Buying a Used Yagi: Hams Ripping Off Hams

I described my research into choosing a tri-band yagi in a series of articles last year (2014). The antenna I eventually purchased was not the first one I looked at. Originally I preferred to locate a TH3 since that seemed ideal to my needs and tower capacity. The Explorer 14 I bought has the same boom length and more efficiency (4 less traps) but more wind load.

While shopping for a TH3 I encountered something unfortunate. There was a transaction involving an antenna where, in my opinion, one ham ripped off another. That story is the subject of this article. No calls or names are mentioned, and indeed I don't know the identities of certain parties. While stories like this are atypical in the ham community, fraud and frayed relationships sometimes just seem to happen when money is involved.

Inspection

I found a TH3 on the local used market and contacted the seller. He explained he didn't know much about antennas, especially yagis, and would willing to negotiate the already reasonable asking price. I drove over to have a look. Problems were already evident from the ground, which is not a good sign. For example, one element tip dipped downward from the outermost trap.

I climbed the tower for a closer look. More problems appeared. He wanted the antenna down and since he was a decent fellow I gladly performed that small task for him. It was already becoming clear to both of us that I would not buy the antenna.

With the antenna on the ground I began to disassemble it. It quickly turned into something like a forensic analysis at a crime scene. What I found shocked both of us. He soon realized he'd been had by the ham who sold it to him.

How we got here

Someone he connected with in a local club offered to sell him a yagi when he was looking for something small and inexpensive to work some DX on the higher HF bands. That individual was friendly and willing to help put the antenna on his tower. A deal was negotiated.

The seller delivered the antenna and assembled it. In retrospect the buyer realized he should have been suspicious since the antenna looked odd and the seller gave him no opportunity to consult the manual that he'd downloaded in advance. He quite sensibly wanted to make measurements and just to have a closer inspection of his first HF yagi. The seller explained he was pressed for time and wanted to be quick with the job. So up the antenna went and coax attached.

Signals were heard and the transmitter seemed happy. Money changed hands. But it was soon clear that not all was as it should be. The SWR was quite high in many places. When contacted the seller told him that was nothing unusual and just to use the rig's ATU to match.

He had no basis for comparison and so could not check gain and I don't believe he checked the F/B. He only knew that when he turned the antenna the desired signal peaked and he was usually able to get through. Of course even a dipole shows directivity, and a yagi is like a dipole in its basic azimuth behaviour: broadside gain and a null off the sides.

This is a real story and a real antenna so I am able to provide physical evidence. You'll soon see why I was outraged by what had transpired.

Busted trap

The cause of the dipping element tip was discovered when the mass of tape holding the tube to the 15 meter trap was removed and the trap opened.

Pretty, eh? The plastic coil form is severed and the tab connecting the trap shell (capacitor) to the element was sheared off. The stand-off toroidal insulators are heavily scored. A violent impact would account for this degree of damage.

Trap covers

All the trap end covers are taped. The few I uncovered were severely damaged by UV radiation. This is a case of the shutting the barn door after the horse has bolted. Tape (the necessity of which is debatable) should have gone on earlier to reduce sun exposure.

The decay of one cover was so bad that the tape was actually holding the cover together. In another case removal of the tape revealed an ill-fitting plastic plumbing cap rather than a Hy-Gain part. Hy-Gain sells replacement kits which are reasonably priced. There really is no excuse for what I found.

Parts substitution

One tube joint could not be taken apart. There wasn't even a clamp at the joint. A closer look (see upper right of above photo) showed no compression slots in the larger tube and that the tube was not square cut. This is a press fit that was likely seized by corrosion or an oversize wall in the larger tube.

I can only presume the original tube was damaged beyond repair and replaced by another tube that was not properly selected or prepared.

Hy-Gain multi-band yagis have few trap varieties. This is good design and good business. All the 10 meter traps are identical. The 15 meter traps are different for the driven and parasitic elements. This allows kilowatt power without heating the driven element traps to destruction. In a conventional yagi the driven element current is higher than the parasitic elements. These traps are wound with copper wire to reduce resistance and therefore I²R loss. All the other traps are wound with cheaper aluminum wire.

The difference can only be discerned by the part number printed on the affixed label (too faded to read in the present case) or by peering through the drip holes to inspect the wire colour. The wire in the trap wound with copper should be brownish instead of silvery gray.

Element-to-boom clamps


I found two cases of substitution of proper Hy-Gain parts. Have a look at the adjacent picture.

On the left is a clamp made from half of the correct part and half that homemade from sheet aluminum.

First, the alloy is unknown and therefore of uncertain strength. Second, it is improperly shaped. The tubes inserted in this clamp had their ends severely crushed.

Why someone go through so much misspent effort when a replacement part is readily available and inexpensive is mystifying. I can only imagine that it was someone with more time on their hands than good sense.

On the right is the clamp for the driven element. This clamp must be larger to accommodate the plastic inserts which insulate the driven element from the boom. This is needed for the beta match feed. Except I found that the wrong clamp was used. To get a compression fit to the boom the clamp had to be tightened so much that the plastic inserts were crushed and split (not visible in the photo). Further, the anti-rotation set screws were replaced with longer hardware to bridge the gap.

Boom

The TH3 boom is a little over 4 meters (14') long. It comes in 2 identical halves that are joined at the centre with a clamp.

The ends of the two halves (off the left side of the photo) are aligned. It is readily apparent their lengths are unequal. The longer is the correct length for this antenna. The other must have come from another antenna since it shows no signs of having been cut. My guess is that it is from a TH3jr (12' boom) since the length is correct for that antenna.It would also explain the lack of copper-wound 15 meter traps in the driven element. That is, this antenna may be a mix of parts from multiple antennas.

Not only is the resulting antenna physically unbalanced the tuning of the elements will be sub-optimal for the altered inter-element spacing. The result will be poorer performance and SWR behaviour. Even if the difference is small it makes no sense to put up a 3-element yagi that and not get the performance you paid for.

Notice the deep indentations in the longer boom half. It's worse than it appears in the photo. Similar damage is present at the position of every boom-to-element clamp. The set screws (2 per clamp, to help prevent element rotation) were overtightened, in a few cases piercing the boom wall. There are more dimples than set screws which tells me that this antenna was improperly assembled at least twice, with the boom partially rotated the second (or third) time.

Although this is not a large antenna there is no excuse for this abuse, which will reduce survivability in severe weather.

Hardware

On a somewhat positive note most of the hardware -- hose clamps, bolts, washers and nuts -- were stainless steel. However the sizes were often incorrect. This is the best I can do to say something positive.

Prescription

There is great fellowship among radio amateurs. We help each other out with antenna-raising parties, sharing expertise and software, advice and training, and even simply offering pointers to help out the novices among us. Elmers -- those hams who mentor others before and after they first join our ranks -- are rightly venerated.

Yet hams are only human. It is a mistake to imagine that we are all cut from better cloth. Just as in the general population we have our misfits, anti-social miscreants and worse. The number in the latter group is small but can have devastating impact when they prey among the novices in our hobby or those who are overly trusting of fellow hams.

We like to believe we are good judges of character. It is not so easy, though it can become easier as we grow older and wiser, often by the "benefit" of bad experiences. Getting ripped off not only cost time and money it can also cause acute embarrassment. I have seen disputes come close to fisticuffs in flea markets. I have looked sellers in the eye and asked hard questions about an item they were selling, and then watched them squirm. Even the dishonest have consciences and you may see it on their faces.

Too often we keep those incidents quiet, fearing that other will think us foolish. That can be a mistake, one that the criminal class counts on for the continuation of their careers. Perhaps my story can, in a small way, shed some light where it's often absent.

In any transaction where you lack expertise about a product or don't know what question to ask, bring a knowledgable friend along. It won't hurt and it can help avert a bad experience. When we lean on each other we are all stronger. Pass along the same favour when you have a chance. Just don't go overboard and become suspicious of everyone; happily the bad apples are the exception not the rule.

