YO 7.5 YAGI OPTIMIZER Copyright 2004 by Brian Beezley, K6STI All Rights Reserved YO 7.5 Yagi Optimizer..........................1 Using YO.......................................1 F/B and F/R....................................3 Frequency Menu.................................3 Height Menu....................................4 Elements Menu..................................4 Element Tapering...............................5 Mounting Brackets..............................6 Boom Correction................................7 Conductivity Loss..............................8 Performance Tradeoffs..........................9 Good-Enough Thresholds........................10 Local Optimizer...............................10 Global Optimizer..............................11 Undoing Changes...............................12 Matching Networks.............................13 Dual Driven Elements..........................16 Frequency Scaling.............................17 Output Files..................................18 Options Menu..................................19 Element-Current Profile.......................20 Frequency Graphs..............................20 Pattern Plots.................................20 Arrays of Yagis...............................22 Gain Figure-of-Merit..........................23 Modeling Limitations..........................23 Verifying Designs with NEC....................24 Notepad.......................................25 Screen........................................25 Mouse.........................................26 YO.SET........................................26 Index.........................................32 ---- YO 7.5 YAGI OPTIMIZER ------------------------------------- YO analyzes and optimizes Yagi-Uda antenna designs on your PC. It runs much faster than NEC- or MININEC-based antenna-analysis programs. YO's mathematical model is more complex than the W2PV Yagi model, simpler than MININEC, and more accurate than either. YO models tapered elements, mounting brackets, the boom, matching networks, dual driven elements, and conductor loss. YO represents antennas in free space or over ground. It can model two identical Yagis stacked in the H-plane in free space. YO calculates mutual-impedance interactions for all of these cases. It also can ignore Yagi-to-Yagi interaction and model large rectangular arrays of Yagis. YO includes an automatic optimizer that can maximize forward gain and input resistance, and minimize backlobes, sidelobes, and SWR. The optimizer iteratively adjusts element lengths and spacings to optimize performance objectives you specify using parameter tradeoffs you decide. It can perform both local and global optimization. YO is calibrated to NEC, the reference-accuracy Numerical Electromagnetics Code from the Lawrence Livermore National Laboratory. YO and NEC results normally differ by less than 0.05 dB in forward gain, a dB or two in F/B, and a couple ohms in input impedance. You can invoke NEC from within YO to verify results. YO's analysis and graphics engines use assembly language with pipelined floating-point code optimized for Pentium processors. ---- USING YO -------------------------------------------------- YO stores information for Yagi designs in individual files. Yagi files contain antenna dimensions, analysis frequencies, and other data. You provide the name of a Yagi file when you start YO. After YO loads the file, you can examine or modify the design. You can save the current design in a file at any time. Yagi files use the extension .YAG. If you know which Yagi file you want, provide the filename on the command line, for example, YO HG204BA (no extension needed). Otherwise, type YO, and the program will list Yagi files in the current directory. To list Yagi files in a different directory, provide the directory name on the command line. Select a file by moving the lightbar with the arrow keys, PgUp, PgDn, Home, or End. Press Enter to select the highlighted file. Or, you can select a filename by typing it. As you type, the lightbar moves to the first filename that matches the characters entered. Press Enter whenever the desired file is highlighted. Finally, you can select a filename by clicking on it with the mouse. Select "Other" to enter a 1 file or directory name not listed. If you enter a directory name, YO will list its Yagi files. See the YO.SET section to organize Yagi files into directories. Press Esc to erase the Main menu. Press it again to erase the tradeoffs triangle. Press Enter to restore the Main menu. Press Esc to return to the Main menu from another. You also can return by pressing the command key a second time (first keystroke only for Save and Notes). This quick-return feature is handy when making a quick check. Instead of pressing Enter after typing input, press Esc to enter the data and exit the menu in one keystroke. Whenever you return to the menu, the cursor will be positioned at the item last entered. This makes it easy to change a value repeatedly to experiment with its effect. Press an arrow key after typing input to enter data and move the cursor up, down, right, or left in one keystroke. Press the spacebar to enter data without advancing to another item. This feature is handy when experimenting with numerical values. Press any command key to exit a menu (except Save and Notes). YO will execute the command, bypassing the Main menu. This is a quick way to go from one menu to another. If you press a command key during data entry, YO accepts the data and performs the command. When you enter data with more precision than that displayed, YO accepts the data as entered. You can enter fractional data; for example, type 1-15/32 instead of 1.46875. Use the # key to enter wire gauge; for example, #12. Press any nonfunctional key (F1, for example) to list control keys not shown in the Main menu. To reduce screen clutter, YO does not label the figures displayed within the Yagi patterns. They are as follows: Frequency Forward Gain & Mismatch Loss Front-to-Rear Ratio Input Impedance Standing-Wave Ratio Elevation Angle or Gain FOM Mismatch loss is displayed only when the option is enabled. Elevation angle is displayed for Yagis over ground and gain figure-of-merit for free-space, single-Yagi models. YO defaults to a generalized definition of front-to-back ratio described in the next section. 2 The notation 12.7-j15.4 means a resistance of 12.7 ohms in series with a reactance of -15.4 ohms. YO displays asterisks whenever a value overflows the space provided. Z stands for impedance. The lambda symbol means wavelengths. The path and name of the Yagi file appear in the lower-right screen corner. YO increments the iteration count in the lower-left corner each time it recalculates element currents; it appends a restart count during global optimization. When YO detects an error, it displays an error message and exits with ERRORLEVEL set to 1. Whenever it invokes NEC, it sets ERRORLEVEL to 2. Batch files can test ERRORLEVEL. ---- F/B AND F/R ----------------------------------------------- YO uses a generalized notion of standard front-to-back power ratio to characterize pattern quality. Conventional F/B is the ratio of forward power (at 0 degrees) to that radiated in the opposite direction (at 180 degrees). YO's generalized F/B is the ratio of forward power to that radiated within a specified region to the rear of the antenna. This is called front-to-rear ratio (F/R). Yagi designs maximizing conventional F/B often have large backlobes at angles other than 180 degrees. Much better patterns result when you optimize a Yagi for F/R. The F/R region begins at 180 degrees and extends forward to a specified angle (90 degrees by default). You can define the forward limit in 5-degree steps and as far forward as 5 degrees. This wide range lets you control sidelobes as well as backlobes. During optimization YO uses RMS weighting to combine backlobe amplitudes within the F/R region. This nonlinear weighting helps minimize peak backlobes. But reported F/R figures use the peak backlobe in the F/R region, not the RMS value. YO samples the radiation pattern in 5-degree steps for all F/R calculations. For H-plane patterns of antennas over ground, the F/R region begins at the rear horizon and ends at the specified elevation angle. This lets you minimize rear radiation over a range of elevation angles. Use the Tradeoffs menu to change the F/R region. Set it for 180 to 180 degrees (a single point) to obtain conventional F/B. The F/R value depends on the extent of the F/R region and on the analysis plane selected. ---- FREQUENCY MENU -------------------------------------------- YO models Yagi designs at a spot frequency or over a frequency band. Specify a single analysis frequency for spot- frequency designs, and low, middle, and high frequencies for designs covering a band. 3 Use the left column to add, change, or delete analysis frequencies. To switch to a spot-frequency design, enter zero for the low or high frequency. While spot-frequency designs optimize much more quickly, the resulting performance often is unacceptably narrowband. Use the right column to set the relative weight (importance) of each frequency during optimization. Nonuniform weighting can help flatten response across a passband. Press the = key to reset all weights to 1. YO-Professional provides two additional analysis frequencies. This provides tighter control of performance across a passband, particularly a wide one. YO models Yagis to 10 GHz. Special tweezers are available for those of you working with Yagis near the upper end of this range. ---- HEIGHT MENU ----------------------------------------------- Use the Height menu to specify the height above ground for a single Yagi, or the free-space, H-plane stacking distance for two. YO models Yagis over