Wednesday, October 29, 2014

Going QRP in CQ WW SSB

I had modest expectations from QRP in this past weekend's CQ WW SSB contest. This is the biggest of the worldwide contests with more that 10,000 participants. The QRM is fierce, as is the competition. QRP doesn't stand a chance. Or does it?

As planned this contest was to be a shakedown of my new antenna system. I did not go in with any competitive objectives, just to do the best I could and have fun. My objective was to see how much better I could do with the new antennas than in the other SSB contest I played in earlier this year: WPX. Typically I avoid SSB and SSB contests when running QRP.

I only operated half of the 48 hours possible, both for practical and personal reasons. So I could have done better, though probably not by a lot. At night it's the low bands that rule and those are the worst with QRP due to the high atmospheric and man-made QRN; it takes power to be heard.

In this article I'll talk about my experiences and suggestions on how to do well in an SSB contest with QRP. If you want to see how I did you can look at my claimed score submission on the 3830 page for this contest.

Station setup

From previous experience I made some small though important changes to my station setup. Compare this picture with the one from my previous setup, which had some problems.


The major change was to do what every contester has already discovered: slide the rig off to the side and put the computer front and centre. For too long I persisted in the belief that the rig must be the primary focus. This is simply wrong, especially during contests. Rig and rotator control is done with the left hand so I can tune and type while moving the yagi. Operating is otherwise hands-free since I'm using a headset and VOX.

Logging software is N1MM. I have the N1MM+ beta installed by I am not yet ready to make the switch. I use only a subset of the available features so this is sufficient for now. One thing I forgot to do was update the countries file, which messed up multiplier counting during the contest. I fixed that after the contest.

I spent some weeks with the current setup to accustom myself to operating the rig with my left hand (I'm right handed). The learning curve was fast. I could even type in calls while fine tuning the other station. The paddles (obviously not used in this contest) and computer mouse are within easy reach. With weeks of non-contest and QSO party practice I am now comfortable sending CW while adjusting the rig or typing on the computer. I purposely purchased an ultra-thin mouse pad so that the heavy Bencher paddles keeps the pad from sliding on the desktop without affecting paddle ergonomics. Even the sheaf of paper is partially held to the table top by the (paddle secured) mouse pad.

The antenna switch is positioned for easy access. This helps with diversity control (see below) by allowing fast switching between the yagi and inverted vee on the high bands. I replaced the round knob on the switch with the one you see. It is much easier on the hand for frequent use, as I do in contests. The benefit is more than you might guess.

The tuner off to the right side is for the 80 meters half sloper, which is cut for CW. I set it up so I could fine tune the match within the SSB segment using only the inductor. As it turns out it saw little use during the contest.

Feeding frenzy

When the contest starts every call you hear is a potential new QSO since no one has worked anyone yet. This creates a feeding frenzy, where the super-stations and rare multipliers dominate the bands and the QRM is fierce. Everyone is focussed on working anything that talks, either by S & P (search and pounce) or calling CQ. In this environment might is always right; if you have power you get through, and if you don't you won't.

On Friday evening I made a total of 25 QSOs. I sensibly stepped away from the radio after a time to wait out the feeding frenzy. There is little room in that environment for QRP. You can't fight it so don't try. It will only frustrate and demoralize you.

I knew the next day would deliver improved results. I got a good night's rest and set the alarm to catch the high bands opening to Europe and beyond early the next morning.

Getting high

No, not that kind of high. I mean frequency. If you looked at my detailed results (linked to above) you may have been surprised by how few contacts I made on 40 and 80, and none at all on 160. The thing is, other than working some multipliers it is perilous to spend much time on the low bands eking out contacts. Few can hear my QRP signal, even US stations, which means a very low QSO rate.

It is better to spend time at higher workable frequencies even if, it seems, the QSO totals on those bands are out of balance with the lower bands. With QRP you will generally do better the higher the frequency, if the band is open. Thus my overweight QSO count on 10 meters, and progressively fewer QSOs on each lower band.

Points are points, so unless the higher bands are closed you should avoid the low bands. I could have spent the hours between midnight and sunrise adding contacts on 40 and 80, but instead opted to sleep after working the easiest multipliers. If I were more competitive or using more power I would have spent more of the nighttime hours on the low bands.

S & P agility

Spending 95% or more of the contest doing S & P operating is fatiguing. Yet it is unavoidable for the QRP contester. Learn to enjoy it. It can be both productive and enjoyable with a little practice. This subject is worth a more detailed discussion.

First off, I did not operate assisted (DX spotting networks) and I don't recommend it for QRP. If you spend your time reacting to spots of needed multipliers you will almost always do worse. After all, many will respond to the spot and you have to fight through a pile-up. It is a fight you will lose. Worse, while you are there you are not calling workable stations.

Conversely, by spending your time S & P you will often run across multipliers that have not been spotted. I worked many this way. Some of these were double multipliers (zone and country), including: JV, KH2, ZD8, A7, among others. The only frustrating aspect of this is that they might not hear you even if there are no other callers. I missed many multipliers this way. After a few calls I give it up and keep tuning. Time is precious.

Another thing to avoid is spending too much time at the low end of the SSB band segments. The super-stations often camp out there, knowing that the bulk of contesters are S & P operators and that they tend to start their search at the bottom of the band. That means lots of high-power QRM that QRP cannot cut through. Therefore the stations with the big signals often won't hear you. Move higher. On 10 meters there was activity spread over 1 MHz of spectrum. My best results were clustered around 29 MHz. Lower-power stations are found there, and they are the ones most likely to hear you; that is, if you can hear them they can hear you.

At my fastest I could S & P contacts at a rate of 3 QSOs per minute. This only happens if the each station I tune is not a dupe and I get through on the first call. This isn't typical, but when it happens it's magical. There is a need for speed in locating each signal, tuning it in properly and typing the call into the log without lag time. When done right I can initiate my call within one second after he finishes his QSO solicitation, secure in the knowledge that it is not a dupe.

After reaching the top end of the band, or the top end of the contest-active spectrum, I will often change bands and repeat the process. When I finish that band I can flip back to the first. This maximizes results by allowing time for others to change their run frequencies and bands, ensuring that a larger percentage of the stations I find the next time I scan a band are not dupes. In any case, be prepared to do a lot of typing since later in the contest many or most of the stations you encounter will be dupes.

