Book I'd like for Christmas...

The more I operate WSPR, the more fascinating I find the curious appearances of isolated spots, often from great DX distances, at times when general propagation is completely closed.

Oi!  Down here, Santa!

For the most part, this seems to be driven by Auroral Es, especially clearly seen in the dead of winter at 14MHz.

Searching for some research in this field, I came across the very well-written and eye-wateringly expensive The High Latitude Ionosphere and Its Effects on Radio Propagation.  Though now quite old, it's nevertheless a rare source of information on this niche subject, making copious references to the heady days of nuclear weapons research done in the 1960s.

I always have a queue of radio things to buy, so this one will have to await kind donations at Christmas, I think!

The curious case of...ZL/K6KWI

WSPR always throws up something unexpected.  Last night, it was a New Zealand station (with the great callsign of ZL/K6KWI - indicating Neil's Kiwi origins) at 14MHz, throwing out a very high, 50W signal.

The strange thing was that I was the only station who heard him around that time (18:20UT) outisde the US and Antarctica, despite the very high output.  Oddly, though it appears on the map, I can find no reference at all to the HB9 spot in the WSPR database, either in TX or RX in relation to Neil's station:

The terminator is correct for the time of the spot.  I am not sure as to the path taken, as ZL is very near the precise antipodean point to the UK, and the geometry of the greyline seems to make either (or both) paths just as likely.  I tend to think it was a path involving antipodean focusing, which was just about enough to enhance the signal to detection limits (I heard him at -26dB).

I didn't hear Neil again until 06:52UT this morning, where a few other stations in Europe were also now hearing him.

Certainly something to look out for in coming days, if the very high winds of up to 130km/h over the next 24 hours permit!

Magnetic loop - 40m WSPR results

Well, nobody can claim I am not thorough!  After a very favourable outcome at 14MHz for my car-portable magnetic loop mounted at 2.5m base height, I had a similarly successful result at 10MHz.

How would this small loop, 3.9m total perimeter, in 15mm pipe, do on 7MHz?  A loop of somewhat larger perimeter and 28mm pipe has done extremely well here in the past, but I didn't expect so much from a loop primarily designed to fit in a car.

Matching is via my usual pseudo-gamma ('wire up one leg') system, which provides a combination of induction, direct coupling and loop mode energising of the main loop.  It all goes through a 4:1 balun, although a 2:1 or 1:1 balun - or no balun - will usually also work.  SWR at resonance was 1.05:1, done with a SARK 110 analyser.

A lot smaller, and infinitely easier to deploy than a wire antenna at 7MHz.

Well, I always choose to put my antennas to the test against the best-performing WSPR stations to be found in the UK.  At 7MHz, M0PAI was the comparison station.

Here's the distance plot for both stations.  Note the very clear dips at post-sunset and pre-sunrise grey line periods.  I've never seen such a clear example before, which was aided by some of the quietest geomagnetic conditions, even at high latitudes, I've seen for a while:

Across all distances, the magnetic loop is only 8dB down on what I think is a full-sized G5RV at 7MHz:

This difference reduces to just 6dB when looking at DX distances, noting there were not many simultaneous DX spots.

Also note that the DX situation is an average.  Looking at the plot of spots, the highlighted station was, at that moment, hearing my loop 1dB above M0PAI:

Once again, the message is clear: magnetic loops work, and can compare very favourably for their size and ease of deployment, against full-sized wire antennas.  This is especially true when one can deploy a loop in a high quality environment.

It's also interesting to note that, even though it makes no allowance for environment, the 6-8dB difference is exactly in line with the efficiency of the loop predicted by this online calculator, which yields a 7.5dB reduction from 100% efficiency, yielding an overall efficiency for the loop of 18%.  This is not the same comparison as testing the antenna itself, of course, as there are various radiation pattern effects and environmental gain to take into account for both the loop and wire antenna.

IMAGES - ESPECIALLY FOR PE4BAS!

Not quite radio,but...

...you may well like to have a look at the Credo website, or have a look at the short video below.

This fascinating, scientific use of your mobile phone is an app to detect cosmic ray showers.  It is fairly efficient in terms of resource use, and even runs on my pretty basic, low-storage capacity smartphone.

Geomagnetic anomalies, 30m

My single turn magnetic loop has been out in the strong and very cold easterly winds over the past few days, doing very well on 14MHz.

Yesterday morning, I decided to re-tune the loop to 30m and see where the 200mW WSPR signal got to.

I am not sure what antenna other people are using at 10MHz, but my loop is doing better than any others I can find from the UK, by quite a big margin.  This is surprising, as a 10MHz antenna is not so large - just 7m per side as a dipole, or 7m tall as a vertical - as to preclude most people from having a full-sized, efficient radiator.

