Wednesday, October 31, 2018

Lurking, Waiting and Scheduling: Circumventing the Pile Up

Now that the Ducie island VP6D DXpedition is wrapping up I thought it worthwhile to review a few of the less competitive techniques for working the rare ones. I used all of them for at least one band slot to work VP6D. I call them "non-competitive" since none involves pile up busting skills, antennas, power or propagation prognostications.

These are complementary to those competitive skills, many of which I've discussed numerous times on this blog, though admittedly not for a while, and mostly in the context of QRP and modest antennas. Just do a search to find those articles. I prefer competitive skills since they are a path to improving my operating and contesting, and it's lots of fun to mix it up in the pile ups with the world's best stations and operators.

Lurking

My timing and luck were not with on 40 meters CW in the early days of the DXpedition. They got on 40 meters close to their sunset, which is prime evening operating time in eastern North America. The pile ups were fierce. Despite having a 2-element yagi up 46 meters my 200 watts could not compete with the mass of kilowatt signals.

According to their band plan they transmit on 7.010 MHz and listen up. One evening I set the receiver on this frequency and programmed a transmit split of 1.02 kHz. I then unplugged the headphones and reduced the volume on the speakers low enough that the noise wouldn't be objectionable. With that playing in the background I concentrated on non-ham tasks that I could do in the shack; my shack and home office share a converted bedroom.

Some time later I distinctly heard a high speed "VP6D up" emanating from the speakers. By the time I stood up and took the two steps to the desk with the rig the message repeated. I punched the memory button to send my call, waited a second and punched it again. My eagerness caused me to double with him. I waited and a second later I heard "VE3VN 599". With a "599 TU" I had them logged on 40 meters CW.

There were a couple of other lurkers but I got through first. To be courteous I waited a few seconds for one more lurker to work VP6D before spotting them. Within a minute the pile up was roaring.

Lurking is boring but it can work to circumvent the pile up. Unless you have a excess amount of time on your hands I suggest you do as I did by scheduling some useful tasks you can do in the shack while you wait. Many times the DX does not show up and you'll be disappointed.

Waiting

If the DXpedition is long enough and you are very patient most pile ups can be entirely avoided by waiting. Eventually all pile ups thin or disappear as most everyone has worked the DX on a particular band and mode.

This is how I worked VP6D on 30 meters. I had tried a few times to break through the pile up early in DXpedition but for some reason I had no success on this band. Waiting was not entirely voluntary because for numerous reasons I often could not get on the air when conditions were favourable. Other times they were not active on 30 meters or were on FT8 rather than CW.

The day after I got home after operating CQ WW SSB from another stration I worked them first call when heard on 30 meters CW. They were calling CQ with few takers. Waiting can work.

The danger with waiting is that DXpeditions can be wrapped up earlier than planned due to weather and other logistical problems. If you wait you lose. Consider the recent TT8KO DXpedition that lasted one day before the authorities shut it down. Wait if you must but don't be surprised when you miss the rare one because you waited.

Scheduling

I was only able to reinstall my 160 meter antenna the day before CQ WW SSB. Since the DXpedition was slated to end a few days after the contest and I would be unavailable during the contest I chose to schedule time to find and work them on 160 meters. It is typical that DXpeditions start dismantling antennas a couple of days in advance of the DXpedition's end and they usually start with the low bands. Waiting to work them on 160 meters after the contest would be risky.

Lurking on 160 meters isn't quite as reliable as on 40 meters since the noise is high and I don't yet have a low noise receiving antenna to the southwest, towards VP6. Instead I lurked at one remove by monitoring their DXA page on the shack computer. I would glance over from time to time to see if the 160 meter activity button was lit. Since I did not have have on 30 meters either I was watching for that one as well.

When it was midnight I became impatient. I needed to rest up before the big contest. So I tuned to their frequency and connected to the spotting network. Lo! There they were. DXA was not updating properly and I missed their first appearance on the band. By this time they were slowing working Europeans, which is a very difficult path. After listening for a few minutes I could see this would continue for some time. I reluctantly shut down and went to bed. However I had not given up.

I decided to risk contest fatigue by getting up 15 minutes before sunrise. Their fantastically strong signal was peaking towards S9 as the sun crept up to the horizon. At 5 minutes before sunrise they were in the log despite a good sized pile up. Perhaps I had the advantage due to a sunrise enhancement, or perhaps I was a little bit lucky. I didn't stick around to monitor their signal strength as daylight flooded through the windows.

Scheduling can feel like work rather than a hobby. But if you want to bag the DX it is something you will need to do occasionally even if you have a big signal. Knowing the schedule of rare DX and planning your schedule on the air to match has a long tradition. In the decades before DXpeditioning was common and spotting networks were non-existent this was a tried and true technique for the serious DXers. If you'd like to read more about it I recommend W9KNI's venerable book The Complete DX'er.

Thursday, October 25, 2018

DXCC: The Remaining 40

When I worked the KH1 DXpedition on several bands earlier this year I briefly celebrated. According to my records it was DXCC country #300 worked since I returned to the air in 2013. As it turned out my records contained a few errors. Baker Island was #299. Eventually #300 came along with Ducie Island VP6D, a DXpedition that is ongoing as I write this. Of these 280 are confirmed via LoTW, which is the only way I QSL nowadays.

It is a milestone that was achieved more slowly than I expected. Consider that I reached 225 countries in two years using a maximum of 10 watts and modest antennas. Spotting networks and information dissemination about activity in rare locales makes running up the country count easier than ever.

Amusingly, the first 100 countries were easy enough despite running QRP and recruiting the aluminum eaves of my house in Ottawa as a random wire antenna. That was great fun. Doing it that way -- with a tiny station and low expectations -- played an important role in rekindling my passion for the hobby.

With DXCC the law of diminishing returns applies: as the country count increases each incremental country requires more work than the last. By the time 300 countries are reached the required effort is quite high and the progress slow.

My lifetime total is perhaps 315 or 320, reduced by deleted countries during the years I was inactive. But that number doesn't interest me. I wanted a fresh start after a 20 year hiatus from the hobby as a way to measure my progress. DXCC doesn't mean much to me and I have never applied for the award. I likely never will. It is the activity of DXing that I find appealing.

Maybe it's the excitement, maybe it's the competition and maybe it's the simply the magic of radio. Perhaps it's explained by catching the DXing bug when I was young, decades ago, when global communication was a rarity, except for those who got into ham radio. Youthful passions tend to linger.

Rather than dwell on the 300 country milestone what I'd like to do in this article is discuss the ones I have yet to work. With 340 countries on the current DXCC list that there are 40 countries left. By the end of this article it should be evident why working them all, or achieving the lesser but still heady level of DXCC Honour Roll, is so incredibly difficult. Most DXers who have done it have taken a lifetime to get there.

I'll group countries together where it makes sense to do so to keep this article from becoming repetitive and too long. Apart from the grouping, countries will be ordered by prefix.

3D2C Conway: I vaguely recall some DXpedition activity but have yet to work this one. Perhaps all this one will take is paying more attention. It shouldn't be difficult to work.

3Y Bouvet and 3Y Peter I: These Antarctic islands are remote and battered by severe sea and weather conditions. There was one recent DXpedition attempt to Bouvet island that failed to land due to the weather. I have both countries confirmed back in the 1980s. All one can do is wait for intrepid and well-financed DXpeditioners to make the attempt.

4U1UN United Nations: Only this summer has the station been rebuilt in compliance with the severe security regulations at the UN complex in New York. I have heard them weakly on 20 meters but not workable. New York is too close for reliable communication on the higher HF bands except by scatter and high power. I would be surprised if I don't have this one in the log by 2019.