Radial Topology Options: 2-element Parasitic Vertical Array for 80

In this article I want to look further at radial topology of vertical arrays for the low bands. In an earlier article I looked at the impacts of base height and radial arrangement of an above-ground mount for a 2-element vertical (ground plane) array on 40 meters, one that had 4 radials per element. That antenna performed well in the model, although I did note the potential pitfalls regarding the precise way in which the radials interlace, parasite tuning and the expectations for ground loss that may differ in the real world.

Although doing that model on 40 meters makes the antenna more mechanically friendly there are better alternatives for gain, directivity and low radiation angle on that band, such as a small yagi that is 20 or more meters high. So it was perhaps more of a modelling convenience than a desirable antenna project. On the other hand, hams without large towers and a disinterest in large yagis have used vertical arrays, including 4-squares, on 40 to good effect.

Dropping down to 80 meters and the situation is markedly different. Even among the big guns of the world a yagi is rare and raised vertical arrays are, while not rare, not common either. Ground-mounted vertical arrays are more typical, with the 4-square being pretty much the big-gun standard.

Standard element

The vertical element I will use here is a ground-mounted monopole tuned to resonance (X = 0) at 3.6 MHz over EZNEC medium ground (0.005, 13):

Top view of vertical; wire 1 is the vertical monopole

• Monopole is 20.2 meters long and 50 mm diameter aluminum, fed at the bottommost wire segment. The chosen diameter is an average value assuming a tapering schedule for telescoping aluminum tubes.
• Radials are 20 meters long, 16 AWG aluminum wire. This is commonly available and economical electric fence wire.
• There are 16 radials. This quantity is a compromise among computation time, ground loss reduction and NEC2 model reliability.
• Mounted at a height of 10 cm, which is just above the 0.001λ minimum W7EL recommends for reliable NEC2 emulation of radials lying on the ground.
• Use of a common mode choke at the feed point and other places is assumed in order to remove the need to model common mode current on the outside of the transmission line. Burying the coax does not eliminate this requirement.

At resonance the antenna has a feed point resistance of 37.6 Ω with a broad SWR bandwidth that is a good match to 50 Ω coax. Its gain is -0.7 dbi at an elevation angle of 15°. This is pretty typical performance for a vertical of this type.

I have chosen 15° as the comparison standard for 80 meters since that is a median value for medium length DX paths, which are the most productive on this band. The longest paths can have angles well below 10°, though not always. Since low angles are the most difficult to attain and are needed to best DX results and high angles are easy with a second, horizontal antenna I choose suitably low angles as the standard of comparison in the majority of my models. In my many articles about 40 meters antennas I used 10°, and lower angles for progressively higher bands.

Adding a second element

I will continue with λ/4 element spacing, which on 80 meters places the identical second element at 21 meters distance. The optimum spacing may be different though, I expect, not by much. My interest here is to evaluate radial topologies. Optimization can be performed later, should one of these arrays be built.

The parasite will be a reflector element. As we'll see identical elements work well to achieve the desired effect in this configuration. Switching between either end-fire directions is straight-forward, being no more difficult than for the design I proposed for the 40 meters array.

With 21 meters spacing and 20 meter long radials the two radial systems overlap. We have a few strategies to deal with this:

• Place one radial field above the other, allowing capacitive coupling between them. Doing so can require making the parasite a reflector since there is near-critical coupling between elements when the height separation is small.
• Lay them in the same plane with electrical continuity where radials cross. This is sometimes done in 4-squares and similar arrays though there is limited benefit in extending radials further than the first crossing. It is usually better to use the extra wire to make a mesh ground plane at the monopole base to reduce near-field ground loss.
• As above but terminate the radials at the first crossing.

I built models for the first and third cases, rejecting the second since it is mostly redundant to the third.

Overlapping, capacitance-coupled radials

The most critical parameter of a Moxon antenna is the distance between the turned-in ends of two elements. That is where the capacitive coupling is strongest since that is where voltage is highest.

Strong coupling drives the parasite current higher than in a conventional yagi. It also restricts the phase relationship between the elements such that above a critical level of coupling the parasite can only operate as a reflector element. This is why I commented in my earlier article on the 2-element ground plane model that it is sensitive to precise placement of the (interlaced) radials.

As I noted above when the second element is added to make a 2-element ground-mounted vertical array the radial systems must be vertically separated and the amount of separation is a critical parameter, as in any critically-coupled array. I modelled the array with a range of separations, each small enough that the vertical offset of the monopoles would not significantly alter the far-field pattern.

With a 10 cm (4") separation (driven element on the right is 20 cm above ground) I achieved the best performance. However I did not try to fully optimize the array: my aim is to generally characterize the array to decide on whether its merits motivate further investigation.

Wire currents at 3.6 MHz, 10 cm radial separation; reflector at left

To demonstrate radial/element coupling I have plotted the modelled currents on the array's top view. The driven element current is normalized at 1 A. The driven element is to the right (red) and the reflector is to the left (blue). Radial currents are roughly maximum about 2 meters out from the monopole: the ground's dielectric constant makes the 20 meter long radials electrical length slightly more than 0.25λ. Strong coupling is evident in the high current in the parasite monopole: 70% that of the driven element at 3.6 MHz. It only appreciably declines when the radial system separation grows to at least 30 cm (12").

Radial currents are not close to equal or sum to that of their respective monopoles. Currents are higher where a radial crosses another, and highest where the far end of a radial is close to another radial. Only where a radial stands clear is the current close to the theoretical 1/16 of the monopole current. It should be evident that radial placement is critical to antenna performance, as it was in the 2-element ground plane.

The forward gain at 3.6 MHz is  about 4 db more than the single element at an elevation angle of 15°. Relative gain of 4.2 db is maximum at 3.5 MHz, and slowly declines to 2.2 db at 3.8 MHz. F/B is poor, ranging from 8.4 db at 3.5 MHz to 9.7 db at 3.8 MHz, and reaching a maximum of 10.9 db at 3.7 MHz.

The SWR is surprisingly good, staying below 2 over most of the band of interest to DXers and contesters. Unlike conventional parasitic arrays the SWR bandwidth is excellent and a good 50 Ω match. The change in SWR with frequency is mostly due to the feed point resistance since the reactance changes more slowly.

Increasing the radial system separation to 20 cm leaves the SWR and F/B nearly unchanged. However the gain at 3.5 Mhz is 1 db lower and the frequency of maximum gain rises to 3.6 Mhz. Even a small change in radial coupling can have a significant effect.

For a simple antenna on which I spent so little time this is good performance. But it comes with some important catches:

• Maintaining the required radial field separation is difficult. Not only is it mechanically challenging it is a safety hazard to have 20 meter long wires 10 cm above ground, whether for pets or wildlife that will inevitably wander into the area. The radials, if bare wire, must never touch any other.
• Weather will alter radial coupling. Winters with snow and ice will almost certainly destroy the radial behaviour, and the structural integrity of the raised radials.
• The antenna is very sensitive to changes in radial coupling. Those changes are usually for the worse.

The study of this radial topology has been interesting and instructive. Since it is not promising for an 80 meters vertical array it is time to move on to the second case.

Radials tied at the mid-point

Terminating and tying radials of adjacent verticals in an array is an old idea, and has been used in commercial broadcast arrays for years. It is also an obvious one, so that it is unsurprising that I independently thought of it before discovering it in the literature. In the amateur field you can read, for example, a discussion by the late W4RNL (Cebik).

There is some coupling in this array, mostly between radials whose ends are close together. However this is dominated by monopole coupling and direct connection of the radial systems. At first blush this would appear to be an array more suited to having all elements driven, with a power splitting and phasing system to achieve the desired result. Antennas such as 4-squares are of this type, though so are 2-element end-fire arrays such as described here.

Connected radial; reflector (blue), drive (red), connected pairs (black)

The radial currents (defined above) vary less than in the array with overlapping radials. First, the reflector current is 58% that of the driven element, which is lower than with overlapping radials though still more than in a conventional yagi. The sums of the radial current for each element are roughly equal to that of the monopole, approaching the ideal situation of a single element where radial currents are equal and sum to that of the monopole (assuming radials lengths near 0.025λ).