It's too early for a full analysis of spots yet.  But I did note that geomagnetic disturbances overnight generated interesting propagation enhancements throughout the dead of night.  Curiously, as the early morning period advanced, a periodicity between spots of an hour, or multiple of an hour, seemed to become apparent.  I will have to think about that one, but this paper is an interesting place to start, not least because the proposed modulation period occurs in agreement with my spots:

Magnetometer stackplot, 19-20/11/2018, Norwegian line.

Standing the test of time...

A quick review of antenna durability for the late edition post today.  I thought this would be useful, because there are plenty of antenna designs out there, but very few that report back some years later, to say what is left of them!

Yesterday, during rather cold weather, I decided to do some maintenance on my 6m 2-element quad.  This was unashamedly based on this web site, way back in around 2012.  Since then, it's been a superb antenna for me - so much so that I have always resisted the temptation to buy an all-aluminium quad to replace it.

I immediately knew that making the antenna out of timber would not last very long in our windy, wet environment.  So I made a spider support from aluminium 90 degree pieces, and (eventually) the arms from spare fibreglass elements from a 7m fishing pole.

The good news is that, if you make the antenna as I did, plus a few minor alterations, then this will last you many years without trouble.  Mine is going into its seventh year now, still with the original components.  The feedpoint has had to be redone a few times, but only because I used TV 75Ohm coax with an aluminium outer for the 1/4 wave stub; this reduced to dust in our salty air in no time.  It now has an all-copper, heavy-duty coax stub.

Originally, for a short while, I recycled a Pfeiffer, 15m quad experimental frame in PVC tube to hold out circular wire loops for the 6m, 2-ele quad.  Of course, it worked very well, with maximum enclosed area.  One day, I hope I can try the linear-loaded Pfeiffer quad at 15m again, because, as you can see, the saving in size is quite considerable:

Early days of my 6m quad.

The trouble with the PVC is that is has a high wind loading area.  Worse, the glued 'tee' pieces very quickly deteriorate under UV light, to the point that they fail in about two years or so outdoors.  Despite this, you see an awful lot of US-based antenna designs using PVC tube.  My advice: avoid it completely!  And for intolerant neighbourhoods (my neighbours know to leave me alone!), that bright white plastic is not very stealthy!

I then made the quad look like this:

Second iteration.

Since this photo was taken in early 2013, the only change I've made is moving the coax down from the boom (which then stupidly ran down the lower driven element spreader again) so that it simply feeds directly the same lower corner (horizontal polarisation) from the mast itself.  This reduced the interaction between coax and antenna considerably, and reduced the length of feedline by nearly 3m.  You can see the arrangement in the last image of this post.

Here's the quad spider detail.  Tape and cable-ties were OK for those nearly 7 years, provided one changes the cable ties every two years or so, but the top arms slip through eventually.

What you  need to do - and what I was doing yesterday, it cut notches with a rotary tool in the 'L'-shaped aluminium, so that the cable tie (or, preferably, stainless steel hose clamps), make better contact with the fibreglass arm.  Typically, the arms are slightly smaller than the depth of the 'L' bed, so there is only slight contact between them and the securing fastener if you don't make notches.

Also, it's a good idea, especially for the topmost bracket, to drill a hole and fix a stainless steel bolt, to physically prevent the upper spider arm from slipping down.  You can do this for all arms if you like, as the tension of the antenna wire does tend to pull all of them inwards.

Also cover the exhaust clamps in plenty of grease to make them last longer.  In an ideal world, I would suggest you use stainless clamps for a square aluminium boom, not a round one.  This avoids slippage and rotation as one could get with round tube, though it has not been a problem for me.

When you think this antenna has gone through nearly 7 years of very strong sunlight and even stronger winds, then, for the peanuts it cost to build, it really is a true winner!

Serviced and ready for another several years.

Figure-of-eight vs. standard magnetic loop.

Many days of data are now in from WSPR transmissions from my 'Meight' (figure-of-eight) loop and a standard, square loop.

Now, the figure-of-eight loop, for reasons of not having to spend money because I had the tube to make the loop already in the shack, and for its ease of forming into shape, is made of 10mm soft copper tube.  The standard, square loop is 15mm diameter tube.

The difference in tube diameter can be expected, if both loops were the same design, to yield a +9% advantage in favour of the wider tube.

And, after many days, we do indeed find, even by eye, that the single loop significantly outperforms the figure-of-eight loop.  A comparison with MX0PHX shows that, whilst the figure-of-eight loop was running (shaded area, left), it was only very occasionally able to outperform that station.  On the switchover to the standard loop, we find MX0PHX is usually under performing the loop:

Indeed, the single loop did admirably well against the UK's typically top-performing 14MHz WSPRer, GI8YJV:

I can't say whether a figure-of-eight loop would perform better or worse than a standard square loop of essentially the same total perimeter, because the 15mm tube, compared to 10mm, does make a difference to efficiency.

