Design, Construction and Performance Notes

BACKGROUND

Yes, I undertook the HUGE process of building an array of four of the 50' long M-Squared 6M9KHW yagis - I highly recommend this array as a "realistically manageable" size, yet with enough gain to really get the job done effectively.  At the time of its raising (October 27, 2001), it was the third largest fully steerable ham radio antenna array for this frequency band in the world.  Why is it so important to have a fully steerable antenna, you might ask?

There are very few 6m DX stations with EME tracking capability, so it is ESSENTIAL to have elevation capability up to 65 degrees.  With my single yagi, I can only elevate to 45 degrees, which is really unfortunate because I was missing EU moonset under the best times of month (which in 2001 were coinciding with most northerly declination, putting the moon WAY HIGH here!).  The new GLEAP array is capable of elevation to 90 degrees.  Although this extra margin of elevation not necessary at this latitude for moontracking, it may be useful for driven element maintenance access, if required, sometime in the future.

 It is very desirable to have at least 1500W on 6m EME if you want to have any success.  Conditions on 6m can be VERY erratic - it is a terrible band for EME....however, it is the ONLY way to work any 6m DX from up here in the "geomagnetic far north".  I think if you put up an array such as four of the 6M9KHW yagis, you will have a killer system for working other kinds of propagation, too (if you happen to live in a place that HAS any other kind of propagation- HI).  And, it should be possible to more easily work DX stations on their moonset with a kw and a single good yagi.

Most of the folks who have tried 6m EME and given up due to lack of success and/or activity have had very little antenna gain, so have been severely limited by the number of  larger stations they could work.  Note that even a four bay array of 5 or 6 element yagis only has the same amount of gain as my single 6M35WL/6M11JKV yagi, which I consider to be the minimum antenna size for 6m EME.  With my single yagi, I DO hear my own echoes  when I am pointed skyward (no ground gain) and the conditions are optimum (quiet background sky, moon near perigee, no auroral disturbances, etc.) but performance is quite erratic.  Just as on 2m EME, though, there is a "magic minimum point" in antenna gain where you can begin to have some reasonable success....and increasing the gain above that point dramatically increases the number of stations you will be able to contact.  The 4.5 dB additional gain from this four yagi array should greatly improve the reliability for EME QSO's with other smaller stations.

MINIMUM ARRAY SIZE

Having said all that about my array, I have to add now that very successful 6m EME results can now be obtained by using JT44 and a smaller array.  I truly believe that a very practical sized array for success using JT44 is an array of four 6M5X yagis.  This sized array is very simple to build and elevate, and has approximately the same gain as my single yagi, with which I hear my own echoes while pointing upward.  As long as it has 14 or 15 dBd gain, such a steerable array should have no trouble making contacts with larger 6m stations on CW, or with many stations on their horizon using JT44.  I suggest constructing such an array using a "double H frame" of aluminum tubing.  It could be very conveniently mounted on a 25' high tower.

MY ARRAY DESIGN

The array was designed to be able to provide a stacking distance of 31'' wide x 29.5' high, with the center of the array at 32' above ground level.  Here is the expected performance of the array:

While pointed at the horizon, the new array appears to have more gain than the maximum ground gain lobe of the single 6M35WL/6M11JKV at 70', over the entire range of 2.5 to 9.5 degrees elevation.  Therefore, I expect it to be a very good performer when the moon is on the horizon, as well as for use on ground wave and during ionospheric openings.
 

SPACE PLANNING AND PREPARATION

Over 2 dozen trees were cut down and stacked in the spring of 2001 to make room for the new array.  Before any excavation could be done, all the slash (the limbs) needed to be burned.  Here we are rushing to burn the slash in the spring before it becomes too dry, and they no longer permit such burning. (You can see the 2m EME array is in the background). Clearance for the array was calculated using my old DOS program "SPACE" (available for free download from the software section of my web site).  Below  is a printout from that program, showing  what the clearance looks like for this array.