Radials that had been capacitance coupled are now directly connected: 16-21, 15-22, 14-23, 13-24 and 12-25. While the currents in these radial pairs are (necessarily) equal where they connect it may seem surprising that currents are quite different at their origins. Power is flowing between the elements, but with a 133° phase difference (at the radial origins). Current nodes occur on driven element side of the radial pairs, closer to the driven element where the current differential is greatest. On wire #23 the node is adjacent to the radial origin (driven element monopole junction).

Despite the relative lack of control over parasite current and phase this array has gain and F/B. However performance is not quite as good as the array with overlapping radials. Nevertheless this is a more realizable antenna since the radials can all be on or just below ground, and coupling requires less fine tuning.

Since the elevation pattern is indistinguishable from the one above I will instead show the azimuth pattern, which is also the same. The plot is at right.

Maximum gain is 2.82 dbi at 3.575 MHz and 15° elevation, which is 3.52 db better than a single vertical. At 3.5 MHz the relative gain is 3.2 db. Going higher, relative gain gradually declines to 2.9 db at 3.8 MHz. F/B is poor. Its maximum is 9.8 db at 3.8 MHz, and an especially bad 3.5 db at 3.5 MHz.

Although I have spoken against the importance of high F/B, particularly for the high bands in a small station with a single yagi, the situation is different on 80 and 160. The bigger problem here is QRN, not QRM. Directivity is needed to improve SNR enough to copy DX stations. If this antenna is built it ought to be supplemented with a low-noise receiving antenna such as a Beverage or compact, rotatable loop. Otherwise be prepare to not hear many stations that call you, especially if you run a kilowatt.

SWR is sufficiently broadband to allow no-tuner use from 3.5 MHz to 3.8 MHz. The above SWR plot is for equal height monopoles (identical elements). This worked well in the array with overlapping radials, though here the resonant frequency drops a little lower than is ideal. A 2,500 pf capacitor should be switched in series with the driven element to bring the SWR below 2 across this frequency range. The value isn't critical, 2,200 pf or 2,700 standard values can be used, but use a ceramic knob capacitor if possible to reduce loss (and potential failure when running a kilowatt).

My take on this antenna:

• It has the makings of a good, simple, broadband switchable gain array without the burden of a power-splitting and phasing system. Even a commercial product to perform these tasks requires work to tune and match the elements and adjust array performance.
• F/B is so bad that a high-directivity receiving antenna be seriously considered.
• Altering the radial quantity and length will change array behaviour. Model first before adding radials to reduce ground loss, since wire lengths may need adjustment.

Again, this is an experimental model and not a design to be rushed into construction without further evaluation.

Disconnecting the common radials

As a final experiment in this phase of modelling 2-element parasitic vertical arrays I will take the previous design and disconnect the 5 radials pairs that are joined at the midpoint between elements. My aim is to see how much capacitive coupling can be achieved and whether it can be used to raise parasite current and thus hopefully improve F/B without the tribulations of overlapping radial systems.

I modified the connected radials model by disconnected the 5 pairs of connected radials and various the separation distance. The results were disappointing. Mutual coupling between elements was low, as evidenced by parasite current (reflector) only around 36% that of the driven element. Gain and F/B performance was, not surprisingly, poor. There was little benefit found by varying the separation distance; I tried values from 10 to 100 cm. More drastic measures would be needed to increase coupling, such as in the overlapping radial case.

The adjacent elevation plot is typical of the performance. Gain is only about 1 db better than a single vertical! F/B was typically below 6 db. Like the connected radials array, gain peaked at 3.5 MHz and F/B peaked at 3.8 MHz.

However SWR performance mimicked that of the connected radials array, being almost indistinguishable and so not worth showing the curve again. I also declined to mark up the antenna diagram with radial currents as it was not worth my time for this unpromising antenna.

One interesting difference was ground loss: it modelled from -1 to -1.5 db worse than the connected radials array. That is undoubtedly where some of the missing gain went. My guess (I didn't look at it more closely) is that the additional loss is due to a third of the radials being about half the length of the others.

Conclusions

Of these experimental models the only one that shows promise is the one with connected radials. It has good power flow between elements, accomplished with a robust physical design. I expect that it can deliver the modelled performance when built.

To get improved F/B from this array it would be necessary to feed both elements and use a power splitter/phasing system to architect the required electrical parameters in each element, in part by "taming" the mutual coupling to do our bidding. That system could also be used for direction switching. Designs are readily available, with perhaps the most comprehensive treatment found in ON4UN's Low-band DXing book (5^th edition), chapter 11.

My approach would be to either build a 4-square or build the 2-element end-fire array described here. In the latter case, going with simplicity and maximum reliability rather than the best F/B performance. A simple, low-cost directional receive antenna would complement the parasitic array. That is, if you have the land.

It is possible to use elevated verticals to reduce ground loss and deal with any local topography and obstacles that impede the view of the horizon. While more challenging on 80 meters it is not necessary to raise the base 20 meters (λ/4) to get the benefits. Half that height can be effective, though probably no lower. There are ample resources on elevated radials, of which I'll point to two available on the internet: by VE2CV and another by N6LF. ON4UN's book also has many ideas in this regard.

But if you do so the antenna must be carefully modelled so that it can deliver the desired performance. You cannot simply lift the arrays discussed in the article and expect that they'll work.

13 db

In this past weekend's ARRL DX SSB contest I did something I have not done since returning to the air over two years ago: I entered a contest using more power than QRP. For the first time I put my recently-acquired FT-1000MP to work in a contest. Running 100 watts is 13 decibels more than the maximum 5 watts allowed to qualify for the QRP category. This is a brief recounting of how it played out.

I entered 15 meters single band since my time was limited and I expected to be competitive by focussing that time on one band. Based on conditions before the contest I chose 15. My choice turned out well. While the band was not open right through the night it performed well up to 3 hours past local sunset, and came up again with the sunrise. The solar flare on Saturday had little impact on me; it had more effect on those further west aiming at Europe and Japan through the auroral zone, and of course those in Scandinavia. I'm far enough east to have consistently good openings to Europe. That makes all the difference.

Run

Unlike with QRP it is possible to make most QSOs by running with 100 watts. True, it is nowhere near as impressive as what a kilowatt and big antennas will accomplish, but still a very effective way to run up the score. It helps that in Canada we can operate SSB below 21.200 MHz, away from the wall of US super-stations. You could hear me and other Canadians lined up from 21.2 downward running Europe for hours on end. Some US stations called me down there, which I could not work both for being out-of-band and worth no points.

This strategy does not work for other areas of the world, even where they can operate SSB lower in the band. But other than Japan the possibilities of runs of stations outside Europe are slim. I did some limited running of JA in the early evenings, and I'd have done more if not for a new local QRN source that peaks to the northwest.

Another advantage of running is multipliers. While working Europe I had a number of callers from Africa, the Middle East and Central Asia. These are multipliers I otherwise never heard. I ended up with 97 countries, just shy of a weekend DXCC on 15 meters. About 10 of those are uniques that I would not have worked but for running. That's what 13 db buys you.

Diversity

With just the one small yagi I am at a disadvantage when trying to work stations off the side or back. It costs time to rotate the yagi to work one or two Caribbean or South American stations when the bulk of QSOs are towards Europe.

As in previous contests I have some diversity in the form of an inverted vee hanging off my house-bracketed tower. The 40 meters element of the cage inverted vee has a somewhat complex azimuth pattern on its 3rd harmonic but does effectively fill the holes in the yagi's pattern. The strategy is to switch to the inverted vee to work those odds and ends that are not always worth rotating the yagi.

With QRP this often did not work out, and I had to rotate the yagi if I wanted the points. Add 13 db and I could make the needed QSOs on the inverted vee. In fact I could often work them when they were directly off the back of the yagi since the F/B is not very high. As I've stated before, this is reason to question the need for high F/B for contests. A notable exception is stations with multiple yagis per band which do benefit from high F/B, since unlike single-yagi stations they can simultaneously achieve diversity and QRM rejection.