Synchronicity

If you do a lot of S & P you'll experience what I'm about to describe. Everyone doing S & P tends to follow a similar pattern of moving up (or down) the band from one running station to the next. It is not unusual to hear the same stations following your pattern (or, conversely, you're following their pattern). You'll hear them on most every station you call for some time.

This is not a problem, though I am amused when it does happen. Synchronicity of this type is not always amusing. There are many other contesters with similar calls, so when my S & P pattern synchronizes with another VE3 with a similar call confusion rules. This affects the calling station and my partner in synchronicity. When this occurs I break my pattern by making a big step up or down (in the same direction of the pattern) to get away from my synchronous partner. I cover the frequency gap later.

A contributing factor to the situation is the use of partial call databases. A feature of contest logging software is to match a partial or full call against a database of known contester call signs scraped from logs submitted to various contests. For those of us with weak QRP signals this sometimes results in the running station guessing at your call from what the database search window provides and asking you to confirm their guess. When used by poor operators this feature is a plague since they aren't listening to you, but rather going through a list. It can be very frustrating. There are about 4 call signs in particular that the database spits up that the poor operator on the other side keeps trying to get me to confirm. That's a clear sign of what they're doing. I don't see a solution to this problem. So be aware of it.

The "Run"

While going about the usual S & P method of working stations I will occasionally call CQ. I only do it at the edges of band activity where the smaller guns tend to hang out, and if I feel confident that the frequency is clear enough and the conditions good enough that I have a chance of being heard well. Usually I get nothing and move on after a minute or two.

I do this for a couple of reasons. First, most casual operators and little gun contesters only S & P. Calling CQ is the only way to work them. Although you would never guess from scouring the bands, these stations are in the majority. Second, if a run can be started the rate can far exceed the best that S & P can offer. The problem of course is that QRP makes running very difficult. If I get more than 2 or 3 responses I'm doing well.

Sunday morning of the contest 10 meters conditions were superb and over 1 MHz of the band was filled with contesting stations. The higher the frequency the less the QRM and the smaller the guns. When I reached the outer edge of activity and had a nice clear spot I called CQ. The frequency was 29.163 MHz, one that I may never forget.

After a few CQs I got an answer. Then another. Within a few minutes later I had several stations calling me at once. The pile-up never got deep but it kept going. Almost all were Europeans, and some were multipliers. When the pile-up got thicker I had to fall back into an operating mode I hadn't used in 25 years. Like riding a bicycle it seems you never forget how to do it. There are techniques to keeping the rate high and keeping the interest of those awaiting their turn. It was a lot of fun. I was smiling as I operated, thoroughly amazed that I was doing this with QRP.

Eventually the QSOs trailed off and I moved onward. Checking my log after the contest I found my run was 113 QSOs in 52 minutes, for an average rate of 130/hour. That's nothing for a station with power and bigger antennas, but it is phenomenal for my station. I'd never guess it was even possible.

The day after the contest I searched the logs of the global spotting networks and found that I'd been spotted 4 times during that run. That likely helped. Many contesters in "assisted" class use software that allows them to identify spots of stations they haven't worked and with one click can QSY and auto-fill the log. After they work me they move on to the next spot.

Antenna diversity

Apart from 40 meters capability the multi-band inverted vee was intended for diversity in contests. By this I mean the ability to work stations off the main beam of the tri-band yagi without having to waste time turning the yagi. Unless it's a needed and workable multiplier it is often not worth the time.

Based on testing prior to the contest I had doubts about my strategy since the yagi works so much better than the inverted vee, sometimes even off the back. With the experience of CQ WW SSB I concluded that the value of the strategy is mixed. Sometimes it works and other times not so much.

When it didn't work to my advantage I would make a note of the unworked stations in a particular area of the world (e.g. the Caribbean). When I had a few of them I would turn the yagi and worked them all in sequence. If conditions in that direction were good I would keep the yagi in that direction for a while longer before turning it back to where it was. The usual direction for the yagi most of the day was towards Europe.

Aggressiveness

The smaller the station the more aggressive the operator must be to be successful. Does that sound odd? If it does, think about it. With a big signal you can call a station and know that you will get through, most likely on the very first call. If you call CQ you will start a run, even if there's adjacent QRM. There is little need for an aggressive stance, other than muscling others aside to hold a run frequency. But if you are a little gun some aggressiveness in your contest operating pays dividends.

I should explain what I mean by aggressiveness. I don't mean rude or obstructive tactics, distorting your signal, or other types of poor behaviour. What I do mean is the drive, even the hunger to put QSOs in the log and to not take 'no' for an answer. Like a short child in a crowd of tall adults you need to assert yourself to be noticed. Too many operators of small stations are timid; they are tentative in calling, slow at completing a QSO and moving on to the next, and have low expectations.

Moving quickly and pushing your call through the QRM and pile-ups is not a crime. Sure, you have to be realistic at times and move on without scoring that rare multiplier. But make that a temporary setback. Note the frequency, note the propagation pattern and make a plan to come back and finally log it, perhaps at your or their sunrise or sunset. Even on the more typical QSOs send your call again, again, and then again.

Most stations call once and wait, so you should call again if there's any delay in the called station's response. Be quick about it. On your second call the other stations will mostly be silent, so you can be heard. Put some urgency into your voice to get their attention. Make them want to work you.

Even if no one else is calling and the CQ machine starts up again after your call you should not give up. Try again. Then again. They probably do hear you but are reluctant to make the effort to pull you out of the mud. Don't let go, but don't waste time either. Make them feel guilty about ignoring you. This isn't being rude or inconveniencing them. They probably need the points more than you do since the larger stations are usually more competitive.

Vary your frequency a bit to attract attention. Speak (or send) more slowly or change up your phonetics. Show them you really want the contact and that you're willing to put in the effort to make their job of copying you easier.

Aggressiveness works. Not always, but it does work. Although you are a little pistol you can and should hold your head high and strive to be noticed. Your score will reflect your confidence. This is true even if you're not in it to win, but just for the DXing or other award opportunities.

Next up

The experience of CQ WW SSB tells me that the yagi is working as well as I expected. Even without the intention of winning my category (SOAB QRP) and operating only half the available hours I appear to have put in a competitive score. I will be very interested to see how I place after the log checking is done and the official results are published next year.

My November contest plans are CW Sweepstakes this coming weekend and then CQ WW CW. I will likely skip SSB Sweepstakes. I will be QRP in both, also with the intention of doing well and not so much focussed on winning. For the present I contest for fun, not to win. That is enough to motivate me. My inclination is to only be competitive in a multi-op effort from a large station, something I haven't done for many years. There will be time for that if I choose to do so.

Before CQ WW CW I hope to build and test a sloper for 40 meters with the intention of achieving the modelled low-angle gain. If it works out I will consider the 2-element version. However that will run into some issues with backyard geometry if I point it to Europe, the most productive direction for DX.

Apart from contests I will continue DXing. The large number of DXpeditions in October have kept me hopping. My overall and per-band DXCC QRP totals are growing apace. After this past weekend even my SSB DXCC count has breezed well past the 100 mark.

Wednesday, October 22, 2014

Building and Using the 80 Meters Loaded Half-Sloper

The prospect of operating QRP on 80 meters is not particularly enticing to me, especially considering my primary interests of DX and contests, even if I stick with CW. Although I need to do it for at least contest operating I had been putting off building and installing the half sloper design I described in an article 6 week ago.

The procrastination ended this past weekend. Two hours in the workshop and then a rushed hour of outdoor work (rain was imminent) was all it took. The morning temperature was a cool 5° C, which is surprisingly comfortable for doing tower work. That is, in calm air.

The antenna works even though it'll never win a prize for exceptional performance. You can refer back to the September article (linked to above) for design details. This article is about construction and my initial experience with the antenna.

Measurements

The modelled length of the half-sloper wire is approximately 17.8 meters, including a 1.25 meters long section going outward horizontally from the tower (due to NEC constraints). With a loading coil of 17.5 μH, medium conductivity ground, the tower and yagi the result is modelled to resonate at 3.550 MHz with an SWR of about 2 (low radiation resistance). However the model is certain to be inaccurate to some degree due to a variety of factors, including: ground, environment, acute angle between wire and tower, and exact contribution of tower width, guys and yagi.

As initially constructed the actual measurements are as follows:
  • Top of wire is 14 meters up, or about 10 cm below the top of the tower.
  • Coax terminates at 14 meters height, with 14 AWG insulated wires from an SO-239. Wire length to the tower connection is ~30 cm, and ~60 cm to the sloping wire.
  • The sloping wire is 12 AWG insulated wire. The top section length is 11.5 meters and the bottom section length is 6.0 meters, with the loading coil between them.
Feed point

Mechanical design of the feed point is shown in the picture below. The wires from the SO-239 are green and the sloping wire (hanging vertically) is black. The set of tower X-braces is the very top of the tower, just below the top bearing plate. You can see the right edge of the egg insulator that terminates the top of the sloper wire. It is closely roped to the tower leg behind the mast.

Yes, it's ugly. After considering constructing a more aesthetically pleasing feed point I decided to keep it simple, yet functional and effective. Even if it stays for a year it is an experimental antenna not worthy of extra effort to make it look pretty.


Sealing putty (used Coax-Seal) protects wire lugs and the SO-239 backside from water incursion. An ordinary electrician's twist connector connects the sloper wire. A screw lug on a strip of sheet aluminum connects to a tower brace with a hose clamp. All hardware is stainless steel.

The longer wire to the sloper can be trimmed as necessary, or the sloper wire can be trimmed at the bottom end. In its present lengthy format the wire to the sloper wire is going all over the place, including pressing up against the mast. This will be cleaned up before the snow flies, as will the final waterproofing of the coax connector.

The potential between all conductors is low throughout the mess of wires so there is no risk of flash-over or erratic tuning. There is no common mode choke on the transmission line since the outer shield will in any case couple to the full length of the tower. A 1:1 current balun to choke common mode current can be used where the coax exits the tower if induced noise or RF in the shack are present. For the present I will go without one.

Loading coil

Per the design, the loading coil is ~65% down the wire, with 11.5 meters above it and 6.0 meters below it as first cut (see above). I used a 17.5 μH coil I picked up in a flea market, which I bought since it looked to be about the right value, and the length-to-diameter ratio is consistent with a low ESR (equivalent series resistance). The inductance is calculated since my inductance meter only resolves to 10 μH.


As shown in the picture I used a dog bone insulator to take the strain of the wire tension; there is no tension on the coil itself. Screw loops and (soldered) connections to the coil are made from solid12 AWG copper, to keep the insulator centred within the coil. The loops permit the use of #12 stainless steel hardware, rather than soldering, to join the wires and coil.

The result is ugly but effective. It's also light enough that the wire sags very little even with low tension.

Anchor

After coming down the tower I walked the antenna towards Europe (northeast) until the bottom end lifted off the ground. With a brick and nylon rope I temporarily anchored the antenna at the edge of my lawn.

The wire end is ~40 cm above the ground. Based on the wire length the calculated interior angle between wire and tower is 37°. This is 8° less than the 45° maximum I will allow to ensure no negative interactions with the yagi.

The temporary anchor will be replaced by a pole that will be anchored to the retaining wall. This will increase the wire angle a degree or two and alter tuning, but is absolutely required for safety.

Safety!

Speaking of safety there is a risk to life with this antenna, and I don't just mean tripping over the wire and falling into the neighbour's rose bushes. The end of the sloper wire and, very importantly, the tower bottom carry high RF voltages when transmitting with more than QRP. With a kilowatt there can be a risk of severe RF burns or even electrocution.

If the tower is grounded the risk at that point is partially mitigated, depending on the quality (impedance) of the ground connection. However you should always assume that the tower is a serious safety hazard at 100 watts or more. Be prudent and take measures to limit the danger.

The end of the sloper wire has no mitigating factors, so the RF potential can be very high. Do not leave any conductor exposed at the termination. Let me tell you a true story about how serious the danger can be.

Back around 1986 I had an 80 meters half sloper (full size) on my 20 meters high tower, also oriented towards Europe. Since I was not so smart in my younger years I tied the bottom of the sloper wire with string to a cinder block anchor positioned adjacent to a steel shed. That placement just happened to give the best match. To improve the match I later added a foot of insulated wire to the end and just let it rest across the string and block.

That night I got on 80, switched on the Collins 30S1 and did some DXing on CW. Minutes later the SWR shot up. Since the 4CX1500B is a delicate tube I reduced drive and did some testing. At low power it worked better but would misbehave at more than a few hundred watts.