CLEARANCE CALCULATIONS FOR ARRAY OF FOUR 6M9KHW 6M YAGIS AT W7GJ

     HORIZONTAL SPACING BETWEEN                   TOWER HEIGHT= 32
     OUTERMOST YAGIS= 31
                                                  VERTICAL MAST HEIGHT FROM
     LENGTH OF REAR HALF                          PIVOT TO LOWEST YAGI= 14.75
     OF EACH YAGI= 22.5
                                                  VERTICAL MAST HEIGHT FROM
     LENGTH OF FRONT HALF                         PIVOT TO HIGHEST YAGI= 14.75
     OF EACH YAGI= 27
                                                  LAST DIRECTOR LENGTH= 8.7
     POLARIZATION= 0 DEGREES
                                                  REFLECTOR LENGTH= 9.75

                   MINIMUM       FRONT         FRONT       REAR        REAR
    ELEVATION      GROUND        CLEARANCE     TURNING   CLEARANCE     TURNING
    IN DEGREES     CLEARANCE     HEIGHT        RADIUS    HEIGHT        RADIUS
    ==========     =========     =========     =======   =========     =======
         0           17.25         17.25        33.51      46.75        30.35
         5           15.35         19.66        34.47      44.73        31.25
        10           13.57         22.16        35.27      42.62        32.03
        15           11.93         24.74        35.89      40.42        32.68
        20           10.44         27.37        36.32      38.17        33.18
        25            9.12         30.04        36.56      35.86        33.53
        30            7.98         32.73        36.61      33.52        33.71
        35            7.01         35.40        36.46      31.18        33.74
        40            6.24         38.06        36.11      28.84        33.60
        45            5.66         40.66        35.57      26.52        33.30
        50            5.28         43.20        34.86      24.25        32.85
        55            5.11         45.66        33.97      22.03        32.24
        60            5.14         48.01        32.93      19.89        31.50
        65            5.37         50.24        31.75      17.84        30.63
        70            5.81         52.33        30.45      15.90        29.66
        75            6.45         54.26        29.07      14.08        28.60
        80            7.28         56.03        27.63      12.40        27.48
        85            8.30         57.61        26.17      10.87        26.32
        90            9.50         59.00        24.73       9.50        25.15
 

THE TOWER

The center of the antenna is 32' high, which will allow me to reach the DE on each yagi with a tall stepladder if I need to for some reason.  This low antenna height means I will not need guys.  I am using 35' of Rohn 55G tower (with the bottom 5' anchored in concrete), with "tripod guys" of 1-7/8" OD diameter steel pipe (three pieces of 1.5" IPS Schedule 40 pipe x 21' long) solidly supporting the tower at the 17-1/2' height, and anchored in concrete 5.5' from each tower leg ( in 18" diameter holes, 3' deep).  Back fill from all the holes was piled up over the locations of the tripod guy pipe anchors, in order to increase the frost protection for the concrete bases. The bottom of each brace has rebar through it to attach it to the concrete, and the top end is fastened with a single large bolt, through the brace placed into the center of an 8" length of steel channel (which is bolted to each tower leg).  This is the same support system I have used for decades on my 16 yagi 2m EME array.

The lower 25' of  tower was assembled and single-handedly positioned in the hole with the assistance of a "skyhook", which was a pulley hanging on a 1" diameter nylon rope between two trees.  A come-along winch was used to pull the line going up through the pulley and down to the tower.  The tower was held plumb during pouring of the concrete through the use of come-along winches attached to polypropylene ropes attached to the top of the second section.  The top section can be seen sitting on a cable spool in the left of the photo below. The top section of tower was also added by me alone, using the "skyhook" lines.  This photo to the left shows me atop the complete tower, along with the hoisting ropes hanging overhead.

MAST AND MOUNT
	The mast is 1.5" IPS schedule 40 steel, with 1.5" steel tubing (.120" wall) welded inside it.  Atop the mast is a 31" long piece of 1-15/16" steel shaft welded to atop a 26" long piece of 6" wide steel channel,  braced by pieces of 3" wide steel channel, which extend down to the mast at a 45 degree angle.  2"x2" steel angle is welded on both sides of the shaft ends and top.  An 8"x8"x1/8" steel mounting plate is welded at the intersection of the braces and masts.  This mounting plate, along with a large piece of 6" wide aluminum channel, provide the attachment for the reversible DC gearmotor (which requires 200 VDC), right angle reduction gearbox, and idler sprockets  for the elevation drive chain.  These photos show the mount on the top of the tower, with the weather cover removed over the gearmotor, gearbox, drive/idler sprockets.  The photo at the left shows the elevation chain and antenna installed.

