Friday, April 22, 2016

Antipodes DX

While discussing recent DXpeditions with a friend we noted the interesting propagation we have from here to the south Indian Ocean and nearby Antarctic islands. This is due to the antipodes of VE3 being located in that wide expanse of ocean skirting Antarctica to the southwest of Australia.

The antipodes is the opposite point on a sphere. On Earth you can easily locate it by inverting your latitude (north to south, or south to north) and adding or subtracting 180° to your longitude. There is a web site that makes it easy to find the antipodes of any point on Earth. Doing it for Heard Island nets the following humourous picture:


As you can see the antipodes of Heard Island is near Prince Albert, Saskatchewan (VE5). This is approximately 3,000 km west of my location of Ottawa. The antipodes of Ottawa is between Kerguelen (FT5X) and Perth, Australia (VK6).

The Wikipedia article I linked to above provides the following helpful map that shows at a glance the antipodes of any location on Earth.

Original can be found in Wikimedia Commons

Geodesics

It is worth understanding why antipodes propagation is so interesting. It concerns geometry on a curved surface.

Every ham who has ever worked DX knows there is a compass point to which you ought to point your directional antenna for best results. The line from you to the DX is a geodesic, although we more often call it a great circle path when referring to the Earth. A geodesic follows the surface of a geometrical object without veering left or right. It is the shortest distance between two points on a surface, curved or flat.

Radio waves most often follow a geodesic from the transmitter to the receiver. Scatter and skew paths make this less than a universal rule, however a geodesic is dominant on 40 meters and up. You can follow a geodesic in two directions, which in radio we call short path and long path (for a reason that should be obvious). Long path is less common but occurs more than you might realize. During solar maximums it is even possible to hear your own signal going right around the Earth back to you 135 milliseconds later. That is long path in the extreme (40,000 km).

Antipodes propagation and geometry

I would have liked to experience the recent VK0EK DXpedition from VE5 since antipodes propagation is quite fascinating. It is a relatively rare experience that I get from working VK6 and the nearby French territorial islands. It was very much in play during the Amsterdam Island (FT5ZM) DXpedition, which I worked without too much difficulty with only 10 watts, even on 40 meters. They could be heard on at least one band most of the time they were active.

Great circle is a circumference (from Wikimedia Commons)
The reason for the unusual propagation is geometry. It begins with a few basic facts that, I think, everyone is familiar with:
  • Radio waves follow a great circle. Skew path and scatter can also be exploited, though they are less common and are most often weaker than the direct path.
  • There is a unique great circle connecting two points on a sphere. The shorter distance between those points is the short path, which is the most common one we experience.
The second point has one important exception, for points that are at or near their antipodes. There are an infinite number of great circles between your location and your antipodes. That is, every compass direction carries the potential for direct path propagation.

This is both an opportunity and a dilemma, since there is no unique answer to the question: where do I point my yagi? The answer is to point it in the direction that works best! The direction that works best will often be the same as short or long path to distant locales that are concurrently heard. This is most notable on 20 meters and above. Indeed, the higher the frequency the fewer the choice of directions since on such long paths you are increasingly likely to encounter a patch of ionosphere with a lower MUF in many directions.

On 30 meters, which is less sensitive to the traversal of parts of the globe that are in darkness or daylight, the direction is less predictable since the MUF is more amenable and absorption is lower in the D and E layers of the ionosphere than on lower bands. On 40 meters and below the constraint is on paths in darkness or at least not far into daylight beyond the sunrise/sunset terminator. Unless you have a directional low-bands antenna you likely can't determine the actual path, only that the path is or is not propagation.

Some experimentation may be required on the high bands to know what direction works best. If you have more than one antenna, even a fixed wire antenna, the task is easier since signal strengths can be compared. For example, if the signal is stronger on an inverted vee than a yagi you know the yagi is not pointing in the best direction. On 40 meters where I have two inverted vees I can often resolve an approximate path.

I have noted signals from VK6 ranging over many compass points at different times of day and year. That is the curiosity that is antipodes propagation.

Antipodes propagation can be compared to lottery tickets. When you buy one ticket you get one chance to win. With antipodes propagation you get many tickets for the same price, and so many more chances to win.

Geographical dependence

The antipodes of North America falls across the Indian Ocean and a sliver of Antarctica. The lack of amateur radio operations in that area results in antipodes propagation being fairly rare for us. Perhaps that's why I find it so intriguing, due do its rarity, often only making a regular appearance during the occasional DXpedition.