The next morning I went outside to investigate. The end of the extra length of wire had drifted close to the painted steel of the shed. There must have been some fireworks since a 10 cm square section of the shed had the paint burned off and soot radiated outward. The wire was a carbonized mess. I had inadvertently created an RF arc welder.

Tuning and match

As originally cut the 2:1 SWR bandwidth of the antenna is ~180 kHz and dips to 1.1 at 3.450 MHz. This is 100 kHz lower than planned. This is either good planning or good luck since this is equivalent to 1 meter of wire length, and I had added an extra 0.8 meters for tuning (trimming).

The minimum SWR and SWR bandwidth are better than the EZNEC model. This does not surprise me. It has been my experience that half sloper of this type, with the wire end so low, tend to match well to 50 Ω systems.

I suspect the reason for the differences is loss in the ground and adjacent buildings that is not encompassed by the model. That is, an additional series loss resistance raises the feed point impedance close to 50 Ω, and that also lowers antenna Q. The additional loss isn't good but the match is, perhaps, some compensation. There are limits to what can be achieved with a simple low-band antenna on a suburban lot.

On the other hand the high Q and narrow bandwidth could be due in part to pushing NEC2 to its limits, as briefly discussed in the design article. This could mean that environmental losses might not be too bad. Unfortunately I don't have a way to properly test either hypothesis, except by on-the-air results.

The antenna can be trimmed at the bottom or by trimming the mess of wires at the feed point. In the former case it can help to raise the antenna off the ground. I opted for a bit of both: taking half off the top and the rest off the more-accessible bottom. If you like, a coarse and fine adjustment, respectively. However it does not do to make all the changes at once (including a permanent, safe-to-walk-under anchor) since any change affects the geometry. The change in wire angle and distance from ground can affect resonance and match more than a change in wire length. My approach is to get the anchor properly positioned and only then adjust the wire as needed.

Final tuning was not done as of the publishing of this article. It will be routine. For operation higher in the band in CQ WW SSB this weekend I will use a tuner. Tuner and transmission line loss will likely be no worse than -1 db. Alternatively, a short section of wire at the bottom of the antenna can be manually switched in and out to choose between CW and SSB segments of 80.

Interactions

I spent some time in previous articles discussing the possibility of interactions among antennas, a serious concern with so many wires antennas and yagi in close proximity. The interaction model I built with EZNEC helps identify problem areas, and how to avoid them. Now that the 80 meters half sloper is installed you can see the mess of wires around the tower in the adjacent, annotated photograph.

The actual distance between wires is not obvious since there is a loss of 3D perspective in the photo. The north upper guy wire is about 2 meters from the end of the 40 meters element of the multi-band inverted vee, and it is a similar distance from there to the half sloper wire. Looks bad, doesn't it?

In actuality there is no measurable interaction between antennas, with respect to SWR or (so far as I can tell) pattern. This is in accord with the modelling experiments I did. Provided that the antennas are no more than mildly resonant with respect to others and the crossing angle is far off the parallel the mutual coupling is small.

This is good enough for my purposes but not necessarily for others. Coupling to the outer surface of coaxial cables can bring RF and local noise sources into the shack and receivers. If you use more than one radio at the same time, whether in a contest or for weak-signal DXing, these types of interactions should be considered and solved with common mode chokes placed at strategic points.

On the air

Just my luck that by the evening of the day I installed the antenna a long period of geomagnetic stormy weather set in. US stations were heard but little in the way of DX. I listened to a few weak German stations during the WAG contest, and the large US contest stations trying to work them. There was no hope for me. It takes a DX opening to properly test the antenna so I waited several days until the ionospheric absorption relented enough for DXing to become possible.

In the interim I measured the SWR and tested antenna interactions. I did not tune the antenna until after making some QSOs. The SWR of 1.5 at the bottom end of the CW segment was good enough to begin operating.

On Tuesday I made my first QSO with the antenna: W1AW/8 in WV. I called a weak LZ station who was having a sunrise enhancement but he got only part of my call. But that was very promising, telling me that the antenna has possibilities even under poor conditions. Remember than I'm only running 10 watts.

My objective for CQ WW SSB this weekend is more modest: to make some short-haul QSOs to bump up my score. I will continue trying my luck at CW DX as 80 meters conditions permit.

Wednesday, October 15, 2014

Two Weeks with a Yagi

Of course a yagi at 15 meters height is going to perform better than an inverted vee at a similar apex height. In that respect what more could possibly be said other than the recently-raised Explorer 14 works great? I think there is more to be said, even if the news is less than revolutionary.

This is especially true when QRP is brought into the mix. Before talking more about the yagi's performance in my QRP station I think it would help to review the several distinct philosophies regarding QRP:
  1. Striving to do the most with the least. Everything is small, not just power, but aim for maximum results.
  2. Low impact amateur radio.
  3. Portability and emergency use.
  4. Level playing field, except for power. Otherwise, operate like the majority.
This diversity of approaches to QRP struck home for me in a recent exchange on a contest forum. There are those who operate QRP contests and there are those who operate contests with QRP. I fall into the latter category. Further, my QRP philosophy is that of the last bullet in the above list (though I briefly started off at #2 when I returned to the hobby almost 2 years ago). Thus my drive toward larger, better antennas in my current domestic situation. Which brings me back to the yagi.

As I hinted in my first experiences with the new yagi, it is a clear winner over the inverted vee. While the fact alone is obvious I wanted to quantify it. My original estimate was a broadside improvement of about 10 db. This is a combination of forward gain and height. Height contributes two ways: lower radiation angle, and rising above the local pattern-modifying, metal-rich suburban houses.

Accurate A-B comparisons are difficult due to Faraday rotation, path-dependent elevation angles on every signal, and other effects. These create time-dependent gain differentials and pattern sensitivities. Even without the ability to properly quantify the improvement, the testing I've done over the previous two weeks convinces me that my expectation of a 10 db improvement has been borne out.

My observations and results of my first 2 weeks with the yagi comprise the rest of this article. My next article will get back to antennas. In particular the first of the low-band wires I've been planning.

DXCC

When I put up the yagi 2 weeks ago I had 207 countries (QRP & CW, starting in January 2013). I now have 215. Good propagation and a number of DXpeditions helped. Actually only 6 of the 8 new ones were accomplished with the yagi. I first worked VK9DLX (Lord Howe I.) on 30 and YJ0X (Vanuatu) on 40. The inverted vee is still pulling its weight.