THE ROTATOR
	The rotator is a "medium sized" prop pitch motor.  This size motor just barely fits inside the tower, but cannot be installed through the assembled tower, and had to be mounted before the top section of tower was lowered over it and bolted in place.  In usual fashion, the prop pitch motor itself is bolted to the underside of the motor mounting plate.  A circular 1" high strap was welded to the rotor mounting plate to act as a barrier to prevent water from running down into the top of the motor.  You must protect the prop pitch from weather coming down inside the center of the motor. The unit used had a pipe fitting welded inside the drive sprocket, so a mast coupler was fabricated to attach to this fitting.  Inside the center of the bevel gear, a pipe adapter (of almost the same diameter as the inside of the bevel gear) had been welded, as you can see in the top photo to the right.  What I decided would be easiest to do, would be to attach my mount to that pipe adapter. As you can see in photos, I sawed with a hack saw, and filed to create two flat faces on the pipe flange so that I could place some 2" angle iron (painted brown) against those flats on each side.  I then secured these two pieces of angle iron on opposite flattened sides of the the pipe flange with (3) 3/8" diameter hardened bolts.   The two outer bolts go around the pipe flange.  The center bolt sits in a slot that I filed into the  two edges of the pipe flange; this center bolt resting in the slots, and the two opposite angle iron pieces tightened against the flattened faces of the pipe adapter, makes the flange assembly very well seated so it will NOT slip. Then, onto the top of this platform formed by the two brown angle iron pieces, I bolted the yellow mast clamp assembly, which is shown in photo to the left.  It was mounted onto the brown angle flanges so that the mast would be centered directly over the center of the prop pitch axis.  The yellow mast clamp was  made in two halves.  Each half has a piece of angle iron welded to a piece of iron plate.   Hose clamps were then used to tighten the clamp halves around the mast, and the clmap was then bolted down to the brown flanges. A conical weather shield was then fabricated from sheet metal and installed to cover the clamp and mounting plate.  As with the top tower section, the XYL (the official painter on the project) added the camoflauge paint job ;-)  I then sealed the top of it with rubber and plastic flanges from the hardware store, which permit the mast to rotate without binding on the weather shield. Although the mount as connected to the prop pitch cannot slip at all, the mast can slip inside the clamp.  Since the array is tied down when not in use, this feature assures that nothing in the mechanical coupling between the motor and the antenna will break.  The azimuth indication is obtained by coupling the rotation of the mast to the rotation of a potentiometer, using a timing belts, so only the actual position of the antenna is displayed.  If a less forgiving connection is desired, bolts could be installed through both the yellow clamps and the mast.

THE H FRAME
	I am using 30'-2" long aluminum vertical masts for the yagis (constructed by using a center 20' long piece of 2" IPS schedule 40 6061-T6 PIPE, with 2" 6061-T6 TUBING, .125" wall, telescoped 11" inside it to lengthen each mast end an additional 5'-1").  These vertical mast materials are very stout and are standard sizes, available through most any supplier of aluminum. The 2" tubing fits quite easily inside the Schedule 40 Pipe (with .070" clearance), and a  round shim for each of the four 12" long overlapping joints was made by forming a piece of 6.25"x14"x.020" aluminum to fit between the two members.  Each  joint was then bolted using two 3/8" hardened steel bolts and washers, oriented orthogonal to each other, and located 3" in from the end of each member (5" between the bolts). The masts are rigidly braced from the center to the midpoints with 1.5" aluminum tubing, and also guyed out to the antenna mounting locations with 1/8" steel guy wire (supported by a 4' long length of 2"x2"x.125" aluminum angle, and supported with 1"x1"x.0625 wall square aluminum tubing). One of the mast splicing bolts was used, in conjunction with a U-Bolt, to secure the mounting bracket for attaching the rigid bracing.  The masts are also braced sideways (at the antenna mounting bracket points) with Phyllistran cable.  The masts were mounted so that the center of gravity  the mast (when both upper and lower antennas with feedline attached) was in the center between the two horizontal towers. The horizontal framework  is a "double H frame" using a total of eight 8' sections of Heights AC14-100 aluminum tower (for two 32' long towers), spaced approximately 2' on center from each other. I used a single 48' long piece of this in my 16 yagi 2m array, and it worked fine, but twisted slightly.  The double H frame using that material, when properly guyed and supported, should be far superior, very lightweight, and more than adequate (this array only weighs half as much as my 2m array).  The length of the pair of assembled aluminum tower sections is actually 31'-8". The centers of the masts were each mounted 4.75" in from each end, with the antennas mounted on the outside of the masts.  This resulted in a horizontal stacking distance of 31',4-7/8" between antenna centers, which was very close to the targeted 31.5'. 4" aluminum channel is run diagonally above and below the two horizontal towers, onto which the 1-15/16" pillow block bearings from the elevation mount are bolted.  With this configuration of the double H frame and diagonal mounting, the center of gravity and pivot point will be exactly in the center of the array.  The entire properly balanced array is very easily elevated.  The photo at the left shows one of the pillow blocks bolted onto the central elevation shaft (as viewed from below after the antenna was installed on the mount). As is more apparent after the antenna is in the air (photo at lower right), the upper horizontal tower of the H frame assembly is mounted to the rear of the vertical aluminum support masts.  The lower tower is  mounted to the front of the vertical masts.  There are a pair of 5' long vertical aluminum masts similarly mounted between the two towers (immediately adjacent to the central support channels) used to support the 3/8"  heavy duty Phyllistran cables from the center of the array out to the ends of the horizontal members of the H frames (to prevent sagging of the H frame when the array is pointed at the horizon). The array is also braced out the front, from a center point with cables running out to the end of each bracket where the tower is attached to the main vertical masts (to keep the frame from sagging when the array is elevated.  As shown in the bottom photo at the left, the weighted precision 3 turn elevation potentiometer that is used for elevation readout is mounted in a pair of  metal film canisters.  Nylon cable ties hold the assembly inside the lower tower section next to the mount.

