earlier versions of YAGICAD which were released under a shareware banner(not that I ever made any money out of them) this latter version 5.0 is presented free. I offer it to the worldwide community of antenna experimenters in part repayment for the advice and invaluable input I have received from them over the years. It is my hope that it will be useful to them and also help to introduce new people (and especially radio amateurs) to the joys of antenna experimentation. I do however retain the copyright, and if someone wanted to include this program in some book or software collection I would appreciate being at least asked first. I can be contacted at the addresses below. Similarly if you have any comments or feedback please feel free to email me. I can't guarantee a speedy reply but I probably will get there eventually PAUL McMAHON VK3DIP 47 PARK AVENUE WATTLE GLEN VICTORIA AUSTRALIA 3096 WEB - http://www.users.bigpond.com/pmcmahon/ CREDITS This program could not have been what it is without the following: Ham Radio Magazine A sadly missed source of things technical. Over the years I have never missed a May issue. In particular the articles by Lawson in the early 1980's. Joe Reisert, and Bill Orr for the many snippets here and there. YAGICAD Users. The feedback from users is always appreciated. 73 and good designing PAUL McMAHON VK3DIP.