It is less rare in other locales. For example, between New Zealand and the Iberian peninsula. But with 71% of the Earth's surface covered by oceans the odds are that your antipodes is not populated, just like mine. Except, that is, when a DXpedition to an uninhabited island takes place.

Saturday, April 9, 2016

Height vs. Gain: 40 Meter Yagi Implications

When I began my survey of larger 40 meter yagi antennas not long ago it was with the intention of selecting one for eventual construction or purchase. It is planned for the top of the 44 meter tall tower I currently have stored in my backyard. That survey is now nearing completion. There was more to it than what I published in this blog.

The focus on gain and overall performance in my survey hides an ugly truth behind all the variations and subtleties of these yagis: height can be a more potent factor than gain. Now is the time to look again at all these antennas and see how they stack up (pun intended) when compared at various heights. The heights I'll investigate are greater than in my earlier survey of wire antennas and arrays in order to firmly establish the study in the context of that big tower.

Although the majority of readers will never have a tower or antenna of the sizes I'm looking at there are lessons for everyone. So read on no matter your ambitions for 40 meters.

Antennas in the survey

I selected antennas both as references and as representatives of one or more yagis in a performance class. The metric gain, and nothing else. Not just gain but gain in free space. Where the antenna shows significant gain variation with frequency I chose a frequency that has the gain reference point I used to chart its performance. This is most applicable to 2-element yagis of all types, and typically less so as the number of elements increases.

With that done I placed each antenna at 10 meters height (apex height in the case of inverted vees) over EZNEC medium ground. The gain at 10 elevation is measured and plotted. This is repeated every 2 meters height up to 40 meters, which is approximately the height I am targetting for a rotatable yagi.

Here are the antenna I used, along with the free space gain at the reference frequency. Links to earlier articles are inserted where applicable.
  • Inverted vee with an angle of 90° between the legs (1.5 dbi). This is my choice for a base reference since it is the simplest antenna to mount on a tower.
  • Dipole shortened with a small capacity hat and coil, similar to the driven element of the Cushcraft XM240 2-element yagi (2 dbi). The model overstates the gain by approximately 0.3 db so you may want to subtract that amount when looking at the charts.
  • 2-element switchable inverted vee yagi with an angle of 120° between the legs (5 dbi). Since it is not rotatable it is included as a second reference.
  • 2-element shortened yagi, such as the Cushcraft XM240 or W6NL Moxon (6 dbi).
  • 3-element yagi shortened with loading coils and optimized for gain (7.5 dbi). A comparable commercial product is the M² 40M4DDLL
  • 3-element full-sized yagi (9 dbi). Many hams build their own since the several available commercial products are very expensive and difficult to ship.
  • 5-element switchable inverted vee yagi with an angle of 120° between the legs (10.5 dbi). I chose this as an upper-bound reference since it is the simplest antenna that exceeds the gain of a full-size 3-element yagi.
Antennas include matching networks where required and conductor loss (copper wire or aluminum tubing). Since the 2-element yagis of all types have a narrow gain band width I chose a 6 dbi reference gain for all of them, since that is more typical than the maximum value often quoted.

Not explicitly listed is a hybrid wire and tubing antenna from ON4UN's Low-Band DXing book. This is a design by NW3Z that resembles a Spiderbeam. The text claims a modelled gain of better than 7 dbi, but I was unable to achieve that in a NEC2 model despite care in construction of the model. I experimented with a few variations to boost gain. The best I achieved was 7 dbi and that took a physical design that is more fragile than that in the book. The model view of the antenna is adjacent. It appears it comparable to the XM240 or W6NL with respect to gain, F/B and SWR. That is, no better than a 2-element yagi.

Elevation angle

As in the past I will continue to use 10° as the target elevation angle for 40 meters (for example, see the ARRL Antenna Book) since it is the median for inter-continental DX paths, such as between VE3 and Europe. Longer paths have a lower median angle, sometimes as low as 1°. Although the gain figures will be lower at these angles the comparisons remain valid since all horizontal antennas behave similarly as the elevation angle approaches 0°.

Height vs. gain

First the chart with the modelling results and then my analysis of what it can tell us.