THE ANTENNAS
	Side-to-side bracing is provided to prevent the yagis from flopping over while elevated.  This is done through the use of 1/8" diameter Phyllistran cable, run just underneath the elements on each yagi, from the boom to a fiberglass cross member mounted near the center of each yagi.  The 8' long pieces of fiberglass poles were cut from broken pole vaulting crossbars (acquired through the courtesy from The University of Montana  track and field athletic department).  Plugs were put into the ends of the fiberglass poles so the U bolts holding the angle aluminum mounting brackets (for the Phyllistran cable turnbuckles) would not crack the fiberglass.  The plugs were standard nylon fittings for bed casters (there is quite a size selection of these at the local hardware store).  The fiberglass poles were lightly sanded, cleaned, and painted with three coats of white paint formulated for use on fiberglass (expensive, but also available from the local ACE hardware store). The paint used was "Marine Topside Polyurethane Enamel" by Valspar Corporation. The antennas came with dacron ropes to support the booms from a temporary 24" high central support post that needs to be clamped to mast after the antennas are mounted.  Galvanized 1/8" steel guy cable was substituted for the dacron line, and 25.75" long pieces of 1.25"x1.25"x.125" square aluminum tubing were bolted to the small guy support on each of the top two yagis.  This extension of square aluminum tubing (weighing 1.35 pounds) makes a tight fit extending 4.25" down inside (past the top three mounting clamps - see photo) the tops of the round 2" masts (1.75" ID), and provides adequately stiff, permanent support for the boom guys.

The balance point for the yagi of course depends on the particular feedline and support systems used.  For my yagis, the balance point was 20'-1/2" from the end of the N connector on the DE feed block.  By bending the feedline away from the mast at a point 30" from the center of the yagi, approximately 19' of LMR feedline weight  is added to each yagi.  The completed, trussed up yagi (less feedline) weighed in at approximately 49 pounds.
 

RF CABLING

The 50 ohm phasing lines were constructed from LMR600 cable.  The length of each line (with connectors installed) was 42'-7.5", which corresponds to 2.5 wavelengths at 50.150 MHz. This was confirmed by the short shown by sweeping the open feedlines at  25.075 MHz (where the length is 5 quarter wavelengths).  The weight of each completed phasing line with connectors was approximately 5.83 pounds, approximately 2.6 pounds of which was on each yagi.. With the extra electrical lengths inside the power divider and the two N chassis connectors, it was estimated that the actual overall length would be 2.5 wavelengths just under 50.100 MHz.  The 4 way 1/2 wavelength power divider was built from 1.25" OD square aluminum tubing with a .125" wall thickness.  The inner conductor was 15/32" OD round hobby brass tubing, creating a line impedance of 50 ohms.  Four .375" thick teflon spacers were evenly spaced at approximately 20" intervals (between the spacers or RF connections) through the inside of the square tubing, to keep the center conductor supported and centered.  The overall length of the aluminum tubing for the power divider was 10', with 57.75" (using a velocity factor of .98) from the center connector to the point where each pair of N connectors were connected.  (Next time I suggest using a velocity factor of .975, since the low end of the 6m band was just on the upper end of the bandwidth for a perfect match with the device built as described above). The connector for the main feedline was a 7/16 connector, to insure plenty of extra durability in terms of handling RF power.  The flexible coax around the rotator was LMR600 ultraflex, with 7/16 connectors on both ends.