I am finally having more success with long path DX: western Pacific, Indian Ocean and east/southeast Asia. But it wasn't just new countries. I added many more band-countries. Even casual DXing is becoming more casual since more stations can copy me well.

Pile-ups

QRP in a pile-up remains challenging, though less so than before. I can mix it up in the pile-ups. If they're not too intense I can get through just fine. The past 2 weeks these included C21GC, S79KB and T30D.

Contests

I contest-tested the antenna in the California QSO Party. For the 9 hours I operated I put in a decent score. I had substantially more success that in the Florida QSO party earlier this year. Few were the stations that didn't hear me -- for some reason (perhaps QRM at the other end) those were mostly on 10 meters. This also goes for SSB, which I didn't even bother with in FQP.

My next test of the yagi is CQ WW that is coming up later this month. Although I was briefly tempted to accept an invitation to join a multi-op operation from a large station.

Diversity

For contests my plan was to supplement the yagi with the inverted vee on 20 and 15 for the purpose of diversity. That is, to catch openings and quickly work stations off the main beam direction of the yagi. For example, in late afternoon Europe is good for running stations on 20 meters, but there are also openings to needed multipliers in Asia and the Pacific.

That might not work out. The yagi is so much better than the inverted vee that other than off the sides the yagi outperforms it. Even with the F/B of the beam, signals off the back are usually no worse than on the inverted vee. I will have to take this into account during contests.

Wind

My unconventional guying arrangement (trees) necessarily includes a measure of uncertainty. The trees must not only support themselves in the wind but also the lateral and vertical (upward) force of the guys due to the same wind. This is a topic I want to explore further in a future article.

The tower survived a brutal wind storm, before the yagi was raised. My station was right in the cross-hairs of the greatest sustained wind (90 to 95 kph) and gusts (over 120 kph). It was hard to watch, yet I did. The tower performed admirably. Or perhaps I should say the trees performed admirably.

The question is whether the added load of the yagi would have had the same successful outcome. My calculations say 'yes' but I can't be certain. This week we had some moderate winds with gusts up to 70 kph. I monitored the changing tension on the most windward guy to gauge the stress placed on the tree anchors.
There were no surprises.

WARC bands

I now know that I am missing on 17 meters by only having the inverted vee. If I had a yagi to cover 17 and 12 I could boost my success. But a compromise was necessary, and I opted for a tri-band yagi for best performance on the contesting bands rather than more modest performance on all bands from 20 to 10. Since I don't care too much for 12 meters it is only 17 meters where I notice the lack of gain. I can live with that for the next year.

Saturday, October 4, 2014

Pile-ups: Zig vs. Zag

Getting through the larger DX pile-ups with QRP and small antennas can be challenging. This is even more true on 40 meters and below where the atmospheric QRN makes it difficult for the other station to copy you even without the competition of a pile-up.

Sometimes it's better to just step away from the radio and not waste the time trying. But if you have some time to burn it doesn't hurt to try. I faced this choice with ZD9XF on 40 meters, after having worked him on 30, 17 and 10. An inverted vee and 10 watts versus the pile-up is not a fair fight!

For those willing to make the effort this is the story of how one QRP'er did, eventually, get through. Since I am no longer so focussed on the absolute need to work every country on every band I was able to look at the problem with some dispassionate analysis. That is, I don't have to work him, I just want to work him.

Be realistic

Even if I were the only station calling ZD9XF on a clear frequency the QSO would be marginal at best. I've been working QRP for long enough to gauge my chances. When you add to that the depth and breadth of a pile-up on a rare DX station the little guy's chances can dwindle to nothingness.

Only when you realistically assess your chances and the competition can you make a plan. Sure, you can just jump in and call anyway, if you have time to waste. However if you want to actually make the QSO you will benefit from a plan based on honest self assessment.

Pattern

Good DXpedition CW operators establish a pattern designed to not only make it easy for them to work more stations but also to help out those in the chase. They need to encourage callers to spread out, creating enough separation so that their receiver filtering can, ideally, isolate one signal. When callers are bunched together the QSO rate can dramatically drop, to the detriment of everyone.

The typical pattern is for the DX to keep shifting his or her listening frequency by a small amount on every QSO. This can range from as little as 100 Hz up to several times that. Perhaps more frequently they'll drift higher until a signal can be picked out, either by its strength or lack of immediately adjacent signals. Eventually they'll reverse direction or jump to the lower end of the listening range, again moving upward.

Learning the pattern can be helpful, though not always. If too many callers adapt to the pattern it no longer works since too many callers will be packed tightly together. As part of learning the pattern it is important to further discover how the DX breaks the pattern. This can involve a sudden frequency shift or even a period of randomness. Either way you need to learn the pattern the DX station follows. That is the basis of all that follows.

Zig and zag

The caller's response to the DX station's pattern is to zig and zag. Here is what I mean by these terms:
  • Zig: You QSY after every QSO according to the observed pattern. For example, when the DX QSYs up 200 Hz after every QSO you QSY the same amount. If your signal is competitive with respect to the other callers, or if they are simply less observant, you make the QSO. Repeat as necessary. When the pattern changes, so do you. If the pile-up is successfully zigging, you should occasionally do a big zig. Per the above example, QSY up by more than the pattern typically exhibits. This is in the hope or expectation (by observation) that the DX tries to out-manoeuver the horde.
  • Zag: When the pile-up is on to the DX station's pattern and you cannot compete by signal strength alone it makes sense to anticipate pattern changes. For example, at some point the DX will reverse direction or abruptly QSY to the other end of the range. So instead of zigging you do the opposite of (you hope) of everyone else. This I call zagging. For example, if the DX QSYs up 200 Hz after every QSO you QSY down 200 Hz. Most of your calls will be fruitless but when the pattern reverses you are there and (hopefully) few others are. Similarly, you can plant yourself at the other end of the range and lie in wait for the abrupt QSY.
Whether you zig or zag it is important that you listen carefully to the frequency of every successful caller. That is how you establish the pattern, and changes to the pattern. If you can't hear the other station that is a sign that you should zig (or zag) regardless, as if the caller was at the predicted frequency, or the pattern has reversed or shifted. Go ahead and call anyway but you need to find out where the DX is listening now so that you can have a chance.