Notice how the gain of the yagis largely track each other with a change in height. There is more spread with two single element antennas: the inverted vee and short, rotatable dipole. In the latter case the near isotropic pattern in the vertical plane interacts more strongly with ground resulting in larger high angle lobes at certain heights which "steals" from the power in the main, low angle lobe.

The narrower main lobe of the yagis is more immune to this effect but it is still present for yagis with a low element count. The pattern shown at right is of the 3-element full size yagi at 40 meters height.

No matter the element count all the antennas with inverted vee elements suffer a sharper drop in gain as low heights. This is due to the lower average height of antenna current in the sloping legs. At 10 meters apex height the bottoms of the inverted vee elements are only 2.5 meters above the ground.

Nulls and domestic coverage

In the above pattern I marked the angle of the first and only major null. The null at 31° is significant. It reduces gain by -5db or more for elevation angles from 25° to 42°. That's a problem even if you are only interested in DX. For contests it's worse since the antenna will perform poorly for much of the US from my location (or within Europe for Europeans).

The null angle rises as the antenna height is lowered. That is why big contest stations either use a switchable stack on 40 meters or have at least one yagi at a lower height. This also allows for diversity by, for example, pointed the high yagi to Europe and the low one southwest through the US heartland. Unfortunately a height of 40 meters is not high enough to achieve good stacking gain with a large yagi for this band. The upper yagi ought to be at least 50 meters high and preferably 60 meters. This is extreme and well beyond my ambition. If you've ever visited a station with one of these stacks they are certainly impressive sights! If you do want at these heights you are better off using 2-element 40 meter yagis.

If the yagi is no more than 30 meters high it is possible to successfully cover both short and long paths with the one antenna. You need only be prepared to turn it a lot throughout the evening during contests. If you do go higher such as I intend it is necessary to have a low antenna as well. This could be as simple as a fixed dipole or inverted vee up 20 to 30 meters.

Note that for the inverted vee antennas, including the switchable yagis, the null angle will be higher by a few degrees. In the case of a 40 meter height the null is at 33°.

Diminishing returns, and a little calculus

As you go higher the main lobe points to a lower elevation angle. This is what we usually need for effective DX communication. Unfortunately the amount of additional gain we get for each height increment is not constant. There comes a point of diminishing returns. Where we draw that point is one of personal choice and circumstances. That is, for those who have the option to go higher.

I took the data for the previous chart and took its derivative. If you don't know calculus this is simply a rate-of-change calculation whereby we determine the rate of gain change for a change in height, for every height for every antenna is this survey. In calculus terms this can be written dG/dH. We'll be doing a discrete calculation, not develop a continuous function, since this is easier done and sufficient for our purposes. That is, numerical differentiation.

I did this by taking the gain at adjacent heights, which are at 2 meter intervals, taking the difference and dividing by 2. This is trivially accomplished with any spreadsheet software. Here is the result:

The curves are surprisingly complex. This appears to be due to the development of the pattern at high elevation angles, with cancellation resulting in more power in the main lobe(s) and reinforcement resulting in the opposite. The general trend is downward for all antennas.

As before the rate of change for the inverted vee antennas is large at low heights. Single element antenna are more variable since their free space pattern is close to isotropic in the vertical plane, which allows development of a strong upward lobe at heights that are an odd multiple of λ/4. The effect on yagis is subdued because their free space patterns have diminished gain directly upward and downward.

In my earlier height vs. gain article I stated that dG/dH ≈ 0.6 db/m for heights between 10 and 25 meters. While roughly correct the truth is more complicated. We can see that for yagis dG/dH is already below 0.4 db/m at 25 meters height. Going higher we see that dG/dH levels off just below 0.2 db/m.

This is useful information. If you are planning a 30 meter tower with an M² 40M4DDLL and you decide to pursue 1 db more gain you have two kinds of choices. One is to increase tower height to 35 meters. The other is to put a full-size 3-element yagi on the 30 meter tower. Actually you'll get 1.5 db additional gain in the latter case, and the first null is more compatible with domestic length paths. A third option of stacking smaller yagis can also work but will do little to improve low angle performance at this height.

If instead you want to choose between the 40M4DDLL and XM240 at 30 meters height you need to go 6 or 7 meters higher to compensate for the lower gain of the XM240. Since the gain bandwidth of the smaller yagi is narrow you should really aim for at least 10 more meters. These are good questions to ponder when building an antenna farm even if your antenna objectives are modest, such as a more typical tower height of 15 to 25 meters.