7O Yemen, YK Syria: One word says it all -- war. Very sad and tragic, and these won't be resolved soon. Going back a few decades I have lots of cards from 7O and a few from YK. But for now they are not workable.

BS7H Scarborough Reef: For one elderly ham who lives nearby this is the only country he needs to have them all. Another longtime DXer of my acquaintance doesn't believe it'll ever be active again. I have no opinion. Even if it were active it would be very difficult to work during a solar minimum from this part of the world.

BV9P Pratas: Like Scarborough Reef I know little about this one. Both have flown under my radar since they appear to have become entities during my 20 years away from the hobby.

CE0X San Felix: I've worked CE0Y and CE0Z, yet failed to work this one. If there's been recent activity I've missed it since I keep confusing these Chilean islands, forgetting which is the one I haven't worked. Like many other entities I worked this one decades ago. I'll just have to pay closer attention to catch the next visitor to San Felix island.

EZ Turkmenistan: This former Soviet republic was uncommon but not rare back in the days of the USSR. Since then Turkmenistan has been jinxed with totalitarian rule. Although that is not necessarily incompatible with amateur radio, it is not the case here. There have been signs that the political situation may be improving. We can only wait.

FR/g Glorioso: Another rare island waiting for a group to launch a DXpedition. Again, all I can do is wait.

FR/t Tromelin: There was an excellent DXpedition a few years ago, one which I completely failed to work. It happens. At the time my antenna and power situation made it difficult. I will just have to wait for the next DXpedition, and that may be years away.

FT5W Crozet and FT5X Kerguelen: Like the US, France has become fiercely protective of the environment on isolated islands with unique flora and fauna. French hams have been trying and failing to gain permission to land on these islands. There is (was?) a ham on Crozet Island although I am only aware that he has worked /MM when asea, and I did work him there. But that doesn't count for DXCC.

HK0M Malpelo: No recent DXpeditions. As with some other environmentally sensitive islands permission to land and operate is not routine. I am unfamiliar with how restrictive Colombia is with respect to Malpelo island.

JD1 Minami Torashima: As with a number of entities on this list, there is occasional activity from this one. Unfortunately the path is not the easiest during a solar minimum and I have yet to catch up with them. I need propagation and activity to coincide, and to pay enough attention to notice when it does occur.

JX Jan Mayen: This one ought to be easy and there has been some activity. Maybe I just need to pay attention and try harder.

KH3 Johnston, KH4 Midway, KH7K Kure, KP5 Desecheo: Permission for Kure island was recently denied (again) and the others are similarly difficult. When DXpeditions are permitted I expect these to be relatively easy to work. Back in the 1970s and 1980s I worked Johnston and Midway islands many times, and Kure island at least once. I have never worked Desecheo island.

KH8 Swains: The propagation path is fairly easy, so all I need right now is a DXpedition. I can't recall whether there has been any recent activity from this Pacific island.

KH9 Wake: The K9W DXpedition was a good one for working what has become quite a rare DXCC entity. At the time I was running QRP and little antennas. The best I did was "VE3?" on 17 meters CW. Although there is a ham on Wake Island his activity level hasn't made it easy. In fact I've never heard him. It's likely I'll have to wait a number of years until another DXpedition is permitted.

P2 Papua New Guinea, XU Cambodia, XZ Myanmar VR2 Hong Kong: Southeast Asia is a difficult path from this part of the world. During the most recent solar cycle peak my station was tiny. Now that I have bigger antennas suitable conditions are elusive. I have heard activity from all of these countries, including a few big DXpeditions, but have not succeeded in getting through. They'll come to me eventually if I pay attention and make the effort.

P5 North Korea: The short bursts of activity that pop up at rare intervals have satisfied few DXers. Some have been doggedly persistent while most that have worked North Korea have simply been lucky. A difficult propagation path and little activity makes this a tough one for many, including me.

PY0S St. Peter & St. Paul Rocks: Another one waiting for a DXpedition. I don't recall there being one over the past 5 years.