Rotation

Concentrating on one band means that you will at times exhaust the pool of available stations to work. If you have a kilowatt and a big antenna you can keep running, seemingly without end when the band is open to Europe by attracting (and hearing!) more of the majority who have small stations. That doesn't work for me, even with 100 watts.

For a all-band effort, as I always have done with QRP, it is best to make frequent band changes. This allows time for a rotation of stations to occur. When, for example, you return to 15 meters there will be a number of new stations to work. Even when the strategy is to run this focus on rotation can work well.

Rotation works since most stations are either single-operator or casual participants, and they can only be on one band at a time. Although this is increasingly less true with the emergence of SO2R entrants (single-operator, two radios), it is not enough to make much a noticable difference. By spending time away from a particular band, or even from the shack entirely, a different bunch of single-ops will fill the band as they QSY from band to band. The available pool of casual operators who will answer your CQ also rotates. You will benefit from this rotation whether you S & P or run.

In my case -- single band and intermittent operation -- my frequent breaks made rotation work for me. If I'd operated full time my rate would have suffered unless I were to operate all  bands. An extra 13 db can do wonders, but not perform miracles.

Working QRP stations

My small antenna and 100 watts was enough to attract quite a few QRP stations while I was running. They are easy to identify in the ARRL DX contests since, for non-VE/W stations, power is part of the exchange.

Some of the 5-watters from Europe, even far-eastern Europe, had good signals that were copyable through the QRM and QRN. Some were a struggle to pull through, but then so were many 100-watt signals. So while SSB QRP is a challenge it should not be dismissed as not worth the effort involved. Being on the other end I was only too happy to pull them through and earn the points. Hearing them also made me smile since I know what it's like to be in their shoes.

It also pays to listen closely to the weak ones as you S & P across the bands. I worked one multiplier this way, a VP5, who was running 5 watts. He was CQing to little effect. I heard no other VP5 that weekend, so it's a good thing I was paying attention to every weak signal I ran across.

On an amusing note someone mused on the N1MM Logger group after the contest about how to log the station who gave his power as 500 milliwatts. Apart from the logging challenge it goes to show just how far QRP (or QRPp) can go, even on SSB. The lowest power station I logged was running 3 watts.

The exclamation point

Later Sunday evening while exporting my contest log and reporting my results I tuned around 20 meters. I ran across the fierce North American pile-up on E30FB, Eritrea. Of course I jumped in. I was still running 100 watts. Mix pile-up tactics and a heaping load of good luck and I got through in only 5 minutes. That was the exclamation point on my +13 db contest effort.

Next up

I have no immediate contesting plan except, perhaps, CQ WPX SSB later this month. So after 3 straight posts about contesting I am likely to return to antennas in my next article.

I know from web site statistics that antenna articles are by far the most popular, especially those that are about a particular antenna and not on theory or other general aspects of the topic. Since antenna articles take some time I can't produce them at the rate of one per week. So you will keep on seeing many articles that are about other topics that interest me, such as the one you are now reading.

I intend to cover some more general points about vertical antennas that I touched one previously. This will set the ground work for specific antenna designs.

CW Skimmers: The QRP Contester's Friend

Having been a QRP contester for almost 2 years now I have concluded that this is a really good time in the hobby's history to be doing this. The reason is technology. We often tend to think of recent technological progress assisting the more serious big-gun contesters (networked logging, global spotting, remotes, antenna design, etc.). Of course the technology also helps average and even smaller stations. What I want to argue here is that QRP contesting may be getting a better-than-average return on modern technology.

The technology I want to look at is not stuff like transceivers and station software since both big and little guns often use the same stuff. It's networking where I believe the benefit lies.

First we need to understand just how challenging QRP can be in a contest environment. My antennas are typical of suburban hams -- tri-bander & wires -- so we're equal in that. But when I run 5 watts I am -13 db weaker than the more typical 100 watts (barefoot) station. This grows up to -23 to -25 db below full legal limit stations. That's a lot when you consider, as hams, we'll endlessly argue about or waver on spending money over 0.5 db of transmission line loss or 1 db more yagi gain.

Listen to me

Let me give you a concrete example of how weak my QRP signal sounds at the other end of the QSO. 3V8SS in Tunisia recorded all his contacts in the recent ARRL DX CW contest. I placed well among the claimed scores in the QRP category of that contest. Go to 3V8SS's contest page and click on the link (upper left) to bring up a search box where you can enter my call (VE3VN) or any other call. Then select the found QSO (by band) to listen to 1 minute of streaming audio containing the selected QSO.

When you finish listening (and stop laughing) you'll understand what I'm up against. When calling other stations I am usually unable to get through when someone else simultaneously calls the other station. Even then I frequently need to repeat my call or exchange. Calling CQ is a particular challenge since on a busy band my signal is an easy one to overlook, and that's on CW with narrow bandwidth filters.

This is where technology comes to the rescue.

Spotting

Spotting only helps if you're sitting on a frequency and calling CQ (running). Someone who hears or works you spots your call and frequency via one of the hundreds of gateways to the global spotting network. Other see the spot and, if they haven't yet worked you, QSY to your frequency and call. Both general logging and contest software make this as simple as a single click. The software may even highlight whether the station is needed and a new multiplier.

I was only spotted once during the contest, which is a little disappointing. Yet I was still quickly besieged with callers many of the times I started calling CQ on a new frequency. This brings us to another bit of technology that brought callers to my frequency.

CW skimmers and the Reverse Beacon Network (RBN)

Spotting networks only work if someone somewhere makes a conscious decision to spot you (note: never self-spot, and it can even get you disqualified from a contest). CW skimmers take that element of uncertainty out of the picture. There are a many stations around the globe running CW Skimmer by VE3NEA, or similar software, on a spare receiver. They may or may not be located at stations active in the contest.

Unlike spotting networks skimmers automatically scan the bands and report on activity. Primarily this is stations calling CQ or otherwise holding a frequency and inviting callers. While a standalone skimmer node has some value to its operator it becomes far more powerful when combined with other skimmers. This is where the RBN comes in.

Since the linked sites on CW Skimmer and RBN describe these technologies in detail I will skip their descriptions here and jump directly to showing why skimming and RBN are so useful to the QRP contester.

Let's do this with an example. The adjacent picture is an image of the RBN search on my call in the final hours of the ARRL DX CW contest last weekend.

As you can see I am getting a multitude of reports of my running attempts, often many every minute. It should be obvious that this is far more productive than relying on other hams to spot me. Most are not motivated to do so in my case since VE3 is hardly an attractive catch for anyone. Besides, most human operators will pay little attention to a signal as weak as mine. Not so the software.

You can see where I spent the final 2 minutes of the contest by calling CQ on 40 meters. It was a way to spend the time since I knew I had little chance of finding someone new in the time remaining. I was quickly answered by 9A8M, which turned out to be my final QSO of the contest. As I ran down the clock the skimmer spots just kept coming on RBN.

Notice the posting of the SNR (signal-to-noise ratio) on each spot. As with the 3V8SS recording referenced earlier you get an idea of how weak I am at many stations. I have few skimmer spots on 40 from the west coast or Europe. That absence of spots most likely indicates that I was below the noise or covered by QRM. This is useful feedback. On 20 and higher bands I fared better with my CQs, as you can glean from the earlier RBN spots.

Call CQ

Even if you're running QRP or small antennas you must spend time calling CQ in a contest if you are to build up your score. Many little guns or casual participants only call other stations, so if you never call CQ you won't work them. That costs you points. In the ARRL DX CW I made a point of calling CQ as often as possible in the final 12 hours of the contest since by then even the big guns are prowling the bands searching for contacts.

With a small number of human-operator spots the existence of skimmers and the RBN is a boon to little guns. Although many S & P operators may pass you by because you are so weak the multi-operator stations and those entering in "assisted" categories will pay you a visit when you appear on RBN. Thus you get a bigger boost from RBN than the big guns, who are stronger and more often spotted or called by the S & P crowd.

So call CQ and let the technology out there help to boost your contest score. Well, at least in CW contests. SSB skimmers will take a little longer to come along.