Terrain

Choosing an antenna and a height for it is not so simple as presented above. Terrain plays an important role. Modelling software treats the earth as an infinite flat plane of constant composition. While this allows consistency in antenna modelling it can severely misrepresent the true far-field pattern.

Serious planning benefits from terrain analysis. I have little to say about this other than to direct readers to tools such as HFTA. When combined with high-resolution topography data it is possible to achieve reasonable accuracy in the real far-field pattern for yagis at various heights and azimuth directions.

In my case I am unlikely to purchase a site based on HFTA. At best it will be of interest only after the fact. By choosing criteria for property that avoids certain negative attributes I expect to not have major difficulties. These criteria include avoiding valleys, bumpy and rocky terrain and the urban jungle. There are no mountains here that need to be avoided. Choosing a home and neighbourhood is not only about radio, unless it is only to be used during contests and not for daily living.

Terrain may be a greater factor when encountering the real world, yet when a location is selected the analysis will add colour to the terrain effects. Nevertheless the analysis in this article does provide useful insights into the benefits and risks of combinations of height and antenna type.

What to do about that tower

In light of the diminishing returns of height and its impact on DX vs. North American coverage it is reasonable to ask whether a tower of of 40 or more meters height is a good investment. A lower height would surely reduce cost, worry and maintenance without sacrificing much performance. So why go this high?

For non-competition operating a few extra decibels of antenna performance is difficult to justify. Some hams do it anyway if only because of the challenge or for the bragging rights. But few people are impressed by braggarts. Even for DXers the advantages are questionable. It may only come down to working that elusive P5 DXpeditions an hour later than you could with the maximum investment. Perhaps a larger, legal amplifier is all you need.

Contests are different. An additional decibel or two pays dividends. The band will open earlier and close later, and you'll reach the many modest-sized stations with greater consistency. If doing well in contests is your ambition it can be very worthwhile to take on the time and expense.

With respect to antenna choice there is something to be said for large wire yagis such as the 4 and 5-element inverted vee yagis presented. Since the bulk of contest QSOs are to be had from a small range of directions a fixed antenna is worth consideration, if combined with a more modest rotatable yagi. However such an antenna requires two tall towers, which will put these antennas out of reach for many. There is also an increased risk of antenna interactions, and thus on multi-op and SO2R operations.

We must keep in mind the matter of bandwidth. Only 2 elements on 40 meters is insufficient for optimum gain, F/B and SWR at more than a segment of the 40 meter band. A low SWR and F/B can be had with a Moxon style yagi such as that by W6NL, but it will not improve gain. Some may be satisfied with this compromise.

I will continue to mull it over. First I need to purchase the property and explore alternative tower and antenna combinations that fit the property. The first year I can make do with XM240 before deciding on a larger antenna. Readers can use the presented analysis to assist making their own choices to fit their circumstances.

Wednesday, April 6, 2016

Sometimes You Get Very Lucky

Since my next antenna article is a couple of days away I am taking the opportunity to add something to my previous article on working VK0EK on 40 meters. In that article I said,
"With my antennas I have almost zero chance on 80 or 160 meters so 40 is the lowest band I can expect to work them."
I am happy to report that I was wrong.


After dinner yesterday I scanned DX spots on my smart phone. Although it was still full daylight I was intrigued by a spot from a W1 reporting VK0EK on 80 meters CW. Since most of W1 is east of me their sunset occurs earlier, meaning that 80 opens there before it opens here.

However since some of W1 is only a few degrees longitude to the east, and a had a few minutes, I went down to the shack to check the band. The local QRN was subdued because it was 30 minutes before sunset and none of the neighbours had yet turned on their lighting systems. I listened around the spotted frequency of 3.534 MHz with a narrow filter to look for VK0EK.

I heard something indistinct so I switched to one of my 40 meter inverted vees. This is a common tactic to improve SNR when local QRN is present. It can't compare to a proper receiving antenna but will often work in a pinch since a horizontal antenna of any kind will reject vertically polarized radiation that is typical for locally-generated QRN.

Now I was able to get decent copy, though far from solid. VK0EK was in there calling CQ and getting no answers. Nothing ventured nothing gained so although they were very weak, my power and antenna are sub-par and the sun was shining I tuned the XIT to -1.4 kHz and sent my call sign at slow speed.