ST Sudan: For a brief period after independence Z8 South Sudan was not difficult to work. However Sudan itself has been and continues to be hostile to amateur radio. This one will likely require a change in the political winds. Years ago I worked Sudan a few times.

SV/a: Mount Athos: I've heard more pirates than legitimate activity from Mount Athos. At the moment I am unaware of any activity at all. The one resident monk with a license is inactive and visiting hams have not been welcome.

T31 Central Kiribati: This one takes an DXpedition, but otherwise should not be difficult. Patience is a virtue.

VK0/m: Macquarie: No DXpeditions in the past few years and little ham activity from the occasional Australian stationed on the island. Despite the great distance this area of the world is not overly challenging to reach, even during a solar cycle minimum.

VP8/o: South Orkney:You would think that since I've worked the other and far more rare VP8 Antarctic islands such as South Sandwich and South Georgia this one would be easy. Sadly, no. To be honest I haven't paid much attention to which island is which and so I've missed a few opportunities.

VU4 Andaman & Nicobar: There have several DXpeditions to this entity. Although I've worked its neighbour VU7 numerous times, and once with QRP, this one has eluded me. I have tried though perhaps not as much as I ought to since I keep thinking there will be a next time. The path over the pole will be a difficult one for the next two years.

YI Iraq: Although there are hams currently active it has proved elusive. From what I can tell their stations are modest and difficult to work from here. A couple that have been heard were very weak and working Europeans.

YV0 Aves: There have been no DXpeditions recently, and this one is notable in that it requires cooperation of the Venezuelan navy. The most recent attempt was ultimately cancelled due to...reasons.

ZL8 Kermadec, ZL9 New Zealand Subantarctic islands: These entities see occasional activity but have not grabbed my attention sufficiently to make the effort to chase them.

ZS8 Prince Edward & Marion: These days about the only activity on this protected island group is when a ham is stationed on the island. Unfortunately activity over the past 5 years has been fleeting and I haven't made a great effort to track them down.

If you took the time to read this far it should be evident that I am not at all fanatical about DXing. I love it, yes, though not nearly to the degree of many others. Notice that for more than a few of the countries I've yet to work I cannot recall whether there has been any activity since 2013 when I returned to the hobby and chose to reset my DXCC count to zero.

Yet there are indeed quite a few of these DXCC entities that have been either completely or almost completely absent from the airwaves since 2013 and therefore impossible to work. For these it will take patience for the political situation to change, permission from the relevant government authorities or money and persistence to mount DXpeditions. I am in no great rush so I can wait.

For me chasing DX is a fun challenge though not an obsessive compulsion. With a little more dedication and persistence I can do better at working the less active and more distant entities that are in fact active from time to time. Whether I'll ever do so I just don't know. Towers and antennas, and of course contests, are more interesting to me.

Amateur radio is a big hobby with diverse activities to meet many tastes. DXing and the DXCC pursuit can be a small or large focus for each of us, and our interests change with the years. There are hams who will only turn on the rig when one of the few remaining entities they have yet to work comes on the air. That's an attitude I'll never understand. I just don't see the point.

Sunday, October 21, 2018

Deja Vu: Planting Another Big Tower

I've been busy getting ready for the planting of my second big LR20 tower, among other urgent tasks to be complete this fall. The blog has not gotten my full attention so the pace of articles has slowed. This will continue for a while.

Earlier this week the tower was planted. Now I am waiting on the concrete to cure and the machine shop to fabricate a few needed items. Then I can proceed. With luck the tower will be up this fall. Luck includes good weather, friends to come out to help and no serious mishaps. It's doable.

This is my last big tower -- two is enough for me to do what I want without incurring excessive maintenance during my golden years. For those who have been following along, this tower was in my original site plan for this QTH, and is located within meters of that plan. The tower will be a nominal 140', and actually ~133', or 40 to 41 meters, taking into account section overlap for splicing and base pillar height. This is an excellent height for DX work since it is 1λ on 40 meters and 2λ on 20 meters, and not quite 3λ at the top of the mast for 15 meters.

Unlike the disaster that was the planting of the first LR20 tower this one took only 7 hours. That's remarkable. It is thanks to the presence of an old hand leading his large crew of four, lots of planning and preparation beforehand and a large measure of good luck.

While it might not seem necessary with this crowd in attendance I was very busy on site working and supervising work. I got a good deal on the crew since for most of them it was their first time planting a guyed tower and my tower was a training exercise. Everyone wins. But the inevitable mistakes had to be noticed and pointed out, then corrected. Fortunately I have lots of management experience.

In this article I'll mainly focus on why the work went so well in comparison to the first big tower, which was a months long expensive headache. This should be more interesting to read about than simple repetition. There are relatively few pictures because I was less motivated to take them. The young guys on the crew took many more to, I suppose, remember what was for them a novel event.


First, the updated site plan. The changes are the approximate final positions of the new tower and the 80 meter array in the north field. Placing the new tower to meet my many requirements was difficult. The major impediment was avoiding the trees and rock wall surrounding the yard while achieving an ideal orientation of one tower face aligned with Europe for side mount yagis. A similar orientation on a line through the original big tower allows experimentation with wire antennas for the low bands.

I got pretty close to my objectives. A lesser objective was to minimize the transmission line run from the tower to the Trylon tower within the yard (yellow ellipse), which is the location of the antenna switch. The distance is similar to that for the other big tower, depending on how I ultimately choose to route the cables, both underground and above ground. There are several approaches to do this while avoiding tree roots and other obstacles. I have until next year to decide.

Surveying was similar to what I did for the first guyed tower. I'll refer you to that article rather than repeat myself.

Pretzel machine


Once again I opted to do my own rebar cages for the anchor and base. This time I had no outside help at all. I recruited material on hand to make the bending jigs and did a few tests to ensure each bend was exactly where I wanted them. I'm getting much better at this compared to the first time.

I again used an LR20 tower section and side mount bracket to build the jig for bending the rebar ties. I call it my pretzel machine. The design is an improvement over what I used before. It worked very well. The main challenge was determining how far the rebar would advance to the left as the hickey was rotated counter-clockwise on the leftmost stud. With that known it was easy to achieve consistent and accurate bends.

The circular stirrups were more of a challenge. The trick was to exploit the natural bend radius the jib imparted and repeatedly advance the rebar when the correct arc for a 12" circle was achieved. A circular template was used to check progress and to correct errors. The resulting stirrups aren't pretty but they're perfectly good.

When all was done I had 12 square pretzels, 18" on a side, and 5 circular pretzels of 12" diameter. Straight lengths of 20M rebar were cut and then combined with the ties and stirrups to form the 3 anchor cages and base pillar cage. The base platform grid is comprised of straight lengths of 20M rebar.

To bend the 20M bars for the base pillar (equivalent to US #6 bars) I needed a heftier jig. As in the past I recruited a suitable tree and steel pipes. After looking at several dozen trees (I have thousands on my property) I found a birch that was perfect. With this jig the 8 bars for the pillar were given a 90° hook on the end for joining to the base platform grid.


Notice the construction of the completed anchor cage. There are extra bars on the top and front that assist with distributing the tension of 4 guys across the load bearing faces of the anchor. Without them there is a risk that the concrete could split in the middle under severe wind load.

The cage is 18" × 18" × 90". This leaves a minimum 3" of concrete on all sides for the 2' × 2' × 8' reinforced concrete anchors. This is standard procedure to ensure long term corrosion protection of the rebar.

Excavation

I had a chance to chat with the backhoe operator before the tower crew arrived. He looked and sounded familiar. I soon realized that he was the plumber who came to my rescue when the house pressure system failed soon after I moved in. Turns out he had been laid off and returned to his earlier job as a backhoe operator.