Getting to 0% Error Rate: "Not In Log"

Logging errors in contests can exact a steep penalty. Not only will the sponsor remove QSOs containing errors they also assess penalties. On average a removed QSO will reduce your score by at least double that amount. If the removed QSOs are multipliers the cost can become very expensive indeed.

With the advent of computer logging, log submission and log checking the wages of sin will always be paid. In the olden days logs had to be manually checked. This was such a labourious process that logs with scores not high within their entry classes were rarely checked. Carelessness that was once penalty-free for 90% of contesters has evolved and is now a guarantee of score reduction. This is not only embarrassing it can also keep you out of contention for awards.

Is it possible to submit an error-free log? Yes, but it takes a combination of diligence and luck. I'll show you what I mean by luck in a moment, but let's first cover the errors that diligence can prevent. Computer logging is both a blessing and a curse in this regard. Software makes it easy to avoid some mistakes and easier to make others.

Call sign and exchange errors

These are the easiest errors to make and often can be completely eliminated by taking appropriate care. Errors include copying mistakes, typing mistakes and pre-fill oversights. Here are some of the ones I am most familiar with, along with ideas on how reduce mistakes.

Busted calls: This can happen on both SSB and CW. In the former case it is usually due to QRM or difficulty understanding a non-native English speaker. In the latter case the causes of copying errors are more numerous. A sequence of short letters (e.g. E, I, T) or characters with 3 or 4 dot sequences (e.g. B, V, 4, 6) at high speeds will often do it. Although there are some contesters with poor CW skills who get by due the short exchanges, everyone is at risk.

If you're running the other station will most often correct your error since you are sending their call to them. However if you're sending at 30 wpm or faster the other guy may not notice the error. I'm surprised how often I hear this happening while waiting to call someone. Worse, they may not care since in the majority of contests they are not penalized for the other guy's copying errors!

Some stations while in S & P mode make a point of sending the running station's call just to be sure they've correctly copied it. But this can be expensive in time if it is done on every QSO. I would only recommend doing this when there is some doubt about the other's call.

Some logging software will flag calls that are suspect by dint of not being known as having appeared in previous contests -- a so-called master database. Flagged calls should be double-checked, just in case. Be particularly wary of partial call databases, used by you or the other station, that will attempt to pick a likely call from the master database if the call is suspect or incomplete. It is often wrong. I am a regular victim of this feature since the other station often has difficulty copying my QRP signal. 

Exchanges: In most contests the exchange is fixed and often predictable from the call. For example, in this past weekend's ARRL DX CW contest my exchange was "599 ON". The RST is of course routinely sent as 599 in every contest exchange. The state/province is a constant and in the case of Canadian calls is strongly correlated with the call sign prefix. Most logging software will analyze the call and pre-fill the exchange field for you, saving you time and potential for a typo. This can also be done for CQ and ITU zones in contests where those are part of the exchange.

US locations in particular are less predictable. It has been many years since a call sign prefix correlated to a state or zone. Logging software makes a best guess but it is up to you to confirm that what the other station sends is what the software pre-fills. It is too easy to get lazy, especially when you're tired, and not confirm the fill before logging the QSO. I can only say that you must be diligent. This is the cause of some of my errors.

If you are uncertain of your copy in contests with a more complex exchange (e.g. Sweepstakes) you should immediately ask the other station for confirmation or a repeat.

Some logging software allows you to use a history file built from previous contest logs (your own or by importing one) to more accurately pre-fill the exchange. The percentage of mistakes will be reduced, but you must still be diligent and confirm what is sent matches what is pre-filled.

Correcting errors on the fly: When I note an error and make a correction on the fly it is not always possible to record the correction in the log. Even if you're highly effective at manipulating the logging software this can take several seconds, during which the other station is wondering what's going on while others waiting for you (if you're running) may QSY. I keep a pen and paper handy to jot down the needed correction since it is sometimes the fastest way to move onward. The log can be updated even just a minute later while your memory keyer or DVM is playing a CQ.

Another technique during S & P is to stick around a few seconds to hear for a second time their call and/or exchange. The call can be quickly corrected this way. However there is a danger if the error is in the exchange. If the running station has a poor rate you could be waiting a while for their next QSO to hear the needed information. Rather than getting into this undesirable situation you should not be shy about asking for a repeat during the QSO.

Not-in-log errors (NIL)

NIL errors are unlike the errors discussed above in that they are not entirely under your control. That is, you cannot be certain that the other operator logged your QSO. You can reduce NIL errors but not eliminate them entirely. This is a topic I previously discussed so you can reference that article for background. What I want to address now is some of the "why" NIL errors are so difficult to eliminate.

I will use a real example. Although the results are not out I am one of the participants in last fall's (November 2014) CQWW CW contest sent a preview log check report (LCR) for review. My error rate is comfortably under 1% so there is little risk of losing my provisional #1 North America position in the SOAB (single op, all band, unassisted) QRP category.

Apart from a small number of busted calls and improperly recorded exchanges there were 3 NIL errors. In trying to understand what might have happened, despite my effort to eradicate NIL during the contest, Randy K5ZD suggested I look through the public logs. He seemed sufficiently confident in their log checking software and processes to punt the question back at me. So I did what he suggested.

Case #1: This occurred while I was running on 15 meters. Getting an NIL while running is odd since the other station would seem to have obviously copied me or would not have called. I have assumed, perhaps mistakenly, that the majority of NIL errors occur while S & P. This is an ongoing risk to those of us operating QRP, and especially so on low bands. Unfortunately the other station's log in this case is uninformative. He was indeed on 15 at the time but his log does not provide the frequency for each QSO, just the band (21000).

Case #2: This one was also while I was running, this time on 20 meters. The other station was multi-single so the log contains the interleaved QSOs of two operating positions. One of those was on 20 at the right time, however that station was running and positioned far from my frequency.

Case #3: Unlike the previous two cases this is one that is easily determined. It was a European on 80 where all of my DX QSOs are marginal. His log confirms he was there and running, and I have some recollection of the QSO. I recall that he did copy my call, with difficulty, though he seemed uncertain and may not have copied my exchange. After a while he moved on, though he first gave me some indication that the QSO was complete. Usually this is a "TU" or something similar before soliciting the next QSO, although I don't remember what it was in this case that decided the case for me that I should log the QSO. I may have been unduly rash since it was a multiplier, something that is very precious to me on 80!

Resolution: As Randy noted confusion often reigns during a contest and some NIL of the sort I had are not uncommon. It is entirely possible for an adjacent QSO (of which you might only hear only one side) can have just the right timing to make it seem that it was with you. This may have occurred in case #1. I have no idea what might have happened in case #2. The call was that of a big gun and not likely to be a copying error on my part, and even if so the log checking software would most likely have determined the correct call from scanning other submitted logs. On this basis there is little I could have done to avoid these two NIL errors.

Case #3 is a dilemma. When there is uncertainty in a situation like this there is no best choice. If I log the QSO and it is not logged by the other station I lose the QSO and multiplier and get assessed a further penalty. If I don't log it not only I will lose the QSO and multiplier, which the other station might have logged, I could cause the other station to suffer an NIL. That would be inconsiderate of me.

This comes back to suggestions above regarding getting as much confirmation of critical data during the QSO to eliminate these doubts about whether to log it. On low bands where the other station is barely able to copy me this is difficult. That does not rate as a justifiable excuse. You just do the best you can.

Hole-in-One

Getting to a 0% error rate is difficult, as I hope my experience has shown. It is not purely a matter of chance: if you are sloppy I can guarantee you will never achieve 0% errors. Yet even if you are extremely diligent I cannot guarantee you will eliminate all errors. The reason is that what the other operators do, the hundreds or even thousands of them you work in a contest weekend, is out of your control.

At its fundamentals the error rate is a stochastic process, where there are one or more random variables. We can draw an analogy from the game of golf. The top players can never guarantee getting a hole-in-one, not in any particular match and perhaps not in their lifetimes. What they have is a higher probability of doing so than everyone else who has ever picked up a golf club. This is because they can more often hit their tee shots onto the green. If you can't do that, or do it often enough, you will never score a hole-in-one. Unless you get very, very lucky.

Operating a contest is much the same. You may never score a 0% error rate but your chance of doing so dramatically rises when you make the effort to reduce errors. Even if you don't get to 0% you will see a big improvement in the difference between your claimed and published scores, and that will push you up the leader board past those who don't make the effort.

2-element Parasitic Ground Plane for 40 Meters

As the solar cycle declines my attention to the low bands increases. This means I have an increased interest in vertical antennas, even though "serious" low-band antennas are impractical on my suburban property. So I instead plan and model in the hope that one day I can start building large antennas.

Experimenting with 40 meters is in some respects easier than doing so on 80 and 160. Antennas are easier to successfully model and prototype at the shorter wavelengths, and then compared to existing antennas. Antennas that work out can then be scaled to the longer wavelengths with predictable results. There are also some things you can do that may be too mechanically challenging on lower bands. The antenna in this article is one of those.

Concept antenna, not a final design

Please keep in mind that this is purely a concept antenna. It is perhaps worthy of prototyping but should not be seriously considered without further work. I did not aim to fully optimize the design. My intention is to play with it to see it if has good DX performance with regard to gain, F/B and match. As we will see there are design aspects that require more work before committing to construction.

Another motivation is to see what can be accomplished with a relatively simple vertical array in comparison to well-tuned, switchable 4-square antennas that many big guns use on 80 and 160. Although this design is for 40 the antenna can be scaled, with some construction effort and expense.

My concern with the 4-square and similar directive arrays is the intricacy of the feed and switching networks. This invites the potential for failure when one or more elements is affected by weather (e.g. snow and ice) or a component drifts in value or fails. That is why 4-squares often have a "dump" resistor and a warning system. The design lacks a degree of robustness.

This is not meant to demean recent efforts at optimum phasing and power division, which can be very impressive. Designs such as those you'll find in chapter 11 of ON4UN's fifth edition of Low-band DXing are so bizarre because the elements are often so closely spaced that mutual coupling dominates. To my mind it is better to use judo rather than brute strength to tackle the problem. That is, to accept mutual coupling as a design partner rather than having to coerce it to behave in a prescribed manner. I try to do that in this article.

Copy and spin

There are many ways to make a parasitic array. You can even do it with λ/4 monopoles with ground planes, if you're careful. Consider the simple 4-radial ground plane mentioned near the end of my recent article on vertical modelling experiments. It is broadband, with low-angle radiation (good for DX), and is a good match to 50 Ω coax.

I proceed in the usual fashion of making a parasitic (reflector) array by copying a resonant single-element antenna and offsetting the copy in the desired direction. To avoid tangling the radials, which are about as long as the inter-element spacing, I spin the radials of one element by 45°.

With the elements mounted 10 meters above ground they a maximum height of about 20 meters. The bases can be light-duty television towers or masts made from aluminum or steel. I prefer the former so that the feed point is easily accessible for assembly and maintenance. Guying is necessary but there is no need for concrete. The radials can do double duty as guys, if you use something stronger than aluminum or pure copper wire.

With a bit of open space this should be an inexpensive and not too challenging antenna to construct. The minimum required area is 500 m² (25 x 20 meters). Even though mounted above ground it is recommended that there be no other structures within 1λ (40 meters) radius, and even farther from towers or other large conductors.

It's a Moxon!

Okay, it doesn't look like a typical (or even atypical) Moxon but it does have the same attributes. The reason is those radials: the way they intermingle causes near critical coupling between elements. Even with the relatively wide spacing between monopoles of 10 meters (close to 0.25λ) those radials keep the coupling high.

You can see this most clearly in the radial currents. Rather than the current in each of the 4 radials being ¼ that of the monopole it varies a lot, depending on the radial's position. One implication is that the parasite must be configured as a reflector. Well, you can try to make it a director but you'll find that task quite difficult unless you can find a way to reduce the coupling.

In the driven element (wires 6 to 10 in the current plot above) the radial currents are 6% and 43% that of the monopole in the rearward and forward radials, respectively. In the reflector (wires 1 to 5) the radial currents are 53%, 34% and 40% that in the monopole in the forward, rearward and side radials, respectively. These values are at 7.1 MHz, near where gain is maximum. No only are the radial currents unequal their sum only equals the monopole current in driven element. The sum far exceeds the monopole current in the reflector element.

The current in the reflector monopole ranges from 60% to 70% that of the driven element across the band. This is typical of a Moxon and higher than in a conventional 2-element parasitic array. As we'll see, gain, F/B and SWR vary less than a 2-element yagi across the entire 40 meters band.

Tuning for optimum performance

Performance is sensitive to element separation (equivalent to boom length in a conventional yagi or Moxon rectangle) and radial arrangement. The thing I found annoying is that although gain would vary a lot when these are adjusted a little, the feed point impedance (and thus SWR) and F/B were far more stable. It's annoying because the latter two are far easier to measure, and would simplify the tuning of this array.

In the model I settled on monopoles 9.95 meters tall and made from 25 mm (1") aluminum tubing. This is an approximation to a real antenna that would use telescoping, tapered tubes. The monopoles are 10 meters apart with bases 10 meters (~λ/4) above medium ground. The radials are 16 AWG aluminum wire, such as the often used (and inexpensive) aluminum fence wire. All radials are 10.525 meters long and slope downward 30°.

I²R loss in the aluminum monopoles and radials is around -0.1 db. The pattern plots are with zero loss conductors, including for the reference single ground plane. The performance chart below includes the loss. Ground loss is calculated by EZNEC to be about -5 db over medium ground.

Considering how peculiar an antenna this is its performance is quite good. Not only is it 5 db better than a single ground plane at 7.1 MHz (comparison in the elevation plot above) the gain varies little up through 7.2 MHz. The F/B, while not exceptional, is adequate to my needs. The SWR is particularly nice, staying below 2 across the band.

Compared to various single element and 2-element antennas the 10° elevation gain is very good. For example, it equals an inverted vee (broadside) 25 meters high and even a 2-element yagi up 15 meters. If you don't have a high tower but do have some open land this could be an attractive antenna choice. Don't expect this performance on a suburban lot since vertically-polarized antennas can under-perform the models.

As I stated at the beginning, I did not really design this antenna with the objective of putting in on 40 meters. If I eventually have higher towers there are superior alternatives. This experiment is about getting a handle on antennas for 80 and 160 where, often, towers cannot be high enough to make a horizontal antenna competitive. When scaled to 80 meters the the required towers for comparable horizontal antennas must be twice as high. Scaled verticals for 80 lose relatively little gain at low height (though half as high in wavelengths) but may suffer from additional ground loss and environmental interaction.

Although the monopoles are the same height the reflector element has a base loading coil. It is a very small coil with an inductance of 0.4 μH. Although small this value is critical. Even if only 0.1 μH higher or lower the SWR and gain will noticably suffer. F/B is less sensitive to inductance changes. Small changes to the radials will affect the required value of this coil since their length and position affect mutual coupling between elements.

To give an idea of the dimensions involved, a suitable 0.4 μH coil would be 5 turns with a length and diameter of 1" (2.5 cm). The equivalent shorted 300 Ω open wire stub would be 16" (40 cm) long at 7.1 MHz. The stub is easier to tune (with a sliding shorting bar) but will be exposed to the weather which can alter its reactance.

Direction switching

It is always helpful in a fixed element array to be able to change the direction of the beam. This antenna is amenable to such an arrangement, though with some stringent construction and tuning criteria. This, too, is typical of Moxon (critically-coupled) arrays.

The switching method I describe here has significant differences to the one I used for the various styles of 40 meter wire yagis I described over a year ago. I am assuming that the support masts for the elements allow convenient access to the feed points which are 10 meters above ground. This is why I recommend light duty television towers for the supports.

The switching system requires 2 plastic enclosures, 3 DPDT relays, 2 loading coils, DC switching circuitry and 50 Ω coax for the transmission line and running between the element feed points. One element will be connected through to the transmission line while the other will be isolated from the transmission line and a coil connected between monopole and radials. Relay DC power can be run by separate cable or on the transmission line. In the latter case a DC cable is still needed between the 2 enclosures at the bases of the monopoles.