Nothing. So I tried again. Then again. That third try elicited a "VN" reply. I slowed down further and sent my call a few more times. Again I got that partial call response. Well, now I was really into it and I just kept calling. On their end a very patient and capable operator also persisted at the task. You need that dogged determination from both operators to push through these marginal situations.

On my end it was a bit arduous since I had to manually switch antennas at the end of every receive and transmit sequence. You don't want to accidentally hot switch with 100 watts!

After several transmit sequences and sending my call sign perhaps 20 times the operator on Heard Island sent back my full call followed by a high speed 599. I replied in kind and in some shock and amazement I logged the 80 meter CW QSO at 2308Z, 27 minutes before my sunset. About 3 minutes later it showed up on DXA (pictured above) to confirm that what just transpired was real.

I kept listening for a few minutes just for the pleasure of hearing them. This allowed me to hear a few familiar VE call signs also succeed in getting through. I don't know how long the opening lasted, only that checked again after sunset arrived I could weakly hear but no longer copy them. Then the lights came on and 80 meters was unavailable to me for the remainder of the evening.

A day later I am still astounded. A spotlight opening, a superb operator huddled inside a tent on a cold night on an uninhabited Antarctic island and me with my 100 watts and an inefficient tower vertical with a total of only 64 meters of radial wire (8 x 8 meters). I live for moments like these. The magic of radio is what finally lured me back to the hobby after a 20 year hiatus. This one QSO is a perfect example of that magic.

Saturday, April 2, 2016

Sometimes You Get Lucky

Several times in this blog I've talked about techniques for working DXpeditions with QRP and small antennas. The strategy required can be quite different from how the big guns go about it. But sometimes it feels good to toss aside the rule book and forget the script, no matter how poor your chances. You may still get through with luck alone. This is one story of this coming true.

I was fortunate to work each of the major DXpeditions currently in progress during their first day of operation. These are VK0EK on Heard and FT4JA on Juan de Nova. With the pressure of those first QSOs out of the way I was able to relax and consider my opportunities to add them on other bands over the remaining days they plan to be active.

Since the solar flux is so low and not forecast to rise before they QRT my thoughts turned to 40 meters. With my antennas I have almost zero chance on 80 or 160 meters so 40 is the lowest band I can expect to work them.

One evening last week I periodically entered the shack to check the signal strength of VK0EK on 40 meter CW to decide whether to bother calling. Since the DXpedition was still young the press of callers was not conducive to success with 100 watts and an inverted vee. At the very least their signal should be strong so that I can be picked out of the pile up on their end of the circuit. I cannot easily compete with kilowatts and yagis.

Yet their signal remained fairly weak that night. The cyclical QSB would sometimes bury them in my (suburban) noise and sometimes lift them to a respectable level. Before bedtime I decided to make a few calls, expecting failure.

Their down split range was narrow and the DQRM was fierce. I narrowed my filter to its narrowest to copy them. Since I was using XIT the filter setting made it difficult to find who they were calling so that I could optimize my transmit frequency. With little hope of getting through anyway I set my XIT to the middle range of where they seemed to be listening and left it there. To fill the time I did some internet browsing to catch up on the news. I kept one hand on the WinKeyer to press the button to send my call. That's about as lazy a pile-up technique you're likely to encounter.

From the arc of this story line you can guess what happened next. After approximately 10 minutes of calling I was shocked to hear my call hazily through the wall of noise followed by "599". I was in mild shock for several milliseconds before I reacted. Not quite believing it was me they were answering I resent my call followed by "599 TU". There are several VE3 DXers with similar calls I am occasionally confused with. That was my worry.

Their "TU" was perfectly synchronized to my transmission, as was their initial response. That gave me confidence that I was truly successful. The Heard Island DXpedition made this trivially easy to confirm since they upload their logs in real time. So I opened their DXA page with my browser and had only seconds to wait for the confirmation to appear.


The screen shot above is of my 15 meter QSO which took place a day or two later. I didn't think to capture the one for 40 meters since I did not have this article in mind at the time. You can see the 40 meter CW square coloured in.

While I don't recommend rolling the dice as a DXpedition strategy there is always a chance. Sometimes you get lucky.