While this is an interesting story what was more important is that I admired the skill and meticulousness he demonstrated on the plumbing job. He brought the same attributes to this job. There are ways to use a backhoe that can greatly improve the excavations, if the operator cares enough and has the requisite skill. This was my lucky day.

Knowing that the anchor holes were to be 2' × 2' × 8' he brought a 2' bucket on the front. He was able to made these 3 excavations between 24" and 27" wide. That's excellent. This was so accurate that I had to get the crew to use their shovel to clear 3" around all sides of the rebar cages, including squaring the bottom corners. There was little margin for error with this quality of backhoe work.


The advantage is that the quantity of concrete was kept to a minimum, which saved me a few hundred dollars. While it is possible to build casings for the anchors it is more costly in labour and materials, and if the casing is to be removed it is necessary to bring back the backhoe a day or two later to backfill the holes. Using what they call "mud holes" is faster and cheaper. But to do it you need the right soil and soil conditions and a good backhoe operator. On this day everything went right. Casing was only used for the base.

Concrete

Readers may remember the fiasco with concrete delivery for the first big tower. The truck got stuck in the boggy November hay field. It and the concrete had to be rescued by backhoe. I had no intention of letting that happen again.

When the driver arrived we did a walkabout to decide what to do. The backhoe has a big bucket (front end loader) for concrete delivery, if necessary.

The driver declared the ground just fine and drove onto the hay field. His judgment was good. Despite the 5 m³ of concrete onboard there were only light indentations left in the field.

Delivering the concrete direct to the excavation by chute is by far the superior method of delivery. It avoiding shifting of the steel when dropping concrete from the bucket and leaving time to push concrete into all the nooks and crannies to minimize air pockets. It is also much faster.

A handheld electric concrete vibrator made it easy to encourage the concrete to flow into all those spaces and level the surface. There is also less risk of shifting the rebar when shovels are used to pump the concrete.

Alignment

The crew did not spend time checking my surveying. There was no need. The crew leader aligned the anchor rods the old school way, with a string. In his experience it's the fastest and most reliable way of doing it. Too often he's found that transits are out of calibration or improperly used, but a string never lies.

When I did my surveying I measured the levelness of the ground. In one case I adjusted the distance to the anchor to account for the 18" drop in that direction. This ensures that the angle of all guys of each set come down from the tower at the same angle. My low tech method was to use a long level on a platform at the future position of the base pillar. Point it to a marked stake at the anchor site, ensure the level is level, then sight along it to measure the elevation difference.

More modern tools were used to set the 38° angle of the anchor rods. Every member of the crew would pull out their smart phones, lay them on rods and use inclinometer apps to adjust the angle. The angle was checked during and after the concrete pour.

The rule of thumb I was taught was that there are 30 minutes after the concrete is poured to move and align the anchor rods. Allow 2 hours before back filling over the exposed concrete, which allows it to firm up.

Back fill

When all but the back filling was complete the crew left, leaving me and the backhoe operator. While we waited for the concrete to firm up I had him move dirt around. This was the excess soil and subsoil from the excavations displaced by the concrete.

Some was left in small piles at the four excavations. I will use those to level the surface once the soil settles. That will take until spring. Since I forgot to leave extra soil at the excavations for the first big tower I also had him leave some at those four locations. I had done some wheelbarrow work in the spring but gave that up since it was tedious manual labour.

As the clock advanced we proceeded to back fill the three anchors. These were done in the same order they were filled with concrete and first poked to check firmness. He started slow with granular soil until the concrete was covered to 1' depth and the space under the anchor rods was filled. For the latter I used a shovel. This helped to prevent the weight of the back fill from pushing the rods downward to a lower than intended angle.

Over the next two days I spent a few hours raking and shovelling the extra soil. Large stones were removed by wheelbarrow. Last year I chose a spot within the bush to place the displaced soil and stones from the excavations.

Pier pin

The pier pin for the base section was embedded at the centre of the pillar soon after the concrete was poured. The pin in this case is a standard pipe with an OD ~1.3". The opening on the base section is 1.625". I may add a shim to fill the gap, however that is not really necessary.