Luck = Chance + Planning

Zigging and zagging are all well and good, and is well-established practice among DXers. But if your signal is uncompetitive you are going to miss out unless you try something different. The reason is that the listening range is going to be full of callers who neither zig nor zag, but camp on any old frequency and call away. They often will do so during each QSO and even when the DX is transmitting!

Sure, you can laugh at their futile behaviour but that will not help you. Your QRP signal is weak and will be covered up by their blather. In other words, as ridiculous as it seems they have a better chance of working the DX than you. After all, regardless of the DX station's pattern they will occasionally be listening on the frequency where these stations have camped. You lose.

Consider what it will take for you to make the QSO with a marginal and uncompetitive signal:
  • Time: You need to be calling at the time the DX is listening. It may seem obvious yet many don't do this. You know, you've heard (and cursed) them.
  • Frequency: The frequency you call on must not only be where the DX is listening but must also be reasonably free of other callers. If there's even one other caller near your frequency you will almost always lose. And then then DX station QSYs for the next QSO.
  • Attraction: Even if you succeed at time and frequency if your signal is unattractive you will still lose. For example, if you are sending at 30 WPM you will likely be passed over. Copying high speed CW when you are so weak is very difficult. Do not send faster than the other station, and consider going quite a bit slower. A pile-up is not a CW speed competition. Once you get the operator's attention you may want to slow down further to make his or her job easier. You hold their attention by at least trying to make their job easier.
All of this will still fail if the other operator is averse to weak signals. You may just have to wait for better conditions or for when the competition is spent. The latter occurs when the DXpedition has gone on for a time, after all the big guns and even most of the little pistols have made their QSO. If the DXpedition is a short one, well, be prepared for disappointment.

Getting through a pile-up is more like poker than a lottery. It's a stochastic process composed of random variables and more deterministic factors. For little pistols the weighting of the random variables is higher than it is for others. It is therefore important to bolster the factors under your control so that when the chance comes you are ready.

The contact

When my chance came I was fully aware of it. I also knew if it didn't pan out I might not get another since the clock was rapidly running down on this DXpedition, as was the time I could devote to it. It did indeed turn out this was the last evening ZD9XF was active on 40.

The operator was sending more slowly than previous nights, perhaps in recognition that all the stations with larger antennas had been worked so signals would be weaker in both directions. The pile-up was still deep and wide but less so than before. It was under similar conditions that I worked FT5ZM on 40 so I reasoned that I had a chance.

The chance didn't appear to be working for me. After an hour of calling I got nowhere. There were just too many others calling for me to be heard. I even tried the strategy of ignoring the pattern and camping out on a frequency in his listening range that was mostly clear of other callers. That is not only time consuming it failed since when his listening frequency approached my camping spot the ziggers and zaggers covered me up anyway.

What finally happened is that during a descending frequency pattern the DX went lower down that before. When he worked a station with a small split of only 1.4 kHz I noticed that the frequency was otherwise quiet. And so I saw my chance. I QSY'd down to a 1.2 kHz split and called. The frequency was clear except for me. But would he keep tuning down or, more likely, QSY up or simply not hear me?

He sent "R3". I took it as a hopeful sign and called, knowing that if I was wrong I was unlikely to be interfering with a potentially real R3 station. Then I heard "R3FN". The turnover timing between my transmission and his was further evidence that it might be me he was hearing. I slowed the keyer and got "K3VN" in reply. That's when I was certain I had his attention. I slowed all the way down to 17 wpm and sent my call twice. That succeeded and with a quick "599 TU" he was in the log. This was later confirmed in his online log.

In the grand scheme of things this was no great accomplishment. What it did was to demonstrate that a combination of chance and skill can achieve results for small pistols with DX ambitions. If it can work for me it can work for others.

Skill

No, not mine! This contact would not have been possible if not for the skill and dedication that G3TXF put into his operation, including making the effort to pull out and work the weakest of signals.

It takes two to QSO, so always remember that. Tell them "thank you" when you have the opportunity.

Wednesday, October 1, 2014

Raising a Yagi Onto a Guyed Tower

Ready to rock 'n roll on the high bands
This past weekend I devoted a day to raising a tri-band yagi onto my DMX-52 tower. As with everything else in my present station I accomplished this feat with no outside help. The system I designed for raising the antenna would not have materially benefitted from the presence of a helper.

The antenna in its final mounted position is show in the photo at right. The antenna is a Hy-gain Explorer 14. It's slightly larger than the type of antenna I had planned on putting on this tower. It weighs 45 lb (20 kg) and has a boom length of ~14 ft (4.3 meters). It can be lifted and manoeuvered by one person with some care in planning and safety procedures.

Considering its benefits of higher-performance and broadband match over a fully-trapped 3-element tri-bander I was willing to go for it. There is also the matter of chance in that I had a limited shopping period which limited my selection to what came available on the used market.

Tram

Super-stations have the art of lifting large yagis onto high towers down to an almost routine procedure. Some sort of tram is required to lift and guide yagis around and over guy wires. Trams, though simple in concept, can be a nightmare to build and operate by the inexperienced. The alternative of lifting the yagi in pieces is rarely recommended since it can be a difficult and dangerous procedure, even for small yagis.

A tram was required to lift this yagi. That the tower and antenna are small is no escape from this requirement. I had several planning choices to make.
  • Element or boom first: In most tram systems the antenna is lifted with the elements pointed toward the tower top. This is necessary since the elements must fit around the tower and mast. Pointing the boom toward the tower is mechanically easier since the yagi bottom can rest on the ground along the back element and no cradle is needed to protect the elements from ground contact. But unless the yagi is lifted over the tower and mast it isn't possible for the yagi to approach the tower this way (you may need to visualize this in 3D if you are unfamiliar with the procedure).
  • Starting position: With 3 guys there are 3 areas between guys from which to lift the yagi. The best one is free of overhead obstructions and has suitable anchor points for the rigging.
  • Yagi to be lifted through this gap. Rigging visible at tower top.
  • Pulley and rope system: The pulley should be above the point where the yagi will be mounted, and leave enough room for the boom sling between mounting point and pulley. As with a gin pole it is necessary to consider the lateral and vertical forces on the tower, mast, pulley and rope since, during the lift, the yagi will be hanging some distance from the tower. A way to quickly and safely lock the lift rope is highly recommended.
I chose to lift the yagi from the south side of the tower. There are tree branches to be dealt with up to about 10 meters. They must also be contended with on the other two sides, though to a lesser degree. What sealed the deal for me was the ability to utilize the trees that anchor the guy wires to also anchor the rigging. Pretty much by definition they are in precisely the best location and of suitable strength.