Initially, 3/4" diameter 75 ohm CATV hardline was used for the feedline, which required an impedance transformation.  This  was acheived by constructing an "L/12" impedance transformer and mounting it at the top of the tower.  The transformer was designed around 1.25" OD (1" ID) square aluminum tubing (the outer conductor, sized to accomodate the connectors selected), with appropriately sized hobby tubing to construct the required 50 and 75 ohm sections.  The 75 ohm section (toward the 50 ohm antenna) was an 18.75" long piece of 5/16" diameter tubing; the 50 ohm section (toward the 75 ohm feedline) was an 18.75" long piece of 15/32" diameter tubing.  A single teflon spacer was used at the center where the tubing sizes changed.  At  the antenna end (to connect to the 50 ohm LMR 600 ultraflex), a 7/16 DIN female chassis connector was mounted to the square aluminum tubing of the transformer.  On the feedline end (to connect to the 3/4" CATV 75 ohm hardline), a male HN connector was mounted to the transformer.  The feedline to the shack is 200' long. In the shack, a traditional 1/4 wavelength long coaxial impedance transformer (of 61.5 ohms) was used to convert the 75 ohm CATV hardline back to 50 ohms. I constructed mine from 1" "Type M" copper tubing (OD=1.125", ID=1.055), with a 57.6" long piece of 3/8" OD hobby tubing as the center conductor. The 75 ohm CATV hardline was later replaced with 1-5/8 Heliax, and the impedance transformers at both ends of the feedline run were bypassed.

 

ARRAY ASSEMBLY
	The H Frame was first raised up and hung against the tower so the bottoms of the vertical masts just cleared the ground.  Even so, the sign company ladder truck could not reach up high enough to raise the antennas up to the tops of the top of the masts (around 32').  Therefore a crane was used to raise the antennas into place, and the bucket on the ladder truck was used to stand in and reach up to tighten the boom-to-mast clamps on the upper antennas. The lower antennas of course were within easy reach of the ground, and were mounted by hand.  The crane was then used to raise the complete array up to the top of the tower, where the H frame was bolted to the pillow block bearings on each end of the central support shaft of the elevation mount. Power supplies to run power the elevation and prop-pitch motors, along with relays to enable actuation from the house, were enclosed in a waterproof plastic box mounted on the bottom of the tower. A messenger cable (low to permit clearance of the yagi reflectors when the array is pointed up at the moon) carries the feedline (3/4", 75 ohm CATV hardline), AC power cord, and 8 conductor control cable from the house. The main tie-down ropes are 4 nylon lines attached to the center of the two vertical masts.  "S" hooks were also mounted in the bottoms of the masts so 4 more secondary  stabilization lines could be easily hooked on (by reaching up with a 15' long aluminum pole) when especially severe winds are anticipated.  From experience with my 2m EME array, it was found that tie downs attached only to the central horizontal boom do resist winds trying to rotate the antenna in azimuth, but the "rocking" effect from forces acting on the elevation mechanism can completely destroy the coupling there.  By also using tie-downs on the bottoms of the masts, such forces acting against the elevation mechanism can be limited.

PERFORMANCE

The VSWR of the antenna is nearly perfect between 50.100 and 5.200 MHz, showing only a couple of watts reflected with 1500 watts forward. During the first weekend of testing, echoes off the moon were present over half the time, and moonset signals from the 9G5AN DXpedition were detected on almost every transmission they made.  The charts at the top of the page indicate a beamwidth of about 10 degrees down to the 1 db point; although this may seem extraordinarily sharp to 6m operators used to being able to watch a very wide range of azimuths for incoming DX signals, it sure seems very comfortably broad for EME (at least compared to my 16 yagi array on 2m)!   Before I had the indicators and remote controls set up, I was limited to visual aiming, and I confirmed that the antenna seemed to work fine as long as I went out to move it every 30 to 45 minutes!  Even pointed through the trees at moonrise, echoes are quite good, and the antenna seems to perform noticably better than the single 70' long yagi at the 70' height, at least between 90 and 235 degrees azimuth.  Aiming the antenna at northern cold sky produces a noise of 2.5 dB, compared to a 50 ohm load here in the shack.

More testing will be necessary to confirm, but the initial comparisons show that the single yagi works better in directions (such as Japan), where the lower yagis are essentially at ground level within a dozen wavelengths, due to local topography.  In other directions in which the EME array has a clearer shot ( such as toward the Caribbean), the EME array is clearly a better performer.  A chart of the horizon provides a more complete picture of some of the VHF challenges of living in the northern Rocky Mountains!   The antenna is normally tied down at 110 degrees, which is aimed at Missouri (1xEs), Georgia (2xEs) and eastern Caribbean (F2).   It has been very helpful to use the single big yagi at 70' to locate DX stations, and then bring the GLEAP array around to the proper azimuth when a few extra dB is necessary.  Hope to contact you with it soon!