Friday, April 1, 2016

King of the Munchkins

With the recent release of the 2015 CQWW SSB official scores I can now confirm that I have, for the second year in a row, won #1 QRP SOAB (single-op all-band, unassisted). This makes me King of the Munchkins. Or something like that.


QRP gets little respect among the majority of serious contesters. Reasons I've heard range from "life's too short for QRP" to "all that effort for such a small score?" For most, bigger equates to better. The pinnacle is always considered to be high power with a large antenna farm where you pit your skills against the elite ranks for the highest score in the world or continent. That and a good location, preferably tropical and a desirable multiplier. No matter your antennas or power, competing from VE3 is limited by propagation and prefix.

Ranking first in a QRP category is beneath the attention of most contesters is a mixed blessing. We're the short guys always getting underfoot, difficult to copy and get into the log, and contribute little to club scores. That's a shame since QRP exercises skills that often go underutilized when using higher power.

Some contesters with lots of experience and big stations try QRP and are seriously tried in getting a respectable score. The techniques required are quite different than when running higher power. You might expect that a big antenna farm will compensate, but that is only true to a point. There are 23 db between 5 watts and 1,000 watts. As impressive as those long boom yagis at dizzying heights might appear it will rarely make up the deficit.

From 807 to 4CX1500B

I see both sides. Until 2013 I had never operated QRP. The closest I came was an 807 transmitter in 1972 (my first rig) that may have put out 40 or more watts at times. I had no wattmeter so I am only guessing. Since then, as money and circumstances allowed, I strove to build bigger and higher antennas and run more and more power.

This trend culminated from 1985 to 1992 when I ran a small stack at 20+ meters height and a 40 meter wire yagi all powered by a Collins 30S-1. That is still modest but very good for a suburban lot. From 1980 to 1990 I also did multi-op contests from other stations, usually those with larger antenna farms.

When I went QRT in 1992 the 30S-1 (and everything else I didn't sell or give away) went into storage. The 30S-1 was sold about a decade later when a persistent Collins collector caught me at a weak moment.

Less than it seems

My claim to the Munchkin kingship is perhaps not as secure as I first claimed. Category ranking often has more to do with the vagaries of location and scoring. There are QRP contesters in the US northeast with bigger antennas and no less skill. Yet in CQWW I often outscore them. I cannot do so in the ARRL DX contests where the availability of QSO points between US and Canadian entrants is more equitable.

On the other side of the Atlantic the European QRP contesters are handicapped in CQWW by scoring 1 point for inter-country European QSOs while I get 2 points per US QSO. This rule might be overcompensation for their rich mine of nearby multipliers and daylight low-band QSOs.

All that said it is still pleasant to receive some recognition for doing my best with 5 watts and a modest antenna farm. My global ranking in CQWW CW is likely to rise this year because of fewer QRP operations from attractive multiplier locales. If the claimed scores hold I will place second behind GJ2A. We'll know for sure in a few weeks.

As far as all-time records, I only have one, and that is for Canada. I cannot compete with the solar cycle or the occasional QRP operation from a station with a large antenna farm. I hold the QRP SSB rate record for Canada, but in CW the record is held by VE6EX. There are many records from solar cycle 23 that could not be touched cycle 24.

QRO to QRP and back again

I've pretty much had my fill of QRP contesting. Looking forward I will be in the low power (100 watts) category in most contests for the next while. With the prospect of building a new station with a large antenna farm I expect to gravitate back to "bigger is better", but in stages rather than all at once.

While it's been fun operating QRP it is not a lifelong attachment. I expect the number of high contest rankings I've achieved to drop drastically even with bigger antennas. I have no illusions about my abilities when compared to the best in the world. Even so I expect to have a lot of fun, including more QSOs in the log. For me it's more about the fun and the station building, not mererly the score.

Although the 30S-1 is long gone I don't really miss it since it was not a good amplifier for contesting. Tuning was a nervous procedure due to the fragility of the power tetrode. Even with the plug compatible and more robust 4CX1500B in place rather than the original 4CX1000A it was not fun. Careful tuning was mandatory for every large frequency excursion, band change and antenna change. A simple error is all it takes to fry a grid with a power dissipation rating of 0 watts! It wasn't even all that great for DXing because of the 3 minute warm up it required.

My next kilowatt amplifier will most likely be no-tune solid state or use a grounded-grid triode. With the latter you can at least use dial markings to speed band changes and not panic over small mistakes.