Since the boom is 14' long and the elements are as long as 31' I made the sensible choice of lifting the yagi with the elements pointing up. There is less potential of tangling with trees and guys wires. The stress on the rig and tower is reduced since the yagi doesn't have to be moved out too far (laterally) from the tower. Unlike with the simple gin pole the tower, mast, rotator and northward guys are sufficiently strong to resist the lateral force of the offset yagi's weight due to the tram.

Complete rig with tree limb obstacles
The pulley is secured by a muffler clamp at the top of the mast by threading it over one side of the U-bolt . Rope cleats are positioned on both sides of the tower for flexibility during the operation. I believe that every tower should have a cleat permanently attached. If you've never used one before you'll wonder how you ever got by without one when you use it for the first time.

The procedure

First, I apologize that there is a dearth of pictures to show the steps in the lifting procedure. I had planned to take more. But when it came to take a picture or deal with the serious matter at hand I chose the latter. So I'll explain using a diagram superimposed on the picture of the rigging I designed and constructed.

The rig (or tram, if you prefer) is comprised of two ropes secured to the tower top (see picture above) and the ground. The ground attachment is a rope with anchor loops running between the trees I use for guy anchors. It is designed so that it is can be easily collapsed for the early steps of the procedure and then pulled tight to support the antenna. It is in the shape of a narrow triangle: 15 meters high and 5 meters wide at the bottom. This is similar to the type of tram that Hy-gain describes in some of their literature.

Step 1: The fully-assembled antenna is placed at the bottom of the tower. It is upside-down and the boom clamp (blue) faces outward. The tower goes in the widest gap between elements, in this case the driven element and reflectors. The rigging at this point is fully collapsed, with ropes stowed under the antenna.

Step 2: The antenna is lifted to a height that just exceeds the length of the longest half-element, or about 16' (5 meters).

Step 3: Climb up the short distance to the boom and rotate the antenna (counter-clockwise in the diagram) until the elements are not quite pointed straight up. This is to ensure the elements do not get inside the guy wires when the antenna is lifted.

At this point one of my neighbours from down the street came running over. He wanted to know if I needed help to lift the yagi. I chatted with him for a while, explaining the procedure and why it would best if he just watched. My plan was solid and would only be complicated by inexperienced hands. I do not want an injured neighbour on my conscience. For the most part my neighbours are neutral to actually supportive of my hobby, which might be unusual!

Step 4: The rigging is reassembled and pulled to not-quite full tension, pulling the antenna outward from the tower. I ensured that one side of the tram was inside the 10 meters reflector and the other side between the director and driven element. A couple of branches of the large spruce (visible in the picture part of the diagram) had to be guided around the reflectors and boom. This involved a combination of bending the branches and pushing the antenna around them.

Step 5 (not on the diagram): Lift the antenna until the upper end of the elements fully clears the lower set of guys (guy station at 23' height).

Step 6 (not on the diagram): Pull the rigging to its full tension and continue lifting to the top of the tram. The added tension is needed for this step since the tram is closer to the upper guys than the lower guys. The bottom halves of the antenna's outer elements ride over the upper guys as well as the tram in this step.

The yagi left dangling while I took a lunch break
With the antenna lifted to its maximum height (where the boom sling touches the pulley) I broke for lunch. It was 2.5 hours of hard work and I was hungry. The yagi was left dangling and slowly swinging in the wind, within the small range of movement allowed by the sling (see picture at right). I was unconcerned with leaving it since I knew it wasn't going anywhere. The wind was near calm all day, so perfect for this job.

It wasn't only hunger that drove me to take a break. All that work will tire anyone. As a rule I will not approach a tower or other demanding work without being physically and mentally alert. An hour's break made sure that I was ready for the next set of tasks.

As detailed below in "what went wrong" there were some unanticipated tasks to be done. Although it was annoying I did leave ample time in the schedule for that sort of thing, and I knew the fair weather would continue throughout the day.

The remaining tasks are in point form, since there are so many of them (and perhaps more than you care to read about):
  • Rotate the yagi back to horizontal, but this time right side up. This was easily accomplished now that the yagi was well above every obstruction.
  • Position the boom-to-mast clamp and tighten it down. I had placed the clamp on the mast some days earlier so, again, this was quick.
  • The 4 bolts to secure the boom clamp to the boom-to-mast clamp (threads pre-greased) were installed. The bolts are Grade 5, not the weaker stainless ones that came with the antenna. This took a little longer since the clamp faces had to be almost perfectly parallel and the holes lined up to drive the 5" bolts through. The most difficult was the second bolt. After that they went in easy.
  • Lift the coax and temporarily hook it up to the antenna. Antenna direction was set towards Europe.
  • Test the antenna in the shack across all 3 bands. I made a few QSOs, including an OD5 with a small pile-up on 15 SSB. SWR and gain comparison with the inverted vee were good.
  • Seal the coax connector and wire lugs on the beta match and driven element.
  • Increase slack in the coax in preparation for lifting the yagi. At this time the coax is held up by only its lift rope.
  • Untie the boom sling and remove the pulley from the mast.
  • Loosen the boom-to-mast clamp and push the antenna up to its planned height. It is bolted tight without (yet) worrying about its direction.
  • Fit the plastic cap to the top of the mast to keep water from getting inside of the mast.
  • Loosen the rotator clamp and turn the antenna north, which is the centre of rotation and where I'd earlier positioned it. Tighten the clamp and test by pulling/pushing on the boom.
  • Untie and drop all ropes (except the one holding up the coax) to the ground.
The final step was to dress the cables down the tower and towards the shack. I hauled up two runs of RG-213: one for the yagi and one for the (future) 80 meters half sloper. This saves the trouble of dressing cables twice. By supper time the yagi was operational, fully rotatable, the cables dressed and the rigging and tools were put away. I left the yard and tower clean and tidy. There is a good feeling that comes with accomplishment of a large job done well.