ground as horizontally polarized. Set height to zero to model in free space. Set stacking distance to zero to model a single Yagi. YO computes forward gain, F/R, and the azimuth pattern at the specified elevation angle. To maintain accuracy, YO accounts for mutual impedances between a Yagi and its ground image, and between stacked Yagis. YO models perfect ground with infinite conductivity and unity dielectric constant. Yagis over real earth may have several dB less gain overhead and as much as 1 dB less gain near the horizon than the figures YO reports. Elevation-pattern nulls may be less deep. Still, YO can provide useful relative comparisons among Yagi designs over ground. Use NEC/Yagis to obtain better absolute accuracy for Yagis over real earth. ---- ELEMENTS MENU --------------------------------------------- Use the Elements menu to change element positions and electrical lengths. YO displays half-element length (boom center to element tip), but you can enter whole-element length for convenience. Lengths displayed are of electrically equivalent untapered and boomless elements. These are not necessarily physical lengths. Use the Taper Schedule menu to inspect and change physical lengths. YO can model Yagis with up to 50 elements (70 for YO- Professional). Press Del to delete the element at the cursor. Press Ins to add one below it. Use ^Ins or ^Del to avoid bunching or gaps by automatically respacing elements when adding 4 or deleting. (^ means press Ctrl together with the indicated key.) Use the X key to fix or free element position or length for automatic optimization. The * key fixes or frees all elements. Use * then X to free just a few. YO highlights fixed element dimensions with color. The reflector position is fixed by default since one element position is arbitrary and that of the reflector least affects performance. The D key toggles between single and dual driven elements. YO does not add or delete driven-element mounting brackets when you use the D key. Use the ~ key to round element positions or lengths. Rounding is a good way to test design sensitivity to dimension tolerances and to generate final dimensions for construction. Use Undo to undo rounding experiments. (Since YO rounds tip lengths, untapered-equivalent lengths for tapered designs normally won't be multiples of the rounding value.) The = key equalizes all element spacings or lengths below the cursor. This can provide a reasonable starting point for optimization. Use Undo to recover from equalization experiments. Changing the position of only the last director has a special effect: All element positions after the last one fixed change to expand or shrink the design. The entire Yagi scales if only the reflector position is fixed. This feature is handy when experimenting with boom length. YO expects every Yagi to have a reflector element. You can model a two-element Yagi with driven element and director by placing a short dummy reflector a great distance away from the two active elements. This arrangement will satisfy YO while minimizing modeling errors. Before accessing the Elements menu, you can select an element with the left/right arrow keys. The selected element blinks in the Yagi sketch. When you bring up the Elements menu, the cursor will be positioned at the selected element. This feature makes it unnecessary to count elements. ---- ELEMENT TAPERING ------------------------------------------ Fast Yagi-modeling algorithms represent each element as a thin cylinder of constant diameter. A tapering algorithm converts elements of tapered telescoping tubing to untapered cylindrical equivalents. Tapered and untapered elements are electrically equivalent when their self-impedances are equal. Tapered elements are physically longer than their untapered equivalents. 5 YO uses the tapering algorithm developed by Dave Leeson, W6NL, described in the ARRL book "Physical Design of Yagi Antennas." This algorithm is based on fundamental principles and is more accurate than the earlier tapering method developed by James Lawson, W2PV. Use the Taper Schedule menu to change taper lengths and diameters. YO can model up to 20 taper sections per half- element. Use the Ins and Del keys to add and delete taper sections. Use the = key to equalize taper lengths below the cursor. This feature is handy when making wholesale taper changes. Enter a taper length of zero for elements that don't employ a particular tubing diameter. To keep the initial electrical design after making changes, exit the Taper Schedule menu by pressing K instead of Esc. YO will adjust element tip lengths to maintain performance. This feature is handy when mechanically redesigning a Yagi. YO does not display untapered-equivalent element diameters. You can inspect them by saving an untapered AO output file. ---- MOUNTING BRACKETS ----------------------------------------- A conductive element-to-boom mounting bracket increases the electrical diameter of an element at the bracket. The increase depends on bracket size and shape. YO uses the method of D. Jaggard to model flat, rectangular mounting plates, Hy- Gain element clamps, L-angle brackets, U-channel brackets, Cushcraft element saddles, and general brackets of arbitrary geometry. For details of the method, see "Physical Design of Yagi Antennas." Use the Bracket menu to define a mounting bracket. Select bracket type with the tab key. After you enter bracket dimensions, YO calculates the length and diameter of an equivalent taper section. Its length is the half-length of the bracket (except for Hy-Gain clamps). Its diameter is electrically equivalent to the tubing and bracket combined. When you exit the Bracket menu, YO updates the taper schedule with the new section. If you enter zero for a bracket dimension (YO lists "None" for the equivalent taper), when you exit the Bracket menu YO removes sections that represent mounting brackets from the taper schedule. YO creates identical mounting brackets for all elements but no bracket for insulated driven elements (direct feed, hairpin match, dual driven elements, Hy-Gain clamps). A bracket is always represented by the first taper section and its tubing by the second. To remove a bracket from an individual element, 6 from the Taper Schedule menu add the length of the first taper section to that of the second, then zero the first. The length of a mounting bracket is its dimension parallel to the element. Enter the total length (both sides of the boom). Width is the dimension parallel to the boom. For flaring Cushcraft saddles, width is the distance between the rounded tops (for example, 1" for Cushcraft part number 190156). Thickness is metal thickness (both halves of a Hy-Gain clamp; ignore Cushcraft U-bolt backing plates which YO accounts for itself). Height is the outside-face flange dimension for an angle bracket, channel bracket, or Cushcraft saddle. YO assumes that elements touch only one surface of an angle or channel bracket. The channel-bracket model is accurate for elements mounted either inside the channel or on the outside surface. For a general bracket, consider the assembled bracket and tubing. Enter the cross-sectional perimeter (the distance an ant would travel when crawling around the assembly) and the cross-sectional area (the metallic area plus the area of any fully enclosed spaces). If you can fit the bracket inside a circle whose circumference is smaller than the bracket perimeter, enter the circumference instead of the perimeter. Before defining a Hy-Gain clamp, be sure you've measured the lengths of the first parasitic-element taper sections from the centerline of the boom. But measure the first driven- element taper section from the feedline/hairpin attachment point since that is the excitation location. A Hy-Gain clamp electrically passes the element through the boom. This shortens the clamp's electrical length. Fig. 10 of NBS Technical Note 688 suggests that an HF boom nullifies about one boom-radius of clamp length. YO therefore subtracts half the boom radius from taper sections that represent Hy-Gain half-clamps. See the YO.SET section to change this compensation. It's important to account for element mounting when modeling Yagis. Large mounting brackets can significantly alter performance. In extreme cases they may move a desired response completely outside a frequency band. Small brackets can distort a carefully optimized pattern at a spot frequency. But mounting methods that use compact hardware little larger than the element diameter generally require no correction. That's also true for insulated mounts. ---- BOOM CORRECTION ------------------------------------------- A conductive boom shortens the electrical length of an element that is close or passes through it. For through-the-boom element mounting typical of VHF/UHF Yagis, Ian White, G3SEK, curve-fit measurements by Guenter Hoch, 7 DL6WU, to derive the following boom-correction formula for noninsulated elements: C = (12.5975 - 114.5B)B^2 C is the element half-length correction and B is boom diameter, both in wavelengths. The experimental data for this formula came from boom diameters smaller than 0.055 wavelength (1.5" at 432 MHz). YO inhibits entry of larger values. G3SEK reports that the correction for insulated elements that pass through a boom with small plastic spacers is close to 50% of C. YO implements the Hoch-White formula in the Taper Schedule menu. Press I to define boom diameters in the first menu column. A separate boom diameter for each element accomodates tapered booms. For quick data entry for uniform- diameter booms, press = to equalize boom diameters below the cursor. Press I to toggle element insulation. To eliminate the boom, delete the boom column with Del. W2PV determined that mounting an element on a flat plate in contact with a boom electrically shortens its length by 6% of the boom diameter. He states that this small effect diminishes rapidly as the element is spaced away from the boom, even by a short distance. Because this correction typically yields a frequency shift of less than 10 kHz at 28 MHz, YO does not implement it. To change boom-correction parameters, see the YO.SET section. ---- CONDUCTIVITY LOSS ----------------------------------------- YO accounts for ohmic losses in Yagis due to imperfect element conductivity. Loss normally is small for tubing elements at HF, but it can be significant for rod elements at VHF/UHF. To calculate conductor loss, YO accounts for the sinusoidal element-current distribution, the length and diameter of taper sections, and the skin effect, which depends on frequency and material resistivity. Press C from the Elements menu to select conductor material. Pick one of the following: Silver Pure silver Copper Int'l Annealed Copper Std 6063 6063-T832 aluminum alloy 6061 6061-T6 aluminum alloy Zinc Pure zinc Brass Yellow brass (35% zinc) Steel Stainless steel (type 302) Note that the volume resistivity of 6061-T6 material is 23% higher than that of 6063-T832 alloy and 51% higher than that of pure aluminum. 