The recommended arrangement has the transmission line terminate at the enclosure at the base of one of the monopoles. Make it the one closest to the shack to keep the total length of coaxial cable to a minimum. The length of coax between that enclosure and the one at the base of the other monopole is non-critical, except that it must (of course) be at least 10 meters long, the element separation. That section of coax will be fully isolated on both ends when the second element is a reflector. Similarly the transmission line will be fully isolated from the first element when it is a reflector. The outer conductor of the both lengths of coax must not be connected together or to the radials, except when connecting to the driven element. This is why the enclosure cannot be metal. Metal will also alter the coil's inductance if it is inside the enclosure.

One relay selects the element to connect both conductors of the transmission line. The other two relays switch each element between the coax feed and the loading coil. The coil provides the inductance to configure that element as a reflector. A shorted stub can be used in lieu of the coil. Obviously when one element is connected to the transmission line it becomes the driven element and the other is a reflector. The normally-off position of the relays should point the array in the most important or commonly used direction.

The transmission line outer conductor will act as a radial when the first element is driven. To avoid asymmetry and mistuning in that direction it is best to run the transmission line along one of the radials and choke it where the radial terminates, making them the same electrical length and at the same potential. A coax coil choke can be used if tuned to be effective on 40 meters. There should also be a choke where the coax enters the second enclosure to ensure it also does not become a radial when the second element is driven.

The two supporting masts and the isolated connecting coax are λ/4 long and so should not interact with the array. The supporting masts should be isolated from ground to make them non-resonant.

Broadside

It is possible in many vertical arrays to create a feed arrangement (power division and phase) to convert an end fire array to broadside. The idea is that it's often more convenient to come up with an elaborate feed than put up more elements, whether two independent 2-element arrays or a 4-square. With the antenna described in this article this is not possible.

Pattern flexibility requires that the factors under one's control (phase, power division, element tuning) are sufficient to sculpt the desired result. With a near-critically coupled antenna like this one the mutual coupling tends to make any effort in that pursuit moot: the mutual coupling dominates. Coupling can be reduced by moving the elements farther apart, but that would negate the parasitic benefits of the array. More elements can be added, but that too has significant costs.

The best that can be achieved is an omnidirectional pattern by splitting the feed in the centre and running equal lengths of coax to both elements. The resulting feed point impedance is 15 Ω, which would require a 3:1 unun. In the broadside configuration here is less than 1 db of gain in comparison to a single ground plane.

I believe it is better to leave the array as is and use another antenna, even just an inverted vee, to fill the holes in the array's end fire pattern.

Next steps

This is an experimental design that I do not recommend be built by anyone. Because if you do you'll have to deal with its various idiosyncrasies either in your own modelling effort or (literally) in the field.

Before I would undertake construction of this antenna I would model variations in radial deployment to modify coupling in a way that preserves the array's best attributes while making it less sensitive to minor variations in tuning. Another thing I'd like is to find an easily-measured quantity that correlates well with antenna performance. For example, in the various 2-element switchable 40 meter antennas I've discussed in the past this metric was the frequency of maximum F/B.

Of course all of this modelling and planning is contingent on having the space to build it. That is tentatively in my long term plans, although for the present I can only design, plan and model. It's time well spent.

KX3 Review by a QRP DXer and Contester

I've been using an Elecraft KX3 as my main station rig for 2 years, ever since returning to ham radio after over 20 years of being QRT. I purchased it since it met my needs at the time: high performance, and; low-impact operation. Now that I've added a FT-1000MP to my shack it seems a good time to review how the rig performs, and how it meets my primary interests of DXing and contesting. The latter activity was not in consideration when I bought the KX3 since I didn't realize that I'd return to contesting.

There are ample reviews of the KX3 so there is no need for me to do the same. In any case I do not have the test equipment needed to do a proper technical review. You can find those elsewhere, such as the QST review and user feedback on eHam. My intent is to give an operator's impression of the rig, both its high points and lows, as applied to DXing and contesting.

My review is in some degree going to be unfair to the rig. It was not designed to primarily serve as a base station, or for the most demanding applications. Take this perspective into account as you read onward. For its intended use the KX3 is a superior product. I mean that whole-heartedly.

Overload

The receiver front end is fragile. There are two significant categories of signals that overload the receiver: strong signals within the roofing filter; strong signals farther away, even very far away. This is with the optional KXFL3 roofing filter installed, which I ordered with the rig and is factory calibrated.

My impression from regular use is that the roofing filter does the intended job of reducing IMD from strong adjacent signals. However when very strong signals fall within the roofing filter the rig may turn off the pre-amp. This is reasonable when that signal is from the ham who lives a couple of blocks away, but less so when the signal is from W8. There is a brief message on the display when this happens which I have rarely noticed while tuning across these signals. A minute later I'm wondering why the band is suddenly quiet. When the rig was new I was always puzzled as to why I would occasionally notice that the pre-amp was off. Eventually I discovered what was going on.

The other problem is fundamental overload where a strong signal into the front-end, even if well outside the ham bands, is detected regardless of VFO tuning. This can be very inconvenient when it occurs. For example, there is a shortwave broadcaster in the US with an exceptionally strong signal here. Even with just a dipole that station can push its way through the receiver and be AM-detected no matter where the rig is tuned in the 20 and 17 meter bands. A narrow bandwidth on CW reduces the interference though not enough in many cases. Turning off the pre-amp or using the attenuator will usually cure the problem, but at the cost of sensitivity.

The better your antennas the worse the overload problems since signals presented to the front-end are going to be stronger.

Receiver DSP

As with all electronics the speed and capacity of DSP hardware has remarkably improved over the years. When coupled with effective software there is the potential for superior performance. In my experience the KX3 does very well in this regard.

The continuously variable receiver filter bandwidth (50 Hz steps) is a delight to use. At all settings the audio signals come through very clearly. Even at the narrowest bandwidth (50 Hz) ringing is low. It is also convenient that the roofing filter automatically adjusts in response to the filter setting.

There are limits to what the DSP filtering can accomplish in this direct conversion receiver. Filter skirts are shallow. For example, I typically set the filter to 200 Hz on CW and yet can hear well outside that narrow range. It is how I compromise between noise reduction (local QRN) and not missing much as I tune across the bands. During contests I often narrow the filter to 100 Hz for S & P (search and pounce) and only need to open it up to 300 Hz when calling CQ, and I don't miss much. The APF is pretty much superfluous on this rig.

Multi-pole crystal IF filters are needed to get the steep skirts needed for best QRM and noise rejection. This is not only impossible in a direct conversion design these filters could not fit inside the KX3's small cabinet.

A bigger issue is opposite sideband rejection for single-signal reception. This is another case where the shallow filter skirts limit performance. Elecraft is well aware of the issue but their suggested solution is one that has side effects I don't care for. So I live with it. It is a particular problem in popular CW contests where stations are packed close together. Even middling strong signals the other side of zero beat interfere with the desired signal. It can also be confusing in that a seemingly interfering signal is in fact 1 or more kHz away. I find it necessary at times to jog the VFO to determine whether the signal is really within the pass band.

Noise Blanker (NB) and Noise Reduction (NR)

I have a lot of noise to contend with throughout the day when neighbours are awake. Most seems to come from LED lighting power supplies, which is added to by various appliances and computers. The worst band is 80, followed by 30, and lesser noise on 20 through 10 which varies with the yagi's direction. For some reason 40 is mostly quiet.

Since I can't change the world or hope to track down every noise source and convince my neighbours to fix the problems the receiver must have effective noise control features. NB and NR are important features.

The NB in the KX3 sometimes works well and sometimes makes matters worse. The NB was not adjustable until a firmware version that came out in (I believe) 2013, after I'd owned the rig for a while. Unfortunately the variability helped little so I typically keep it at or near maximum (15).