But what I can say is that a 15mm square loop is considerably easier to make than a figure-of-eight loop.  My own view is that I would prefer to have a wider-bore standard loop of, say 22mm pipe in practice (the increase from 15 to 22m gives roughly +7% more efficiency).  A figure-of-eight loop in 22mm pipe might slightly outperform the single loop (but of course, I don't yet know), but it would become quite top-heavy to the point of necessitating a more substantial mounting system to tolerate strong winds.  A permanently-located, stout, treated timber mast of about 4m height would do the job well, easily and fairly cheaply.

One other thing in concluding is that I realised the figure-of-eight loop needs significantly more tuning capacitance to bring it to resonance than the standard loop.  The answer as to whether or not this affects overall antenna efficiency is something that will have to await another day. The tuning range of the single loop is also significantly different, tuning about 7MHz up, from the figure-of-eight loop, which tunes from roughly 10MHz up.

All change!

The figure of eight loop, as earlier reported, has done very well in WSPR performance, and showed itself able to tolerate strong winds up to, so far, about 110km/h.  Tuning is stable in all weather - a satisfying vindication of my simple, very cheap approach to remote tuning).

Having gathered about a month's worth of data with the 10mm pipe, figure-of-eight loop (total tube perimeter of 3.67m), I decided I might as well see how a single loop of 15mm tube, 3.9m total perimeter (1m per side, minus 10cm gap where the cap sits) compares from exactly the same location and height (noting the single loop base is the same height, but its top is only half as high as the top of the 'meight' loop).

Single loop test.

In fact, I had stopped running the 'meight' loop a day or so previously.  I had been running my vertical delta loop at 1W from the rig overnight (shaded period in the graph below).  At 08:06UT, I switched over to the single magnetic loop running 200mW.  The outcome is surprsing:

Of course, my delta loop might well have seen its spots graph increase after this period; it usually slightly outperforms GI8YJV most of the daytime.  But that doesn't really matter, because the magloop running 200mW is certainly reaching as far as the delta running 1W, and doing as well at this moment as the full trapped wire dipole at GI8YJV, who is one of or often the best WSPR performer on 14MHz from the UK.

At the time of writing, as it's possible to see on the extreme right, the magloop has started - and has so far maintained - a significant distance advantage over the wire dipole.

I will now continue to run the magloop for at least several days, to see how it compares to that figure-of-eight loop, which is of almost identical perimeter (though, again, the single loop is of 15mm, not 10mm pipe as in the case of the 'meight').

Other than that, this Sunday, it's time to go flying (using VHF AM radio). 

Image: Wiki Commons.

Propagation anomaly, 2018 November 15

Last evening's Norwegian line geomagnetic stackplot looked like this, showing a relatively minor disturbance in the early evening, and some much smaller disturbances later in the morning:

The last that was heard of WJ1I anywhere other than the US was at 21:02UT - except for here at MW1CFN, where a single reception was made at 01:20UT (now reverted to the vertical delta loop antenna):

The fact that only one US station made it across the Atlantic at this time, and that he was not then heard again, shows there must be very localised and extremely transient propagation changes related to the passing of the auroral oval at mid-Atlantic longitudes.  Again, the anomalous spot comes at the time of field restoration.

60m - Wow!

It's been a bit quiet here at Copper Mountain HQ of late.  Normal operations have been largely suspended in favour of long-term beaconing on WSPR, using the figure-of-eight loop.

But, as the commercial FM radio blurted out the latest shouting about Brexit from the circus that is the UK Parliament, I turned over to 60m FT8.

Busy on 60m this evening!

I was amazed to get an immediate reply to my 'CQ' from BH4IGO.  Wow!  China on 60m.  That's a first for me.

60m is really coming alive now, as more and more countries approve allocations for their amateur population, and the weaker higher bands propagation encourages its use.

Certainly, it's worth trying a few calls around sunrise and sunset, so why not give it a go? 

LED Security Light - no RFI!

In the past, so-called security lights were incandescent units with bulbs that typically blew up after a few months.  Due to the cost of the whole lamp being close to the cost of the bulb, most people would not bother replacing the light, leaving it to rust on the side of the house.

But now, we have LED lights!  These offer about the same light of a 250W incandescent unit in a package that is small, lightweight and typically only uses 10W.


Of course, as someone involved with astronomy, I would very much encourage you to fit a PIR-operated unit, and ensure it is pointed downwards, where it won't uselessly light up the sky, disturb neighbours, and cause glare that diminishes, not aids visibility.

For us hams, LED lights are yet another potential source of RFI, often from unfiltered mains connections that allow the switch mode PSUs in these lights to inject interference directly into our wiring.

No detectable RFI!