One day later I filled the pipe with no-shrink grout and cleaned the pipe of stray grout and rust. It will be painted before the load bearing plate is slipped over the pin and grouted.


As you can see the 10' ground rod is already in. It was placed at a corner of the excavation then hammered down into the undisturbed soil. They crew would normally drive it below ground level and fill around it with gravel for future access. I prefer the ground rod to poke up a few inches. The difference in lightning protection is not large. I can drive it down later should I change my mind.

With that we're done! All that's left is final preparation of the tower sections and building upward. That stage of construction is scheduled after the CQ WW SSB contest. It should go faster than the first tower since the rigging can be reused and the lifting process has become routine. Hopefully I can get the same hams out to serve as ground crew. It's a race against winter.

Tuesday, October 9, 2018

3-element 40 Meter Yagi on a 40' (12 m) Boom

My recent article about long boom yagis reminded me of an omission in my blogging practice. The 15 meter and 20 meter yagis had references to earlier articles or elsewhere so that readers could learn the details of design and performance. I did not do the same for the 40 meter yagis on a 40' (12 meter) boom.

The designs were done some time ago, which is probably why I forgot about them until then. Not long before I wrote that article I sent the EZNEC files to a ham in Europe who was inquiring whether I had a design handy for a 12 meter boom. But in that article I only mentioned them in passing. It is worth writing about them because they do have interesting performance characteristics in comparison to the usual 48' (14.7 meter) boom length commonly used for 3-element 40 meter yagis, including those with a fourth element (coupled resonator, OWA design).

Before diving in it is worthwhile to briefly recap a few points about 3-element yagi design.
  • Optimizing for gain only modestly reduces F/B but will significantly reduce the SWR bandwidth. The latter is due to the low radiation resistance associated with maximum gain.
  • Frequency of maximum gain is above the usable bandwidth, while F/B tends to peak towards the low end of the usable bandwidth. In practical terms, a 3-element yagi typically has maximum gain on SSB and maximum F/B on CW.
  • Designs which maximize SWR bandwidth sacrifice up to 1 db of gain. However F/B can remain good.
  • Gain is optimized by bringing the self resonance of the director and reflector closer together. The spread in percentage is a convenient metric, one that I originally learned about in the venerable Yagi Antenna Design book by W2PV. Tuning of the driven element affects the match (and matching network) while having negligible effect on gain, F/B and SWR bandwidth.
  • Best performance typically has the driven element offset toward the reflector end of the antenna a small amount.
The addition of a fourth element as the coupled resonator in a high bandwidth OWA design increases the mechanical complexity and load, with the advantages of low SWR from 7.0 to 7.3 MHz and little to no sacrifice of gain or F/B. The worth of the trade off is an individual decision. I include 3-element and 4-element designs for both 15 meter and 12 meter boom lengths in this article.

Overview of the yagis to be compared

The 3-element yagi on a 48' boom in the ARRL Antenna Book is a useful baseline for comparison because of its balance between gain and SWR bandwidth. If that antenna interests you, look there because I won't get into the details here. To summarize, its 2:1 SWR bandwidth is ~200 kHz, gain rises from ~7.8 dbi at 7.0 MHz to ~8.7 at 7.3 MHz, and F/R is 21 db at 7.0 MHz, rises to 27 db at 7.1 MHz then falls to 12 db at 7.3 MHz.

The tuning spread of the parasitic elements is 15%, or ±7.4% of the design midpoint. I will use the latter convention in this article. It is not half of 15% since the exponential mean must be used. You get this with the square root of the ratio between the reflector and director lengths. This is good enough for our purposes though not absolutely precise since the tubing taper schedule has an effect.

For comparison, the 3-element yagi I discussed in an earlier article has  a tuning spread of 6.4% to increase gain by ~0.5 db. F/B continues to be very good, although 2:1 SWR bandwidth is a little less than 200 kHz. With a coupled resonator added the SWR bandwidth is low from 7.0 MHz to 7.3 MHz.