What went wrong

Can you admit an error? Acknowledging and investigating mistakes is at the very core of achieving future improvement. This is done when a airplane crashes, a train explodes or someone is injured or killed in a work-related accident. Tower work is no different. I made some mistakes and underestimated the difficulties with some steps in the antenna-raising procedure. Let's use my experience as a learning exercise.
  • Rigging: The rope rigging was constructed the day before antenna raising. It was all carefully measured and tested, ready for the big day. I then separated the sections to stow the ropes out of the way for safety. The next day I put it all back together (see Step 4 above), paying special attention to the marks I made on the rope to properly place them. But I didn't check the reconstituted rigging before proceeding to lift the yagi. It was only afterward that I noticed the rigging was laterally offset by 1.5 meters. I had misinterpretted one of the rope marks. That explained why the yagi kept skewing to one side (and increased tangling with tree branches) during lifting. Luckily this mistake didn't cause any serious difficulties beyond costing some time.
  • Boom sling: In the picture of the feed system below you can see the thick rope of the boom sling. The sling must allow easy rotation of the boom during the lifting operation yet not encourage unintentional rotation. My first sling used thinner rope. When I lifted the antenna (see Step 3 above) the thin rope could not be easily manipulated. Down went the yagi and on went the new sling with thicker rope. The second one worked fine.
  • Snags: To avoid snagging of antenna hardware (especially tube clamps) with the rope rigging or guys I made sure that all the hardware on the underside of the antenna (the side that would rest on those lines) was rotated to avoid snags. For example, the open end of hose clamps. Snags occurred despite my precaution. The pressure of the antenna on those lines made it possible for even the tapered trap end caps to snag. The 1" of boom protruding past the outside elements and those element-to-boom clamps also snagged on the rigging, guys and tree branches. This cost me time but no serious problems. The heavier the antenna the worse the problem of snags.
  • Boom bolts: The boom is in two pieces which are secured in the boom clamp (see the picture below). The bolts of the clamp, and the boom-to-mast clamp squeeze the boom sections to hold them tight. There is also one bolt per boom half that pierces the boom and clamp for added security. You can see the head of one of these bolts peeking around the left side of the boom-to-mast clamp. I installed these on the ground with the heads on the other side of the boom. The bottoms of the bolts protruded far enough to prevent connection of the boom-to-mast clamp. I had to remove these tight-fitting bolts at the top of the tower and reverse their direction. That was not fun. I should have checked this on the ground.
  • Tilt: When mounted and tightened onto the mast the elements were not horizontal, even though they were all in the same plane. The tilt was only a few degrees, which although it does not affect performance it is unsightly. I wanted the cleanest appearance so that neighbours would wonder why the antenna was crooked. Again, I should have checked this on the ground. I added a few washers between the boom clamp and boom-to-mast clamp to correct the problem.
  • Final lift: After completing antenna testing I temporarily secured the antenna to the mast. I then removed the boom sling and the mast pulley. I then needed to lift the antenna ~50 cm (20") to its final position at just under 15 meters above ground. This small difference has negligible performance impact but does minimize interaction with the top ends of (future) low-band wires antennas. The 45 lb antenna is easy to lift on the ground and even up the tower. It is less easy when you are perched at the top of the tower and somehow have to push it upward, hold it in position and tighten the bolts of the boom-to-mast clamp. The pulley and sling had to be removed since there was no room for them in this operation. It took some minutes but I finally found a way to brace my feet, bend my knees and push it up with my shoulder. Once lifted in this way it was (relatively) easy to tighten the bolts. It worked but was too ad hoc for my liking. In the past I have usually worked on larger towers on which this procedure is easier.
  • Cable dressing: The coax is dressed to the boom and tower with a combination of vinyl tape and UV-resistant cable ties. The ties are for security and the tape is to spread the pressure and weight of the dressing to avoid crimping the coax. The cable ties were too short for the task. I corrected the error by chaining together two cable ties.
Modified antenna feed

The stock Explorer 14 comes with a BN86 balun: coax goes in one end and the other (balanced) end connects to the driven element and beta match. But this is a voltage balun which is not the best choice to choke common mode current on the coax outer shield.

I removed the BN86 and made a not-quite-as-bad choke by coiling the coax. There are 9 turns of 4.25" diameter. It is an acceptable choke though not ideal for the 20-15-10 range.

Modified feed for the Explorer 14 yagi

The picture shows the incomplete feed system atop the tower, temporarily connected for on-air testing. The rigging is still attached since it would be needed if a problem required the antenna to be moved. It worked fine. I completed the feed system by installing a plastic block (one side concave) under the aluminum "L" with the SO-239 (quickly made in my workshop) and securing it to the boom. This is necessary to isolate the coax from the boom to ensure good electrical balance.

The rubber boot (a discarded bicycle tube) over the PL-259 is for weather protection. Unlike most yagis the Explorer 14 has the driven element closer to the director than reflector so the yagi points to the right of the picture. The slack section of coax to allow rotation (sometimes called a drip loop) is placed to the left of the boom-to-mast clamp to keep it away from any antenna elements.

Performance results

The benefit of the yagi over the inverted vee was immediately apparent. I plan to spend some time comparing them over different band and paths so that I can better quantify their respective attributes. Small stations don't often have the ability to compare antennas so this is an opportunity to learn. What follow are my observations from the first 48 hours with the yagi.

Seemingly quiet bands like 10 and 15 came alive with signals with the yagi. In a way it's subtle and sneaks up on you. It's like travelling from the city to country at night and suddenly the sky is full of stars. In particular 10 meters became a far more interesting band. Europeans poured in each morning. It took some time but I eventually cracked the ZD9XF pile-up on 10 CW. On the inverted vee he was barely audible over the local QRN, and I wouldn't have bothered with my 10 watt signal. Polar paths were poor the last two days but from the signals I did hear it bodes well.

SWR across the bands is mostly very good. It is low in the CW segments and slightly higher at the top of the bands, which is the opposite of what the manufacturer specifies. On 15 the SWR rises well above 2 higher than 21.3 MHz but still works well. SWR is below 2 across 20 and 10. Gain seems good everywhere but is difficult to quantify, as gain always is. F/B ranged from good to fair.

I am looking toward contests and future DX opportunities with tremendous optimism. I should be able to improve my contest placing in the QRP category. DX pile-ups suddenly seem far more crackable. The neighbours for the most part seem comfortable with the look of the tower plus yagi, which is a welcome bonus.

Next come the low-band wires for 40 and 80 which I'd like to get in place over the next 3 weeks.