8 Enter a resistivity value to model an unlisted material. Enter zero to model perfect-conductivity elements. The increase in forward gain that results is the total conductor loss. ---- PERFORMANCE TRADEOFFS ------------------------------------- Maximizing Yagi forward gain produces the strongest transmit signal in the forward direction. On receive, it yields the best signal-to-noise ratio for omnidirectional noise. But it also results in a narrow beam, a large backlobe, multiple sidelobes, low input impedance, and small SWR bandwidth. Yagis optimized for good patterns sacrifice gain, while those favoring wide SWR bandwidth give up gain and pattern quality. Because of these inherent tradeoffs, YO is designed to optimize a combination of forward gain, F/R, SWR, and input impedance. Use the Tradeoffs menu to select the performance parameters of interest and their relative importance. Use the Frequency menu to adjust the relative weight for each frequency. YO expresses parameter tradeoffs in percent to indicate relative importance. For example, to maximize forward gain regardless of the other parameters, set the gain tradeoff to 100%. To maximize gain but also take F/R into account, set the F/R tradeoff to some nonzero value, say 50%. This tells YO to weigh both forward gain and F/R when optimizing, and in this case, to give them roughly equal importance. YO does this by maximizing a weighted sum of the forward gain and F/R values. If you allow local optimization to run to completion, the resulting design will be optimal in the following sense: Any small design change that improves forward gain will necessarily reduce F/R. Any F/R improvement will degrade gain. In other words, there will be no free gain or free F/R to be had. In addition to gain and F/R, YO provides SWR and impedance tradeoffs. You can use the four tradeoffs in any combination. The SWR tradeoff minimizes SWR. The impedance tradeoff raises input resistance. This is useful when maximizing forward gain without F/R or SWR tradeoffs. When optimizing receiving antennas, select Mismatched Gain in the Options menu and set the SWR tradeoff to zero. Power reflected at the antenna terminals, which is reradiated and lost for receiving antennas, will be subtracted from the forward gain figure during optimization. This directly accounts for the effect of SWR on power delivered to the receiver. Equal values for the gain, F/R, and SWR tradeoffs often yield a reasonable, balanced Yagi design. (Normally the impedance tradeoff is unnecessary for multi-frequency designs, as is the SWR tradeoff for single-frequency designs). The default tradeoff values (press the = key to restore them) provide a starting point for custom optimization. Vary the 9 tradeoff percentages according to the relative importance you place on the design objectives. For example, to emphasize F/R, increase the F/R tradeoff value. If YO sacrifices too much gain for SWR, raise the gain tradeoff or lower the SWR tradeoff. By adjusting the tradeoff percentages, you can home-in on a design that meets your requirements. The design will be optimal in the sense that an optimal tradeoff will have been made among the performance parameters according to the importance you assigned to each. ---- GOOD-ENOUGH THRESHOLDS ------------------------------------ Sometimes a performance parameter may greatly exceed your design goal at one or two frequencies. For example, you might get 35 dB F/R at the top end of a band when all you really need is 20 dB. Whenever something like this happens, invariably another parameter is compromised to enable the extra performance. To address this problem, you might try reducing the F/R tradeoff. But this will affect all frequencies. You might reduce the high-frequency weight, but this will affect gain and SWR as well as F/R. Good-enough thresholds can help in this situation. These numbers provide limits beyond which YO does not regard a performance increase as an improvement. For example, if you set F/R OK > 20 dB in the Tradeoffs menu, whenever F/R exceeds 20 dB at any frequency, the optimizer will not regard it as an improvement. It will have no incentive to increase F/R at that frequency as long as it stays above 20 dB. This frees the optimizer to change the design in ways that maximize your other objectives, such as forward gain. (Good-enough thresholds are soft limits. They take effect gradually and smoothly near the limiting values.) It's best to disable the good-enough thresholds (by setting them to extreme values) when you first optimize a design. YO defaults to this condition. Some designs sacrifice little of one parameter to achieve high performance in another. If you habitually set good-enough thresholds near tolerable levels, you'll never uncover these remarkable designs. Use the tradeoff percentages to apportion optimizer driving power according to your design priorities. Use the good-enough thresholds to dynamically redirect that power once an objective has been met. Because performance characteristics tend to correlate across a frequency range, good-enough thresholds cannot completely flatten antenna response. But often they can help. ---- LOCAL OPTIMIZER ------------------------------------------- YO uses a conjugate-gradient optimizer that combines the Fletcher-Reeves and Polak-Ribiere methods. It works as follows: 10 First, YO changes each nonfixed element length and position by a very small amount, with all other dimensions unchanged, to calculate the sensitivity of the objective to each variable. The objective is the weighted combination of gain, F/R, SWR, and impedance defined by your tradeoffs and good- enough thresholds. The set of element sensitivities is called a gradient. The gradient is an n-dimensional vector, where n is the number of nonfixed element lengths and positions. The gradient points in the direction that locally maximizes the objective. YO recalculates Yagi response at points along the gradient until the objective no longer increases. YO then recalculates the gradient and modifies the search direction. The new direction incorporates prior gradient information in a way that minimizes unnecessary iterations. YO repeats this process until the optimizer can no longer improve the objective. This optimization technique is robust and efficient, locating optimal dimensions for an antenna without wild goose chases in a reasonable number of iterations. However, the method does not necessarily find the global optimum. The global optimum is the best possible set of antenna dimensions, where "best" is defined by your tradeoffs and good-enough thresholds. The global optimum is missed whenever the algorithm converges to a local optimum, where small movement in any direction degrades performance. Optimization using perturbation techniques is inherently nearsighted. If you're maximizing forward gain, for example, and the gain curve has a small bump at a certain set of element dimensions, the design may wind up on top of the bump because the algorithm cannot see beyond the immediate vicinity to the highest point of the overall curve. To visualize the problem, imagine trying to climb in dense fog to the very top of a mountain range with multiple peaks. The next section describes a global solution. Start the local optimizer by pressing Alt-Z. Press any key to interrupt it. Interrupting with Esc or a nonfunctional key causes no further action. Interrupt with a command key to execute the command. ---- GLOBAL OPTIMIZER ------------------------------------------ When you press Z instead of Alt-Z to start the optimizer, YO continues optimization after finding a local optimum. It continues by altering each nonfixed element length and position by a random amount and then restarting the conjugate-gradient optimizer. When the optimizer quits, YO compares the newly optimized performance with the prior result. If better, as defined by your tradeoff weights and good-enough thresholds, YO replaces the prior design. If worse, it discards the new design. YO then generates another random dimension variation and restarts the local optimizer. This process continues indefinitely. 