The major problem with the NB is its susceptibility to signals outside the filter bandwidth and within or near the range of the roofing filter. Even if those signals are of middling strength the NB is heavily modulated and pretty much renders the desired signal unreadable. This appears to be a case of where the DSP filtering is unable to make up for the limitations of the basic receiver design.

The NR sometimes works, though as with most rigs with NR it typically works best on SSB, and not so well on CW. This is especially true with weak CW signals, where the NR is of no benefit. In my experience adjusting the NR doesn't help on CW.

When the noise is making a CW QSO difficult I often try the NB first. In the majority of cases I get a better result by tightening up the filter to 50 or 100 Hz. In contests where most of my QSOs are S & P (search and pounce) I use a narrow filter of 100 to 200 Hz to contend with the noise and QRM while ensuring I miss few signals as I tune across the band.

VFO Attributes Memory

Every rig associates a variety of controls with each VFO-band pairing, though not always the same ones. On balance I feel that the KX3 does better at this than some other rigs I've used. For example, the KX3 remembers these important settings with the VFO-band pair: attenuator, pre-amp, NB, NR, mode, frequency, and perhaps some others. Since noise and signal levels are band and antenna dependent the first four of those are very welcome. Other rigs often don't associate these with the VFO-band pair, requiring some button pushing on every band change.

What I wish the KX3 would remember is the filter bandwidth. It associates this setting with the mode but not the VFO-band pair. Yet like the items mentioned above the filter is also a control I set differently per band, for reasons of noise and activity level. At least that's how I operate.

VFO, RIT and XIT tuning rates

These features work pretty well aside from the small size of the controls. The problem I have is one of tuning rate. It can take many turns of the knob to effect the desired offset.

The VFO rate is less of a concern to me than RIT and XIT. When I am scanning the band, either in a contest or in daily operation, the VFO tuning rate is fine. For larger frequency excursions it is too slow. Tuning rate can be sped up with a long press of a button, but then the steps are quite large and unsuitable for tuning in a station. Alternative rate settings in the menu are less than ideal to my style of operating. It's easier to use the rig control pane of logging software for large QSY steps.

RIT and XIT tune so slowly that they are very cumbersome for use in split operation. It is almost always better to use both VFOs for split and leave the RIT and XIT for minor adjustments on ordinary QSOs.

QSK and VOX

The QSK (CW break in) and VOX (SSB) work very well. I have heard some complaints about the rig's QSK though that is the opposite of my experience. I use it full time on CW, and I use CW for over 95% of my QSOs. At well over 30 wpm I can hear clearly between dots. There are no annoying switching transients.

When I decided to enter a SSB contest last year I had planned to use PTT with a foot switch, just as I did years ago. I ran into some difficulty wiring in a foot switch and time was pressing, so I decided to try VOX. It worked so well that now I only use VOX when operating SSB with this rig. It was easy to adjust, reasonably immune to inadvertent or extraneous sound, and never cut enough of the first syllable to cause copy errors at the other end in rapid-fire contest exchanges.

ATU-free operation

I did not purchase the optional ATU (automatic tuner). This could have been added later if I got tired of manually adjusting an external tuner. But once I transitioned from my eaves trough antenna to resonant outdoor antennas I found that tuning was not necessary.

At 10 watts output I rarely triggered the protection circuitry even when the SWR got as high as 4. The only serious problem I encountered was common mode from my eaves trough antenna that was easily cured with a coax choke. It similarly was happy putting out 5 watts on 160 meters when feeding the coax centre pin of my 80 meters half sloper. In this case the measured SWR was close to 5.

Buttons and ergonomics

The KX3 is really small and lightweight. This makes it an ideal portable rig, it's designed purpose. As a base station -- how I any many others use it -- size is a problem. When I describe the rig to others I like to say that I can cover the entire face plate with one hand. It's no exaggeration.

The majority of controls are on one of the main circuit boards, which is mounted directly underneath the face plate. When you press a button you are pressing on the circuit board. Although the physical support for the board is quite good I worry that it is prone to eventual failure. If a button breaks or the board cracks the repair implication can be dire. Further, even with my relatively light touch (I'm not physically abusive with my rigs!) the entire rig will slide backward. The rubber feet don't help since the KX3 is very light.

Another problem with its small size is the overlaying of features on the available controls. In addition to a primary use each button or knob can access (presumably less-used) features by long press, pushing knobs, or by pressing another button first. While necessary in a compact design it does cause operating grief. For example, I have several times managed to QRM a DXpedition when trying to operate split since the A=B key is also a split selector on a long press. Playing back a CW memory requires 2 button pushes making it almost useless due to the delay involved between action and result, and the potential for error.

For some reason the normal CW mode operates on the lower sideband (LSB). This is opposite to most other rigs, requiring that I set the mode to CW-R (reverse, or USB) to suit my operating style. In N1MM Logger+ I had to locate the option to use CW-R rather than the default CW mode, otherwise every clicked (self-)spot reverses my preferred mode. It's a solvable problem though not one I find convenient.

Microphone flexibility and SSB

As originally designed the microphone gain could not be set to accommodate a dynamic microphone element. For some reason Elecraft was reluctant to change this despite numerous requests. I was surprised since, as it turned out, it was correctable with firmware alone. To their credit they did relent and I was able to use my venerable Heil headset with the KX3.

Despite the company's protestations about noise and hum that can come with high gain I have experienced no such problem. Ultimately I am happy with the outcome and the KX3's flexibility in regard to the variety of mic elements and handheld controls they support.

After adjusting the gain and ALC in accord with the manual I also added compression for more SSB talk power. The feature itself is standard on every rig on the market although not always done well. From numerous solicited requests on the air I feel confident in saying that when properly adjusted the KX3 produces a high quality, easily-copied SSB signal. This is with the Heil element that emphasizes the mid-range; I set the KX3's mic equalization to flat.

Computer integration

This was trivially easy: just plug one end of the supplied cable to the rig and the other to a USB port. When I first purchased the KX3 not all popular software supported it since it was relatively new, a lack that has since been filled. I have no issues with any of the standard and contest logging software. At least not since setting communication speed to its highest setting of 38,400 bps. Otherwise the logging software was slow to track mode and frequency changes.

One thing I particularly like is that only the one interface is needed for contesting since it can be used to send CW and play rig-resident message memories. That is, no additional serial port is needed for PC-generated CW. However when used in this way there are some things to be kept in mind.

In N1MM Logger+ a macro can be used in the function keys to tell the KX3 to send CW. For example,

{CATA1ASC KY tu {MYCALL};}

closes off one QSO and invites others to call. The KX3's KY command converts the following text to CW, delimited by the ';'. CW speed is determined by the setting on the KX3.

A problem arises because once the macro is executed and the command sent to the rig the software does not know the ongoing progress of the message since there is no feedback from the rig. In N1MM Logger+ (to give a particular example) the following must be kept in mind:

• ESC can't be used to interrupt transmission. You must tap the paddles.
• CQ auto-repeat interval is from when the command is sent to the KX3, not from end of the transmission. It is best to disable this feature and press F1 as needed.
• Sending speed cannot be easily adjusted from the software. For example, the "<" and ">" speed change commands don't work, nor do the PageUp and PageDown keys.
• If you operate SO2R or switch rigs the CATAASC macros must be swapped out. I have this problem now that I have a FT-1000MP, as soon as I start using it in a contest.

I consider this a small inconvenience, well worth the benefit of interface simplicity. At least for now.

Wrap-up

There are other things I like and don't like about the KX3, though those are relatively minor items that I won't spend time writing up. Features I don't use I can't report on at all, including everything related to digital modes and battery operation.

I plan to continue using the KX3 for contests in the coming months. Although I have the FT-1000MP it requires some work to be ready for the demanding requirements for contesting (CW filter and various mods), and I do not want to risk neighbourhood RFI. That is, my contesting will remain QRP.

Every rig has its problems, including the KX3, so the negative points I've made should be kept in context. There is also the matter of personal preference, where my preferences can be very different from that of others. The KX3 is well regarded in the ham community, a reputation it deserves. Be sure it meets your needs if you are thinking of getting one.