Because reducing the boom length to 40' also reduces achievable gain I further tightened element tuning to 5.3%. This recovers much of the gain while sacrificing SWR bandwidth, as we'll see. The addition of a coupled resonator helps with the latter.


As you likely noticed I have reverted to English units for these yagis. I do this because the tubes and pipes I use are measured in these units and so I model the yagis accordingly. The figures on the left are the distances along the boom the elements are placed and the lengths of each half element. The addition of a couple resonator only affects the length of the driven element. Its length without the coupled resonator is in any case dependent on the matching network: gamma, beta, L-network, etc.

Half element tubing schedule is: 144" of 1.9" pipe; 96" of 1.5" tube; 66" of 1" tube; 66" of ⅞" tube and a variable length of ¾" tube. This is not likely how I would construct the elements. It was an experiment to determine how I could put my stock of aluminum pipe to good use. The idea was spurred by reading Dave Leeson's Physical Design of Yagi Antennas -- it's out of print but the author kindly sent me a copy. For other tapering schedules it is necessary to carefully scale the elements.

For interest I show the 4-element yagi element currents at a higher frequency to highlight one aspect of how the coupled resonator broadens the SWR bandwidth. With the coupled resonator tuned to a higher frequency its current dominates that of the driven element at the high end of the band. In a sense it becomes the driven element despite not being attached to the transmission line. Since it is offset toward the front of the yagi the gain and F/B are slightly affected at higher frequencies.

Performance comparison

In the chart I used the same colour lines for both curves of each antenna -- gain and F/B -- to make it easy to compare yagis. Gain and F/B ought to be easy to distinguish!

I chopped the top of the chart to further improve readability since F/B figures well above 30 db are very difficult to achieve in real antennas due to the precise cancellation of element fields required. It is sufficient to state that F/B is excellent over a portion of the band.

Gains of the 40' boom yagis are about 7.8 dbi at 7.0 MHz and 8.4 dbi at 7.3 MHz. Gains of their 48' boom sisters is a fairly consistent 0.7 db higher. That isn't a large sacrifice to make considering the substantial construction differences for equal robustness.

F/B is significantly better for the 40' boom yagis even though I hadn't planned for that. It may be that further tuning of the 48' boom yagis would erase some or most of the difference. Certainly the yagi in the ARRL Antenna Book does a little better than these gain optimized yagis.

With a fixed matching network the 2:1 SWR bandwidth of the 3-element yagi on a 40' boom is 180 kHz. That's good but not great. With a switchable L-network at the feed point it is possible to achieve a low SWR up to 7.25 MHz. Getting all the way up to 7.3 MHz would require one more switch position. I wouldn't bother though some might want it. At least in the Americas where we have a 300 kHz wide 40 meter band.


The coupled resonator version of the 40' boom yagi can achieve better than 2:1 SWR across the entire band. However I was unsuccessful bringing down the SWR lower than 1.4. Further tuning of the driven element and coupled resonator might do it, but there are no guarantees. My difficulty was exacerbated by NEC2 which has some difficulty correctly modelling the impedance of closely spaced elements even when great care is taken with element segmentation, as I did.
Perhaps in practice it'll do better than the model and can be tweaked once on the tower -- the driven element and coupled resonator are within easy reach. Otherwise it may be necessary to loosen the tuning of the parasitic elements to improve the SWR, at the expense of some gain. I did not explore this path for the present study.

Further work

Despite my best intentions I will not construct and raise a full size 40 meter yagi this year. Maybe not even next year. There is therefore time to play with the models and explore further options. For example, to give up some gain to recover SWR bandwidth. No matter what I do there will be have to be compromises made.

I hope the information in this article spurs a few ideas of your own. These antennas are far more challenging than the 40 meters wire yagis that have been so popular among the readers of this blog.