11 Global optimization can efficiently locate overlooked optima. In effect, YO pops an optimized design off its local peak onto the slope of another and then climbs the slope. The longer you let it run, the more globally optimal will be the final result. You can profit from millions of Yagi evaluations. During global optimization YO normally displays only the best design so far found. To watch the design trials as they're being optimized, press W. Leaving YO unattended in watch mode can slow optimization due to display overhead; press W to toggle the display mode back to normal. You can abort the global trial you're watching and move on to the next by pressing X. If your judgment is good, this can save many iterations. When you're willing to be attentive, such directed optimization can provide the best design in the shortest possible time. The global optimizer uses Gaussian random deviates to vary element length and position (and for dual-driven designs, phasing-line impedance). The default standard deviations of 2% of element length, 20% of average element spacing, and 10% of phasing-line impedance work well for HF designs. Long-boom VHF/ UHF Yagis typically benefit from a smaller position deviation. Use the Tradeoffs menu to alter the default values. YO displays the number of optimizer restarts after the Yagi count in the lower-left screen corner. Press Y during optimization to display calculation speed in Yagis per second. This figure is useful when comparing processors. The value is an average whose accuracy increases the longer it's displayed. YO resets the average when global optimization commences or watch mode changes since these events alter display overhead. Always use the same Yagi file and menu settings when comparing speeds. Avoid moving the mouse because that adds overhead. Press Y to toggle the speed display off. ---- UNDOING CHANGES ------------------------------------------- Whenever you alter a design, YO automatically saves the design parameters in a last-in, first-out stack. The stack maintains a history of design changes. Use the Undo command to recall a previous design. Each press of U moves back one design. Press Alt-U to move all the way back to the first design in the stack. Use the Redo command to undo an Undo. Each press of R moves forward one design, while Alt-R moves all the way forward to the latest design. Undo and Redo let you experiment with abandon without fear of losing a design. They lessen the need to use the Save command to save intermediate results. They make it easier to undo manual changes since you don't need to remember parameter values. Finally, Undo and Redo provide a great way to compare results before and after optimization, especially when combined with graph overlays or graph comparison. 12 You can undo and redo within most menus without losing the menu. This feature is handy when experimenting with numerical values. Optimized designs are saved in the stack, but intermediate designs generated during optimization aren't. If you change a recalled design, the design history from that point forward is lost. ---- MATCHING NETWORKS ----------------------------------------- The Match menu lets you design a variety of matching networks. Use the tab key to select one of the following network types: Ideal Match The ideal match is an idealized, broadband matching network. It's modeled as a series reactor followed by a transformer, both lossless and ideal. The optimizer can minimize SWR while modeling another matching network, but it won't automatically adjust network parameters. The ideal match is useful during optimization because YO automatically adjusts series inductance or capacitance and transformer turns ratio to minimize SWR. Since the transformer has no reactance and that of the series component varies little over a band, SWR variation closely reflects the inherent impedance-bandwidth properties of the antenna. When graphing ideal-match SWR versus frequency, YO takes advantage of the many additional impedance values to refine the SWR curve. The refined curve shows the lowest weighted SWR across the passband, using weights linearly interpolated from the Frequency menu. (YO does not refine SWR for spot-frequency designs, nor does it update ideal-match component values after graphing.) Once you've optimized a design with the ideal match, you can design a practical matching network. Use graph overlays and the ideal-match SWR curve as a target when adjusting matching- network parameters. Direct Feed In direct feed the feedline connects to a split driven element without a matching network. While convenient, Yagis optimized for direct 50-ohm feed usually perform less well than matched, low-impedance designs. You can specify the dimensions of balun leads as well as the shunt capacitance between driven-element halves. See the next section for details. Feed spacing is the distance between balun leads at the driven element. 13 With direct feed you can model a matching scheme once used for some Mosley Yagis. Add shunt capacitance to raise the equivalent parallel input resistance. Then lengthen the driven element to cancel the capacitive reactance. This is the lowpass, L-network equivalent of a hairpin match. Hairpin Match A hairpin match is a highpass L-network that uses distributed reactances. The shunt hairpin raises the equivalent parallel input resistance to that of the feedline. A short driven element cancels the hairpin inductance. A hairpin match usually is fed with coax through a balun. YO models the hairpin as a rectangular loop. For typical hairpin shapes, this model is more accurate than a shorted transmission line. You provide hairpin-rod diameter, rod length, and center-to-center rod spacing. You can also specify the diameter and length of balun leads. Enter the length of one lead. YO assumes that the balun leads connect to the driven element at the hairpin-rod attachment points. Don't neglect to model balun leads. Even short leads can greatly alter feedpoint impedance. You can specify shunt capacitance between driven-element halves. This is particularly useful with Hy-Gain element clamps. Use 26.5 pF for 20-meter clamps and 9.5 pF for 15- and 10-meter clamps. Hy-Gain calls its version of the hairpin match a beta match. The hairpin rods straddle the boom and a shorting strap connects all three. The boom has a negligible effect on hairpin inductance, so a beta match is electrically equivalent to a conventional hairpin match. YO displays hairpin inductance. Use this value and the COIL.EXE program to design an air-core inductor to replace the hairpin. A coil may be more compact, more accessible, and easier to adjust (squeeze turns). Gamma Match YO uses equations by Shintaro Uda and insight by Dave Leeson to model gamma matches. You specify gamma-rod diameter, rod length, center-to-center spacing from the driven element, and series capacitance. To model a lead from gamma rod to coax connector, specify lead diameter and length. YO models the strap between gamma rod and element as an inductive reactance rather than as a short circuit. YO assumes the strap width plus thickness is equal to the rod diameter plus 1/16". 14 T Match A T match is a balanced gamma match. Rod length and capacitance values apply for each side of the driven element. Some T matches use a 4:1 balun made of a half-wavelength of 50- ohm coaxial cable. Enter a 200-ohm feed impedance for these systems, and specify the diameter and length of the lead from balun to coax connector. Other T-match systems use a 1:1 balun. Measure lead length from T-rods to balun for these systems. (YO assumes that lead spacing at the driven element is equal to rod spacing). To model a T match without capacitors, enter a large capacitance value (or zero, which YO interprets as infinity). To model a folded dipole, set the T-match rod length equal to the driven element length. If the driven element has just one taper section, YO will keep the rod length equal to the driven-element length when the latter changes. Driven-element electrical dimensions vary with T-match dimensions because the rod fattens the element and acts as a taper section. Because it's tricky to model this effect, you may need to adjust driven-element length when constructing a T match. -- The impedance value YO reports is always the unmatched impedance at the center of the driven element. The SWR value is that in the feedline and includes matching-network effects. By default YO uses a velocity factor of 0.975 for the transmission line formed by gamma- or T-match rod and driven element. To change this value, see the YO.SET section. Because it affects performance so little (except when adjacent elements are very close), YO fixes driven-element length for the ideal match during optimization. YO allows the length to vary for other matches to help lower SWR. You can override these defaults with the Elements menu. For a split driven element (direct feed or hairpin match), measure the length of the first taper section from the feedline attachment point since that is the excitation location. To avoid disrupting special feedpoint tapering, YO does not add or delete a driven-element mounting bracket when you switch to another matching network. The Yagi sketch shows which elements have brackets and can alert you to a problem. Matching-network performance depends on parameters that are difficult to measure and to model accurately. Calculated and actual matches may differ due to input-impedance modeling error, matching-component geometry simplification, stray reactances, capacitance-estimation error, lead-length indeterminacy, proximity of other conductors, and so on. When 15 constructing a Yagi, always make the matching network adjustable! Occasionally YO and NEC impedance values may differ significantly. This can limit accuracy when designing a matching network. After verifying a design with NEC, access the Match menu with Alt-M to use NEC impedance values instead of those calculated by YO. If you change driven-element length within the Match menu, YO does not invoke NEC to recalculate impedance. Instead, it uses new YO values compensated by the difference between prior NEC and YO results. Exit the Match menu and press the spacebar to verify a compensated-value design with NEC. To avoid displaying misleading impedance and SWR curves, YO inhibits the Graphs command within a Match menu invoked with Alt-M. ---- DUAL DRIVEN ELEMENTS -------------------------------------- YO can model Yagis that use two driven elements interconnected by a crossed phasing line. This type of feed was first popularized by KLM Electronics. It is used today on some Cushcraft and M-Squared Yagis. A dual-driven design may provide lower SWR across a wide passband. Press D in the Elements menu to toggle dual drive. Dual-driven Yagis use balanced feed at the forward driven element. The nominal feed impedance for KLM designs is 200 ohms, and the antennas use a 4:1 balun to 50-ohm coax. Phasing-line characteristic impedance varies with antenna model and line assembly, but it is in the 200- to 250-ohm range for KLM HF Yagis. YO uses default values of 230 ohms for phasing- line impedance and 200 ohms for feed impedance. Use the Match menu to specify other values. SWR of dual-driven Yagis is sensitive to phasing-line characteristic impedance. For KLM line construction, the impedance varies with phasing-line tension. Two insulators determine line spacing at the center crossover point, but the spacing elsewhere varies with line tension. This variation affects average line impedance. Measurement of a 14-MHz KLM Yagi with a time-domain reflectometer and a vector-impedance meter yielded an average impedance of 215 ohms when the phasing line was somewhat slack and 240 ohms when taut. YO uses a special Match menu for dual-driven Yagis. The first option allows automatic optimization of phasing-line impedance. Enable this option when you're willing to modify the phasing line. The second option maintains constant spacing between driven elements while allowing their joint position to vary. This lets you optimize joint driven-element position without reconstructing the phasing line. The next items specify phasing-line impedance, feed impedance, and phasing-line velocity factor. Finally, you can specify balun-lead diameter and length (one lead), and the spacing between leads at the driven element. 16 By default YO allows the lengths of dual driven elements to vary during optimization. One strategy for optimizing a dual-driven Yagi is to first optimize the antenna using a nominal phasing-line impedance of 230 ohms. When you have a good design, save it, then enable automatic optimization of phasing-line impedance. Reoptimize to see if significant improvement is possible. Use global optimization to randomly vary the initial impedance. To realize optimum SWR characteristics, you should be prepared to adjust a constructed dual-driven Yagi. The impedance and velocity factor of your particular phasing line may differ from the values modeled. It's easy to alter phasing- line impedance by changing the conductor spacing in the crossover region. Simply replace the original phasing-line spacers with ones larger or smaller of your own design. Experiments have shown that you can raise average line impedance 50 ohms by increasing the crossover spacing to a few inches, or lower it 50 ohms by decreasing spacing to a fraction of an inch. It's not necessary to measure the line impedance; simply adjust conductor spacing until the SWR characteristics match those predicted by YO. An alternative is to adjust driven-element lengths for minimum SWR. Don't neglect to model the flat straps KLM uses to connect parasitic element halves. The equivalent cylindrical diameter of a flat strap is half its width. The half-length of a strap is the distance from the center of the boom to the strap mounting bolt. Measure tubing length from this bolt rather than from the tubing end. Dual driven-element lengths often are short, and optimized designs may have closely spaced elements. Both of these factors tend to increase the impedance-modeling error. Always check SWR for dual-driven designs with NEC. When YO generates an AO output file for a dual-driven Yagi, it uses a current source at each driven element rather than modeling the phasing line with wires. Note that source magnitudes and phases will not remain strictly valid if you change the analysis frequency within AO. ---- FREQUENCY SCALING ----------------------------------------- YO can scale a Yagi design to another frequency while maintaining performance characteristics. This makes it easy to replicate a design on another band without designing a new antenna from scratch. The scaled design can use an entirely different taper schedule. Scaling can virtually duplicate response at a spot frequency, but performance across the new and original frequency bands will differ unless element diameters scale with frequency and the relative bandwidths are equal. This is unlikely. After 17 scaling a design, reset the low and high frequencies to the new band edges, check performance, and do a quick touch-up optimization if necessary. Frequency scaling in YO does not rely on a simple polynomial approximation for element self-impedance as does the W2PV scaling technique, nor does it attempt to match element reactances. Instead, YO iteratively recalculates element self- impedance with the method it uses for normal analysis while adjusting tip length. YO stops when the phase angle of the complex self-impedance has converged to that of the original element. To scale a design, enter new taper diameters and lengths in the Taper Schedule menu. (Don't bother with tip lengths since YO will adjust them during scaling.) Press C and enter the new design frequency. When you press Esc, YO adjusts element tips to yield a design at the new frequency electrically equivalent to the design that was current when you brought up the Taper Schedule menu. YO does not alter the dimensions of matching networks or mounting brackets when scaling a design. ---- OUTPUT FILES ---------------------------------------------- Use the Save command to save the current Yagi design in a file. You can specify a path with the filename. If you enter the name of a file that already exists, YO asks permission before overwriting. To quickly save, press Enter without a filename to overwrite the last file saved (no permission requested). If no file has been saved, YO writes the design to SAVE.YAG. Save follows the YO quick-check convention that pressing a command key a second time exits a menu (first keystroke only for Save and Notes). To save a filename beginning with V, press the spacebar and then enter the filename. In case you forget to save a design, YO automatically saves the current design in OUT.YAG when you exit the program. Recover the design by specifying OUT.YAG as an input file. Enter PRN or LPT for the output filename to print an output file. This will record antenna dimensions in a compact, easy-to-read format along with any design notes. If you specify the extension .ANT when saving a file, YO uses the AO file format. YO does not model matching networks or balun leads in AO output files, but it does normalize element positions so that the center of the boom is at X = 0. For YO- format output files, YO normalizes the reflector position to zero. Output files use current units. See the next section to control tapering in AO output files. 18 YO 7.5 saves boom, mounting-bracket, matching-network, and extra-frequency data in output files in such a way that earlier versions can read the remaining data. (But versions earlier than 7.0 will always use #12 wire for balun leads and a velocity factor of 0.975 for transmission lines.) ---- OPTIONS MENU ---------------------------------------------- The Options menu lets you control some additional aspects of YO. The E-plane is the plane containing Yagi elements. Select this plane to optimize the azimuth pattern of a horizontally polarized Yagi. The H-plane is the plane perpendicular to the elements. Select this plane to optimize the azimuth pattern of a vertically polarized Yagi. YO can account for mismatch loss for receiving antennas. Mismatch loss is the power reflected at the antenna terminals, which for receiving antennas is reradiated and lost. When the option is enabled, the gain figure on the main screen includes mismatch loss. The loss value itself is displayed following it in gray. YO includes mismatch loss when graphing, optimizing, and plotting. Mismatch loss applies to transmitting antennas only when the source and feedline impedances are close, as for a laboratory signal generator. You can disallow longer elements when optimizing an antenna with nontelescoping elements and you don't want to fabricate new ones. When combined with fixed element positions, this option lets you simply trim elements for improved performance. When you specify a fixed boom length, YO fixes the positions of the reflector and last director during optimization. YO also inhibits changes to these element positions with the mouse. YO can generate AO output files with tapered or with electrically equivalent untapered and boomless elements. Tapered output normally is best. Use untapered output to include boom correction or driven-element shortening due to a T match in AO models, to simplify AO models, or to examine equivalent taper diameters. YO output files are always tapered and boomful. You can control NEC accuracy and execution time by varying its segmentation density. See NY.DOC for guidance in setting this parameter. By default YO uses the dimension units of the Yagi file. See the YO.SET section to have YO convert file dimensions to the units specified in the Options menu. Output files use current units. Wavelength units are based on the wavelength of the middle analysis frequency. 19 ---- ELEMENT-CURRENT PROFILE ----------------------------------- YO draws a profile of element-current magnitudes along the Yagi sketch. Current is plotted linearly, with zero at the boom. YO normalizes the profile so that curve height remains constant from screen to screen. The up-arrow and down-arrow keys scale the curve. Scaling can bring out small detail or position the curve for screen printout. You can scale the profile during optimization without interrupting the optimizer. To disable the profile, set its color to the background color. The element-current profile is fascinating to ponder for its deeper significance, but it has a practical use as well. It can reveal elements with questionable tuning, particularly for long-boom Yagis. Elements with abnormally high or low current are easy to spot in the profile. YO computes the element-current profile at the middle analysis frequency. ---- FREQUENCY GRAPHS ------------------------------------------ YO graphs forward gain, F/R, SWR, and impedance curves versus frequency. Regular command keys are active while viewing the graphs. For example, press M and G to alternate between the Match menu and graphs when adjusting a matching network. Press F1 to list graphs commands. YO automatically scales the graphs. To generate the curves, it recalculates Yagi response every two pixels along the frequency axis. To change this resolution, see the YO.SET section. The gain curve includes mismatch loss when the option is selected. The F/R curve reflects the current setting for F/R region. The impedance curve graphs input resistance. YO graphs spot-frequency designs over a +/- 0.5% frequency range. Press F8 to save a graph (the first graph viewed is automatically saved). Later while viewing another, you can recall the saved graph for comparison by pressing the spacebar. The spacebar then switches between the two graphs. Better still, press O to overlay two graphs. This is a great way to compare the performance of two designs. ---- PATTERN PLOTS --------------------------------------------- The pattern sketches YO draws on the main screen omit grids and scales to reduce screen clutter. For speed, they have just 5-degree resolution. Use the Plot command to draw high- resolution patterns with grids, scales, and annotation. YO calculates the patterns at the middle analysis frequency with 1-degree resolution. The Plot command generates a plot file and then displays it. YO saves the patterns in files so you can later review and 20 compare them. A plot file uses the Yagi filename and the extension .PF. If you plot, change antenna geometry, and press P again, YO overwrites the plot file with new patterns. If you press Alt-P, YO generates a new plot file. Each new file has an incrementing digit appended to its name. Use Alt-P to avoid overwriting plot files generated earlier in a session. You can review a plot from the command line by typing YO PLOT. This bypasses the analysis part of YO and lists plot files in the current directory. Provide a directory name on the command line to list its plot files. To display a pattern immediately, provide the plot-file name on the command line, for example, YO PLOT HG204BA. YO generates plot files in the OpenPF plot-file format. OpenPF is an open, nonproprietary standard for files containing electromagnetic-field data. See OPENPF.DOC for details. YO creates OpenPF plot files with phase data for free-space elevation patterns. These files provide complete elevation data for use with the TA Terrain Analyzer program. YO centers the Yagi on the boom when writing a plot file to provide a consistent phase reference. YO provides two radial scales for polar plots. The default ARRL log-dB scale compresses minor lobes toward the polar center, emphasizing major lobe shape. This scale is widely used in amateur publications. The center of the plot is minus-infinity dB on the log-dB scale, but there is little plot area below -40 dB. The linear-dB polar scale provides much more area below -20 dB. It's good for examining minor lobes that may be hard to see in a log-dB plot. The linear-dB scale cuts off at -50 dB at the polar center. YO-Professional provides a third polar scale, one linear in field strength. This linear scale is useful when comparing YO plots with those generated by automatic pattern-plotting equipment on an antenna test range. See the YO.SET section to use the linear-dB or linear polar scale by default. Dots in the sparse radial lines are spaced 2 dB, while those in the two outer circles are 1 and 2 degrees apart. These calibrations let you read directivity values from the plots with good accuracy. Select rectangular coordinates with the R key. This coordinate system can reveal minor-lobe detail even better than can the linear-dB polar plot, but the overall pattern shape isn't as easy to grasp. The X-axis is degrees off boresight and the Y-axis is antenna response in dB. Use the Y key to change the Y-axis cutoff. This parameter is always negative, but you can enter it without a minus sign for convenience. Change the 21 X-axis cutoff with the X key. The value you enter is rounded to the nearest multiple of 10 degrees. Press the O key to overlay two patterns. Annotation color matches pattern color. YO displays filenames instead of titles when the titles differ. Press the C key to compare two patterns. After selecting a second plot file, press the spacebar to switch between the two patterns. This is a good way to compare patterns so complex that an overlay muddles detail. YO identifies free-space patterns as E-Plane or H-Plane. It uses the annotations Azimuth or Elevation for over-ground patterns. Polar plots are perfectly circular on monitors with standard 4:3 aspect ratio. Adjust your monitor's vertical- height control to correct elliptical plots. ---- ARRAYS OF YAGIS ------------------------------------------- Given the pattern of a single antenna, YO can synthesize the pattern and gain of a rectangular array of Yagis. Since the synthesis is very fast, you can quickly investigate the properties of large Yagi arrays. Press P to plot a single-Yagi pattern. Then press the + key to increment the antenna count in the plane you're viewing. Press the - key to decrement the count. Press Enter to switch to the other plane, and use + and - to set the Yagi count in that plane. Vary array spacing with the up/down arrow keys. Use the Home and End keys for finer resolution and PgUp and PgDn for coarser. Array spacing varies in the plane you're viewing. Spacing can differ in the two planes, but all Yagis are uniformly spaced within a plane. YO uses wavelength spacing units by default. Press the U key to change to inches or millimeters. You can compare the pattern of a synthesized array with another pattern, but you can't normalize or overlay it. YO synthesizes free-space patterns only. YO ignores mutual impedances between individual Yagis when synthesizing array patterns. For most Yagis, interaction is small at realistic array spacings. YO uses pattern integration to estimate array gain. The gain estimate usually is within a few tenths of a dB of true array gain, and it peaks at a spacing close to the true gain peak. By varying array spacing and observing the gain figure and pattern sidelobes, you can quickly determine the best geometry for a desired gain/ pattern tradeoff. Then you can analyze the array from the 22 command line with NEC/Yagis to account for all mutual impedances. ---- GAIN FIGURE-OF-MERIT -------------------------------------- For free-space, single-Yagi models, YO estimates the maximum forward gain practical on the boom length. YO displays the difference between forward gain and this maximum at each frequency. This number provides a handy gain figure-of-merit (FOM) that shows at a glance how close a design comes to realizing all the gain it can. To calculate maximum practical gain, YO uses the following formula: G = 10Log(5.8B + 3.7) G is gain in dBd and B is boom length in wavelengths. The formula yields figures within a couple tenths of a dB of reasonable expectations for booms as short as 0.4 wavelength. The formula does not estimate maximum possible gain. That condition leads to very low impedances and extremely narrowband designs. Instead, the formula predicts the gain of practical, near-maximum-gain Yagis with real-world losses and design compromises. Thus, some designs may exhibit a positive FOM. See the YO.SET section to change the formula coefficients. The gain FOM is not intended as an overall quality rating for Yagis. It rates gain/boom-length efficiency only. Most Yagi applications require good patterns and this constraint always limits forward gain. ---- MODELING LIMITATIONS -------------------------------------- YO accurately predicts antenna performance as long as you carefully characterize element tapering and mounting, and you follow the guidelines in this section. YO is calibrated to NEC, the Numerical Electromagnetics Code. YO closely tracks NEC for element diameters up to 0.01 wavelength. Yagis with thicker elements may exhibit some frequency offset from calculated performance. Still, YO agrees reasonably well with NEC for diameters up to about 0.04 wavelength. (When comparing YO and NEC results, make sure NEC has converged. Some NEC models require up to 60 segments/ halfwave for full convergence.) Models with input impedances of just a few ohms are not likely to be accurate. At these impedance levels, element currents are large and fields nearly cancel. This condition magnifies small model inaccuracies. There are practical reasons to avoid very low-impedance designs as well: Dimensions become critical, skin effect may cause excessive loss, impedance 23 matching becomes difficult, and bandwidth can be severely restricted. Accuracy may degrade somewhat for designs with half- elements shorter than 0.19 wavelength or longer than 0.26 wavelength. Such dimensions may occur for very long Yagis, very wideband designs, or for dual driven elements. YO may understate SWR at the band edges for Yagis with very closely spaced elements, particularly more than two such. Always check SWR-critical designs with NEC. To prevent gross modeling errors, YO won't let you enter element half-lengths shorter than 0.15 wavelength or longer than 0.3 wavelength. It won't let you position elements closer than 0.05 wavelength (see the YO.SET section to change this limit). Finally, it disallows taper diameters greater than 0.05 wavelength. ---- VERIFYING DESIGNS WITH NEC -------------------------------- From the command line you can verify any YO design with NEC/Yagis by typing NY filename, where filename is a YO file (the .YAG file extension isn't needed). NEC will report forward gain, F/R, input impedance, SWR, and losses, and then display patterns. To invoke NEC from within YO, press the spacebar. YO will generate an output file for NEC and exit. The YO.BAT batch file then takes over, sequencing NEC through an analysis equivalent to that on the YO screen. When NEC finishes, YO.BAT restarts YO. YO reloads the output file but displays NEC performance figures and pattern sketches in place of its own. Press the spacebar to toggle between NEC and YO results for comparison. If you alter the design, the spacebar will reinvoke NEC for an updated analysis. Use the Options menu to set NEC segmentation density. NEC generates high-resolution patterns at the middle analysis frequency. To display them, press P to plot YO patterns and then overlay or compare the NEC.PF plot file. The versus-frequency graphs always are YO results, never those of NEC. By default, NEC analyzes over-ground antennas using lossy earth, not perfect-conductivity ground as does YO. Put GND=0 in YO.SET whenever you need an exact comparison with YO over-ground results. See NY.DOC for more information about ground models. Press Esc to interrupt NEC analysis and return to YO. 24 ---- NOTEPAD --------------------------------------------------- Use the Notes command to enter, edit, and display design notes. YO displays the Yagi title at the top of the notepad and you can edit it, too. YO saves the title and notes in all output files. The notepad editor automatically wraps words within a paragraph. A blank line or a line with leading space begins a new paragraph. Use the arrow keys, Home, End, PgUp, and PgDn to move the cursor. Press Ctrl with the right or left arrow to move the cursor to the next or previous word. Del and Backspace delete characters, Alt-W deletes a word, and Alt-D deletes a line. See the YO.SET section to redefine the keys for word delete and line delete. ---- SCREEN ---------------------------------------------------- Press F5 to change the colors of any screen. Use the left- and right-arrow keys to select a screen item, which blinks once when selected. Then use the other listed keys to alter the red, green, and blue color components. When you're done, press Esc. YO saves color codes in the YO.INI initialization file in the directory containing YO7.EXE. YO also saves menu settings in YO.INI each time it terminates. At the next session, YO initializes its menus to the saved values, restoring them. To reset everything to default settings, delete YO.INI. To export YO graphics, press F9 to save a color image of any screen in the PCX file format. The output filename is the Yagi filename with PCX extension. If you press F9 again, YO appends an incrementing digit to the filename and creates another file. YO prints screens to HP LaserJet-compatible printers (including DeskJets) and to Epson-compatible, dot-matrix printers. Press PrtSc to print any screen. See the YO.SET section to configure YO for your particular printer. YO prints monochrome images only. To print color images, generate a PCX file and print it with other software. Except for LaserJet landscape mode, YO does not eject a page after printing a single screen. This lets you print two screens on one page (24-pin dot-matrix images are too big to allow this). YO will automatically eject the page after it prints a second screen. To manually eject a single-page print, press F12. To keep the mouse cursor from appearing in a screen print, move it beyond the right edge of the screen. 25 ---- MOUSE ----------------------------------------------------- You can use a mouse to select Yagi files, but its main purpose is to let you change antenna geometry and optimization tradeoffs graphically. Grab an element at its center to change its position. Grab it at the top to change its length. To simultaneously alter position and length, grab an element at the bottom. Press U to undo mouse changes. You can adjust element dimensions with the mouse during optimization. This feature is handy for nudging the optimizer toward a geometry you believe to be optimum. Grabbing an element during global optimization activates watch mode. YO displays a triangle whenever the Main menu disappears. The vertices of the triangle represent tradeoffs. The location of the dot within the triangle reflects the current tradeoff settings. Grab the dot with the mouse and move it to alter the tradeoffs. When you release the mouse button, YO starts the global optimizer. Grab the dot during optimization to change tradeoffs on the fly. The purpose of the tradeoffs triangle is to provide a simple, intuitive way to explore Yagi performance space. Simply move the dot around within the triangle to find the most attractive design. On a very fast computer, the design displayed is quickly optimal (locally, at least). Pressing the right mouse button is like pressing the Esc key. ---- YO.SET ---------------------------------------------------- The YO.SET file sets YO parameters that aren't changed often enough to warrant inclusion in YO menus. Use any text editor to create YO.SET. Put each parameter on a separate line. Entries can use upper or lower case. YO looks for YO.SET in the directory containing YO7.EXE. Directories Once you accumulate many Yagi and plot files, it's convenient to organize them into directories. You might use the current directory for experiments, saving optimized files elsewhere. YO.SET can tell YO to automatically reference certain directories. YO uses the directories only for reading files. Define directories like this: YAG=HFBEAMS VHF KLM 50MHZ . PF=. HFPLOTS 2MPLOTS LONGYAGS 26 The dot represents the current directory. It can appear anywhere (or nowhere) in a list. You can define as many directories as you like. You may find it convenient to use the same set of directories for both Yagi and plot files. You might organize directories by Yagi type, frequency band, designer, and so on. YO (YO PLOT) begins by listing files in the first directory on the YAG (PF) list. After the files it lists the names of the other directories. Select one to list its files. You can specify a file on the command line from one of the directories without typing its path. YO searches the directories in the order they appear in the directory list. You can specify a directory (not necessarily from the list) on the command line to list its files. To list files in the current directory when it's not first on the list, type YO . (YO PLOT .). Gain Reference To display gain figures in dBd rather than dBi, do the following: DB=dBd To use dBd in free space and dBi over ground, do this: DB=dBdi The gain reference for dBd is the peak gain of a halfwave dipole in free space. The reference for dBi is the gain of an isotropic antenna in free space. An isotropic antenna radiates uniformly in all directions. A dipole has 2.15 dB gain over an isotropic antenna, so YO converts from one gain reference to the other by adding or subtracting 2.15 dB. Units YO normally uses the dimension units of the Yagi file. To automatically convert file dimensions to the units specified in the Options menu, do this: UNITS=MENU Polar Scale YO normally uses a log-dB polar scale to display patterns. You can specify a linear-dB scale as the default with this: POLAR=LINDB 27 To specify a linear scale for YO-Professional, do this: POLAR=LINEAR Plot Normalization YO normally scales the low-, middle-, and high-frequency patterns independently. This gives each plot the same amplitude in the forward direction, making it easy to compare patterns. But gain differences aren't apparent. To normalize plots so that all patterns are drawn to the same scale, do the following: NORM=ON Printer Type YO defaults printed output for HP LaserJet printers in portrait mode. To try a landscape plot (not available on all printers), do this: PINS=HPLJ L For a bigger landscape plot, add a B: PINS=HPLJ LB YO normally draws LaserJet plots with a border. To eliminate the border, add an X to other options like this: PINS=HPLJ LBX YO can transfer compressed data to PCL 5 laser printers for faster output. To try this, add a C: PINS=HPLJ LBXC For an HP DeskJet printer, use the following: PINS=HPDJ Add an X to eliminate the border. No other HPLJ options work. For a 9-pin dot-matrix printer, do the following: PINS=9 Some 9-pin printers don't recognize the special line- spacing command YO uses to reproduce an exact screen image. If your printer won't print YO screens, try this: PINS=9ALT Even with 9ALT, YO may be unable to print to certain old 28 9-pin printers. For a 24-pin dot-matrix printer, do this: PINS=24 Printer Port Output goes to LPT1 by default. For a different port, use one of the following: LPT=LPT2 LPT=LPT3 Notepad Delete Keys Use WDEL and LDEL to redefine the notepad keys for word delete and line delete. Use ^ to indicate the Ctrl key and Alt- to indicate the Alt key in combination with a letter key. You can also use ^Enter and ^Bksp. Use F1 through F10 for function keys. For example: LDEL=^Y LDEL=F3 WDEL=Alt-K WDEL=^Bksp Graph Resolution When graphing performance versus frequency, YO recalculates Yagi response every two pixels. You can trade resolution for speed by changing the number of pixels per frequency step with the following: PIX=number Minimum Element Spacing YO normally won't position elements closer than 0.05 wavelength. To change the spacing limit, do this: MINSP=spacing Accuracy may degrade when elements are very close. Use a setting below 0.05 wavelength only when you're willing to frequently check results with NEC. 29 Gamma/T Velocity Factor By default YO uses a velocity factor of 0.975 for the transmission line formed by a gamma- or T-match rod and driven element. To change this value, do the following: VEL=value Use the Match menu to set the phasing-line velocity factor for a dual-driven design. Gain Figure-of-Merit To change coefficients of the gain figure-of-merit formula, do the following: FOM=a b The formula is G = 10Log(aB + b), where G is maximum- practical gain in dBd and B is boom length in wavelengths. When FOM isn't defined, YO uses a = 5.8 and b = 3.7 by default. To disable the FOM display, do this: FOM=0 Boom Correction To model through-the-boom element mounting, YO uses a formula of the following form: C = (a + bB)B^2 C is half-element shortening and B is boom diameter, both in wavelengths. Default values are a = 12.5975 and b = -114.5. To change these coefficients, do the following: BOOM=a b c d e Parameter c is the maximum value of B for which the formula is valid. Parameter d is the percentage of C to apply for insulated elements. Parameter e is the percentage of boom diameter to shorten a Hy-Gain clamp to account for boom proximity. Default values are c = 0.055, d = 50, and e = 50. VESA BIOS Extension YO uses the VESA BIOS Extension (VBE) for smooth screen animation. VBE may be included in your VGA card's video BIOS, it may be provided as an optional file, or it may not be implemented at all. YO automatically detects VBE. It displays "No VBE" on the file screen when it can't find it, and "Incompatible VBE" when VBE is present but unusable. YO can't 30 always determine that a VBE implementation is incompatible. If your YO screens look or act funny, or if YO won't even execute, try inhibiting VBE with this: VBE=NO YO displays "VBE ignored" to confirm. 31 INDEX # key 2 * key 5 + key 22 - key 22 .ANT 18 .INI 25 .SET 26 .YAG 1, 24 180 to 180 degrees 3 1:1 balun 15 200-ohm feed 15 4:1 balun 15, 16 50-ohm feed 13 6061-T6 8 6063-T832 8 = key 4, 5, 6, 9 ^Del 4 ^Ins 4 ~ key 5 Accuracy 1, 4, 19, 21, 24 Alt-D 25 Alt-M 16 Alt-P 21 Alt-R 12 Alt-U 12 Alt-W 25 Alt-Z 11 Aluminum 8 Always make the matching network adjustable! 15 Analysis plane 3 AO output file 6, 17, 18, 19 Array gain 22 Arrays of Yagis 22 ARRL 6, 21 Arrow key 1, 2, 5, 22, 25 Assembly language 1 Asterisks 3 Automatic optimization of phasing-line impedance 16, 17 Azimuth 22 Backlobes 3 Backspace 25 Balun 14 Balun leads 13 Batch file 3, 24 Beta match 14 Blink 5, 25 Boom Correction 7, 30 Boom length 23, 30 BOOM= 30 Bracket menu 6 32 Brackets of arbitrary geometry 6 Brass 8 C 18 Calculation speed 12 Clamps 14 Coefficients 23 COIL.EXE 14 Color 25 Command key 11 Command line 1, 27 Compare two patterns 22 Comparing processors 12 Comparing YO and NEC results 23 Complex self-impedance 18 Conductive boom 7 Conductivity Loss 8 Conjugate-gradient optimizer 10 Constant spacing between driven elements 16 Convert file dimensions 27 Copper 8 Count elements 5 Crossover spacing 17 Ctrl 4, 25 Current 20 Current directory 1, 26 Current profile 20 Current source 17 Cursor 2, 5, 25 Cushcraft element saddles 6 Cylindrical equivalents 5 D key 5 DB= 27 DBd 27 DBdi 27 DBi 27 Default 2, 3, 5, 9, 10, 12, 15, 16, 19, 21, 22, 24, 25, 27, 29, 30 Del 4, 25 Design sensitivity 5 DeskJet 25, 28 Dev 12 Dimension tolerances 5 Dipole 27 Direct Feed 13 Directed optimization 12 Directories 26 Directory 1, 21 Directory list 27 Disable the good-enough thresholds 10 Disable the profile 20 Disallow longer elements 19 DL6WU 7 Dot 26 Dot-matrix printers 25 Down-arrow 20 33 Dual driven elements 5, 16 Dummy reflector 5 E-plane 19, 22 Earlier versions 19 Earth 4 Eject 25 Element currents 3 Element reactances 18 Element Tapering 5 Element-current distribution 8 Element-Current Profile 20 Elements 29 Elements Menu 4 Elements with questionable tuning 20 Elevation 22 Elevation angle 2, 4 Elliptical plots 22 End 1, 22, 25 Enter 1, 2, 18 Epson-compatible 25 Equivalent cylindrical diameter of a flat strap 17 Erase the Main menu 2 Erase the tradeoffs triangle 2 Error 3, 5, 15, 24 ERRORLEVEL 3 Esc 2, 6, 11, 18, 24, 25, 26 Expand or shrink the design 5 Experiment 2, 5, 26 F/B 3 F/R 3, 9 F/R region 3, 20 F/R tradeoff 9 F1 2 F12 25 F5 25 F8 20 F9 25 Feed impedance 16 Feed spacing 13 Feedline 14 Field strength 21 Figure-of-merit 23, 30 Filename beginning with V 18 Final dimensions 5 Fixed boom length 19 Fixed element dimensions 5 Flat, rectangular mounting plates 6 Fletcher-Reeves 10 Folded dipole 15 FOM 2, 23 FOM= 30 Forward gain 1, 9 Fractional data 2 Free space 4 34 Frequency band 3 Frequency Graphs 20 Frequency Menu 3 Frequency range 20 Frequency Scaling 17 Front-to-back 3 G3SEK 7 Gain figure-of-merit 2, 23, 30 Gain FOM 23 Gain Reference 27 Gain tradeoff 9 Gain/boom-length efficiency 23 Gain/pattern tradeoff 22 Gamma Match 14 Generalized F/B 3 Global Optimizer 11 GND= 24 Good-Enough Thresholds 10 Gradient 11 Graph Resolution 29 Gray 19 Ground 4 H-plane 3, 19, 22 Hairpin inductance 14 Hairpin Match 14 Handy 2, 5, 6, 13, 23, 26 Height Menu 4 HF Yagis 16 High-resolution patterns 20 Highlight 1, 5 History of design changes 12 Hoch 7 Hoch-White formula 8 Home 1, 22, 25 Hy-Gain element clamps 6, 14 Ideal match 13 Impedance 3, 20, 23 Impedance tradeoff 9 Impedance value 15 Inches 22 Ins 4 Insulated elements 8 Intermediate designs 13 Interprets as infinity 15 Interrupt NEC 24 Isotropic antenna 27 Iteration count 3 Jaggard 6 James Lawson 6 35 K 6 Keep the initial electrical design 6 KLM 16 L-angle brackets 6 L-network 14 Lambda symbol 3 LaserJet 25, 28 LDEL= 29 Lead-length indeterminacy 15 Leeson 6, 14 Lightbar 1 Line delete 25, 29 Linear scale 21 Linear-dB polar scale 21 Listing files 27 Local Optimizer 10 Log-dB scale 21 Long-boom Yagis 20 Lossy earth 24 Low impedances 23 LPT 18, 29 LPT= 29 Main menu 2 Match menu 13, 16 Matching Networks 13 Maximum forward gain 23 Maximum practical gain 23 Menu settings 25 Millimeters 22 Minimum Element Spacing 29 MININEC 1 MINSP= 29 Mismatch loss 2, 19, 20 Modeling Limitations 23 Mosley Yagis 14 Mounting Brackets 6 Mouse 1, 26 Mouse cursor 25 Mutual impedance 4, 22 NBS Technical Note 688 7 NEC 1, 19, 23, 24 NEC.PF 24 NEC/Yagis 4, 22, 24 Network 13 Nonfunctional key 2 Nontelescoping elements 19 NORM= 28 Normalize plots 28 Notepad 25 Notepad Delete Keys 29 Notes 25 Numerical Electromagnetics Code 1, 23 NY.DOC 19, 24 36 Objective 11 Ohmic losses 8 Omnidirectional noise 9 OpenPF plot-file format 21 Optimize joint driven-element position 16 Optimum SWR characteristics 17 Options Menu 19 Other 1 OUT.YAG 18 Output filename 18 Output Files 18, 19, 25 Overlay two patterns 22 Parameter tradeoffs 9 Path 27 Pattern integration 22 Pattern Plots 20 PCX 25 Peak backlobes 3 Pentium processors 1 Percent 9 Perfect ground 4 Perfect-conductivity elements 9 Perfect-conductivity ground 24 PF 21 PF= 26 PgDn 1, 22, 25 PgUp 1, 22, 25 Phase data 21 Phasing line 16 Phasing-line tension 16 Physical Design of Yagi Antennas 6 PINS= 28 PIX= 29 Pixels per frequency step 29 Plane 22 Plot file 20, 26 Plot Normalization 28 Polak-Ribiere 10 Polar scale 27 POLAR= 28 Power 3 Print an output file 18 Print any screen 25 Printer Port 29 Printer Type 28 PRN 18 Profile 20 Proximity of other conductors 15 PrtSc 25 Quick check 2 Quick-check convention 18 Radial scales 21 Random deviates 12 37 Random dimension variation 11 Randomly vary the initial impedance 17 Reactance 14 Real earth 4 Rear horizon 3 Recall a previous design 12 Receiving antennas 9, 19 Rectangular array 22 Redo 12 Resistance 9 Resistivity 8 Resolution 22 Respacing elements 4 Restart count 3 Right mouse button 26 RMS weighting 3 Rounding 5 Save a graph 20 Save and Notes 2, 18 SAVE.YAG 18 Scale 17, 28 Screen 25 Screens look or act funny 31 Segmentation density 19 Self-impedance 5, 18 Shunt capacitance 14 Sidelobes 3 Signal-to-noise ratio 9 Silver 8 Sketch 5, 15, 20 Skin effect 23 Soft limits 10 Source magnitudes and phases 17 Spacebar 2, 18, 20, 22, 24 Spacing limit 29 Speed display 12 Spot frequency 3, 17 Spot-frequency designs 20 Stack 12 Stacking distance 4 Stainless steel 8 Standard deviations 12 Strap 14, 17 Strategy for optimizing a dual-driven Yagi 17 Stray reactances 15 SWR 9 SWR bandwidth 9 SWR for dual-driven designs 17 SWR value 15 Synthesize 22 T Match 15 TA Terrain Analyzer 21 Tab key 6, 13 Taper Schedule menu 6 38 Tapering algorithm 5 Through-the-boom element mounting 7 Tip length 18 Title 25 Total conductor loss 9 Tradeoff percentages 10 Tradeoffs 9 Tradeoffs menu 3, 9, 10, 12 Tradeoffs triangle 26 Triangle 26 Tweezers 4 Two-element Yagi with driven element and director 5 U key 22 U-channel brackets 6 Uda 14 Undo 12 Units 18, 19, 22, 27 UNITS= 27 Untapered and boomless elements 19 Untapered-equivalent element diameters 6 Untapered-equivalent lengths 5 Up-arrow 20 Using YO 1 VBE 30 VBE= 31 VEL= 30 Velocity factor 15, 16, 17, 19, 30 Verifying Designs with NEC 24 VESA BIOS Extension 30 VHF/UHF 7, 8, 12 W 12 W2PV 1, 6, 8, 18 W6NL 6 Watch the design trials 12 Wavelength 22 Wavelength units 19 WDEL= 29 Weighted sum 9 Weights 4 White 7 Whole-element length 4 Wild goose chase 11 Wire gauge 2 Word delete 25, 29 Wraps words 25 X 12 X key 5 X-axis cutoff 21 Y 12 Y-axis cutoff 21 YAG= 26 39 Yagi files 1 Yagi performance space 26 Yagis per second 12 YO PLOT 21 YO-Professional 4, 21, 28 YO.BAT 24 YO.INI 25 YO.SET 26 YO7.EXE 25, 26 Z 3, 11 Zinc 8 40