I bought a new LED unit earlier this week, for a demonstration of light pollution at the Snowdonia National Park headquarters.  It cost me just £8.   Even better, it produces no discernible RFI on any of the ham bands.  All the others I've tested have been so incredibly RFI-noisy that I put them in the bin immediately.

If you want one of these in the UK, you can find them here.  Elsewhere, you can probably find the same thing from the product reference number on the box pictured above, as they are mass-market things likely sold under a plethora different names.

Rough!

On the night of 04/05 November, I really, really couldn't sleep.  The kind of night where minutes turn into hours.  Of course, the following morning, I had to be up early and drive 200km (and back!) to take the practical component of my aviation radiotelephony examination!

Lots of coffee and even more chocolate biscuits ensured I passed my RT exam.  But over on the amateur bands, things had become very disturbed indeed.

Whilst the geomagnetic field wasn't energised to a very high level (maximum G2, Kp=6), it was disturbed for a very long time - lasting more than 24 hours.

Lots of disturbance!  Image: IRF, Kiruna.

I've noticed in the past that, like last night, during disturbed magnetic conditions, I tend not to sleep very well.  This is not due to auroral light, because often, the activity is not such that aurora occur over Wales.  I've also noticed that the sleepless nights don't seem to happen when I am in Iceland, directly under the auroral oval.   It's fair to say that science has (a) failed to investigate human magnetoreception in any coherent, sustained manner and (b) most scientists dismiss it.  I remain of the view that other mammals and non-mammals do exhibit the sense, and that therefore, humans might also posses it.

Anyway, the effects on HF were quite pronounced.  Far from killing propagation, it again, as I've noted very many times in the past, led to continued spots of my 200mW from the figure of eight magloop currently under test here.

So instead of 14MHz propagation falling off a cliff at around 20:00UT as it consistently has over the past week under quiet conditions, it carried on supporting spots throughout the night, albeit to a limited number of stations, but which did usefully include TF1A.

As the plot below shows very beautifully, TF1A was, for the past week, hearing me in a consistent pattern that centred clearly around the middle of the day.  You can see that, as disturbed conditions commenced late on November 04, TF1A hears me only intermittently in the daytime, but continues to hear me during the evening period, shutting down just before midnight.

But as the disturbed conditions continue into November 05 and 06, you can see how TF1A now does not hear me at all during the day, but then hears me clearly, and for the sustained period of 3 hours, starting just before (23:30UT) midnight.

High resolution detail of the propagation to TF1A is shown below:

Looking at all spots of my signal, I was heard 175 times from 23:18UT to 03:22UT - a period where no spots were seen in the previous, geomagnetically quiet week (typical latest spot ~20UT, earliest ~05UT), as shown below:

Another magloop comparison.

I was interested and mildly amused to read G0KYA's blog post about his latest antenna improvements this morning.

As you can see from the blog, G0KYA expresses delight at being heard around the globe with his 66' (20m) long Windom.  I don't know whether or not there was a conscious decision to omit the power output used for the WSPR test, but at the time I accessed the blog (08:00, 03/11/2018), it certainly wasn't there.

Of course, it's easy to look up the WSPR database and find out what the output was.  So I did.

This reveals that G0KYA was using 37dBm.  That used to be quite a common output in the days before WSPRlite, when QRPp levels of 200mW or less became the de facto standard.

This plot proves my claim on power output used by G0KYA:

Of course, with such a high power output, there is no surprise at all that G0KYA's signal was making it across the globe at 14MHz.  Unfortunately, this doesn't really provide as much useful information as using a lower power does.  For one thing, because most people are using 200mW now, meaningful comparisons with most other operators when you are using 5W is quite difficult.  In fact, even at 200mW, it's difficult to find enough people who operate long enough, or have information published about their antenna systems to make proper, long-term comparisons.

So, let's now look at how my magnetic loop results - just 200mW - compare with G0KYA's 20m-long Windom at 5W.  The plot includes all distances for simultaneous spots:

The result is that, for an output power difference in favour of G0KYA of +14dB, the difference in received signal in his favour - averaged across all distances - is only +8.5dB.

Across DX (>5010km) distances, the difference in favour of G0KYA (+15dB) becomes more in line with the output power difference (+14dB):

And of course, as a result, G0KYA makes it further across the globe:

So I can confidently say that, when power output differences are normalised, my magloop - just two rings of 73cm diameter at 2.5m base height above ground, is, for shorter haul skips, doing better than a 20m-long Windom in an inverted-V sort of configuration, and matches it across DX distances.

I'm glad to see that G0KYA says he's interested in magloops under his QRZ.com page entry (accessed 10:24am, 03/11/2018).  This comparison would suggest it may well be worth him abandoning wire antennas in favour of a well-made magloop!

And if you didn't catch my video of wind tolerance, posted under construction details for the magloop, here it is again for you: