17m Surprise! (Updated)

A nice, warm summer's evening last night.  Another visit to the copper mine with my grab-and-go portable kit, this time for 17m, was justified, especially as the 'mini-bog' site had been soaked overnight with heavy rain.

All it takes to be effective.  You won't geneally see these advertised, of course, because nobody can make money out of them (but a few try!)

 

I had a few nice QSOs with FT8 and SSB on 3W and 5W, respectively.

Quite amazing to see how effective QRP can be from a good site: reports of FT8 activity, mostly RX of my 3W:

I also ran just two rounds of WSPR (at 1W), which returned a surprising number of spots, given that 17m often has very few WSPR users these days.  Conditions seem to have enticed people out of their slumber:

So, if you continuously read material online or in magazines that you need a £1500 rig, a tower, Yagi and a £2000 amplifier to be a success in amateur radio - you really don't!  A QRP rig and some junk wire, carefully made into a resonant vertical and deployed on a good site, is all it takes.

UPDATE.

The following evening, I undertook some evaluation of the 'mini-bog' site compared to my 17m vertical delta loop, just a few hundred metres away.  First, the map of TX from the mini-bog:

Spots of my 1W WSPR at 18MHz from the copper mountain hill 'mini-bog'.

And, second, the spots from the vertical delta at home:

Now, whilst the two maps do not look very different, there are in fact very important differences.  First of all, just by looking, you can see that the vertical loop is not getting as far west into the US as the 'mini-bog' site.  It also has more spots from Europe, which it to be expected due to screening by higher ground immediately to the east of the copper mine site.

Good stations like K5XL and KFS simply couldn't hear the delta, whilst the 'mini-bog' site was managing it quite comfortably (median at KFS for 1/4 wave vertical was -18dB)  That implies a difference of at least 16dB between the two sites, given the detection limit is -34dB.  As we'll see in a moment, this kind of enhancement is backed up by spots actually heard, not simply not heard.

VE6JY heard the copper mine site at -8dB.  But from the delta at home, it was only -23dB; a 15dB difference.  Always good to put that into context, because it magically turns 1W into nearly 32W.  Very useful, when you are QRP!

The only close match was from KA7OEI-1, which only returned a 3dB enhancement for the copper mine site, compared to the delta back home.

Overall, of nine of the best DX heard, no less than five did not hear my delta at all in the brief test period (17:54-18:18UT)

And, finally, some SSB heard from far-away Arizona at what is, for the conditions and general achievement of most stations at the time, quite amazing signal strength, albeit with occasional, strong QSB (suggestive of very low angles, I would expect):

Brendan Prize: could it all become a mess?

Almost by chance this morning, I came across a forum post on the QRZ.com page, which is a part of the internet I generally avoid for its suffusion by angry, opinionated white men with a very large axe to grind.

The claim made was that the Brendan Prize was available for a QSO between "Europe and the U.S.", a claim that is clearly incorrect.  

The prize rules are, from the IRTS web site itself that the Brendan award is:

'to each of the operators of the two amateur radio stations which first establish two-way communication in the relevant category between the continents of Europe and America (North or South) within the Two Metre Amateur Band [a],'

and 

 
'The two stations involved must be located on land or non-tidal waterways within the continental shelves of Europe and America as defined [b].  Note that the limit of the continental shelf of Europe is deemed to lie along the line of maximum depth between the European land mass and Iceland, while that of North America is defined to lie along the line of maximum depth between Canada and Greenland.'

Note [b]refers to the Times Atlas as the reference for where the continental shelves lie.

This all seems perfectly clear enough.  But this week, John, EI7GL, reports and provides excellent recordings by others of strong Greenland radio signal reception in Ireland.  Over coming days, a fairly persistent pattern of weather will provide for potentially good tropo conditions between Ireland, the UK and Norway to the north Atlantic regions of Iceland and Greenland.  There is at least some chance of a prize-winning QSO.

This all set me thinking about the precise meaning of words for the Brendan rules, and why those words might have come into being.

Let's take it one step at a time, starting with the Iceland-Europe bit.  The boundary is to:

"lie along the line of maximum depth between the European land mass and Iceland"

Here's a map to help us figure-out where this "maximum depth" point might be: 

Hmm.  We see that there is a problem of sorts.  The "on land/non-tidal waterways" bit of the Eurasian plate is all that bit in orange on the right.  The "maximum depth" between the Eurasian continental shelf and Iceland is, erm, well, we can say a bit to the right of the mid-Atlantic ridge and to the SE of Iceland, roughly between Iceland and Scotland.  And/or we can say that much deeper bit between Norway and Icleand.

But geology comes to highlight the arbitrainess of this Brendan rule.  Scotland goes under water to the north, doesn't go very deep, then rises again as the Faroe islands.  You can fairly say, in both geological and bathymetric terms, that there is a barely-submerged continuum - and not by any stretch of the imagination the "maximum depth" - extending from the North Sea part of the Eurasian Plate, projecting towards and including Iceland - or at least the eastern half of Iceland, which is part of the Eurasian side of the mid-Atlantic ridge - a geologically divided country.

Part of the division between continents near Reykjavik.  Eurasian plate on the left, NA plate on the right.  (C) MW1CFN.

 

In other words, there is no actual division, in anything other than political or surface-map terms, between Iceland and Europe.  In ever other way, they are one.

We can of course take things as written, and identify the actual, deepest part between Europe and Iceland, ignoring the fact that eastern Iceland sits on the Eurasian plate. It's a bit difficult to get an absolute answer, but 'eyeballing it' clearly shows that it's either the Aegir ridge part between Norway and Iceland, or the deepest parts of the mid-Atlantic ridge somewhere south of Iceland and between Europe.

Would it be fair to say this results in this kind of boundary?  If not this kind of line, then it would have to be the deepest part between Eurasia and NA, which isn't the definition used.  I'm being extremely generous in my definition of 'Iceland' here:

 

Turning to the second part, relating to the gap between NA and Greenland, then this must lead to this kind of line:

Now, if you put all this together, still ignoring crucial geological and bathymetric facts, you end up with what we could be forgiven for thinking was the aim: to cut Greenland and Iceland out of the picture altogether, leaving an artificial construct that is based on a Ireland/UK => eastern Canada/US perception of how the Brendan Prize would be won.  The white lines are what we can reasonably infer from the definitions given, and the orange line is what we can approximately assume must connect them.  

I have no idea where that leaves Iceland and Greenland in terms of which continents they are perceived by the IRTS to be on, if any.  In practice, Greenland is either on the NA plate, or its own, rather disputed 'Greenland' plate, whilst Iceland is split down (roughly) the middle between the NA and Eurasian plates.

 

I have no idea in fact as to why the limits were defined as they were, other than to assume, which seems reasonable given the title of the prize, that it was to create a transatlantic region that didn't include Iceland and Greenland, but did include Ireland, the UK (and the rest of Eurasia), and North America.  

There is no mention of where a southern boundary might lie between Eurasia and, for example, Latin America ("North or South" America qualify for the prize), which can only add to the suspicion that the rules were drawn up with the expectation that the qualifying QSO would occur between Ireland/UK/northern Europe and eastern Canada/US.

In the end, perhaps the Brendan Prize has, in light of recent amazing 2m DX QSOs, fallen into complete irrelevance.  Rare tropo allowed me to make three 2m QSOs with 'just' a 3 ele and 50W with Cape Verde in 2020 - a distance of  4467km, or ~52% further than the minimum Ireland-Newfoundland, transatlantic distance.  So, if there is any purpose left in pursuing the Brendan Prize, that purpose is certainly not based on distance any longer.  It is now, and perhaps always was, just the notion of crossing an arbitrary ocean, with some harking back to the Marconi days, that provides the attraction.

Maybe you'd like to contribute your own take on any problems with these definitions - or that you see none at all?  The more the merrier.  For sure, the whole thing leaves me with a headache!

 

 

 

 

Horizontal on the mountain

After an interesting outing onto the local copper mine recently, I decided to try something a bit more horizontally-polarised for once.

I say 'a bit', because putting up a flat-top dipole is so difficult as to render it entirely impractical for /P deployment.  So, instead, I used a shallow-sloping dipole, cut for the lower end of 14MHz (resonant at just under 14.0MHz in practice).

Nice and warm on the hill.

 

Even putting up a sloping dipole is, though, a lot more hassle than a 1/4 wave vertical, bordering on the impractical.  The string necessary to open-out the dipole arms mean that the extent of the set-up spans something like 15m on each side of the support.  Not so good if in a place where there are a lot of people moving about (EURAO provides 9 million Euro public liability insurance for just 10 Euro annual membership - you would be silly not to take advantage of this important provision).

The support itself, a ~9m fibreglass pole, is very wind-prone, and needs a strong base support; in my case, I had to take a 'L' section steel and hammer it into one of the few places on a rocky hill that this can be done.

And then there is the local population of snakes to take into account when dancing around the dense heather!  

After all that trouble, I ran the dipole facing NW/SE, and then N/S on WSPR, comparing to my vertical delta loop, just ~500m away, but lower down the hill.

Results were good.  Out of 11 stations I used for the comparison on receive, no less than 6 (within the dipole beaming direction) were not heard at all by my delta.  It can be implied, from the strength seen at the dipole that did hear those stations, that the difference in signal for some was as high as 20dB; e.g. K5XL was -14dB in the dipole, so not being heard at the loop implies a 20dB difference with a S/N detection limit of ~34dB.  

Of the remaining ones heard, the median enhancement seen by the dipole was +5dB.  

On transmit, six stations in the beaming direction of the dipole produced a median +8.25dB enhancement, relative to the vertical loop.

In the case of both RX and TX, the best enhancements, all in the narrow range of  +10.5 to +13dB, were seen at the longer DX range (typically west coast US/Canada).  Curiously, KFS, in the San Francisco region, heard the dipole 2dB weaker than the vertical delta - the only such example seen. 

The 1/4 wave vertical on the hill, which is infinitely easier and quicker to deploy in the field, yielded an enhancement on received signals of about +9dB, relative to the vertical delta at home.  

View out to sea from slightly further up the hill, on the previous day's outing with a vertical.  Unlike the dipole, the vertical can be easily put up, even on hard rocky ground like this.

 

So it's pretty clear that there is not much point bothering with the considerable hassle of the dipole.   Les Moxon's assertion in his HF Antennas for all Locations book, that an antenna on a hill next to the sea "must" use horizontal polarisation to benefit from ground reflections and so yield the best enhancements, is proven incorrect when we consider overall results.  Of course, there may be exceptions in terms of specific paths to specific stations.

2m: still not there!

I'm active, but where is everyone else?

 

Despite being past the solstice now, and despite 'assurances' that people are making an effort to be active on 2m, finding someone who actually is active is still proving as elusive as the transatlantic signal itself!

Here are all the active stations on 2m, for all modes, in the past day.  Hardly encouraging, is it?

What I do occasionally see is people in Newfoundland appearing on the active stations map, only to switch off again after a brief few minutes.  That type of operating really isn't going to turn up the goods.  Nor is thinking that only the tropo map is relevant to the Atlantic crossing.

Acid bath

Spurred-on by a very nice result on the local copper mine a couple of evenings ago, I decided to try what must be another unique experiment last night.

Parys Mountain copper mine.  Lots of precipitation ponds.  My experiment was from the ponds seen at far centre left, near the road.  Coast is seen at top.

 

After the main mine had been exhausted, copper continued to be extracted at the site through a simple replacement chemical reaction, where ponds of mine drainage are created, scrap iron thrown in, and after some time, copper is deposited where the iron once was.  

Just one of a complex of precipitation ponds used at Parys Mountain.  Image (C) Pixaerial/J. Rowlands.

 

The pools were also commonly used to cure rotten cattle's hooves - a treatment that appears to have been successful.  The waters were even written up in a very early British Medical Journal paper, where they were said to cure a large number of human ailments.  In those days, there was no scientific method as such, so there is no real proof.  What they didn't know back then was that, apart from iron and copper, the waters contain dangerously high levels of toxins, such as lead and arsenic!

For those with an interest, the displacement reaction to extract copper is nicely explained, courtesy of ScienceABC:

It's been a wet spring and early summer, so most of the ponds that have always dried out in previous years by now, are still full of water.

I decided to take advantage of the situation and seek a pond that looked to be more than just accumulated rainwater, and instead full of water that had percolated over time through the rock waste heaps.  

Out in the bath-warm acidic waters of Parys Mountain.

I found a fairly dark, acidic pool which probably had a good mineral content, and deployed my antenna in the middle of the warm water, which was at about body temperature. 

By coincidence, the pond was half a wavelength in length and breadth at 14MHz!  It had a reasonably open aspect, but only the top of the antenna, where current is lowest, had any view of the sea.  The lining layer of clay at the bottom was just thick enough to provide some support for the clothes line screw that holds my antenna up, helped with some rocks to steady it against a fairly stiff wind.

Unfortunately, it wasn't a great night for DX.  There was a very strong 'D' layer formation, so just about all signals were ~1000km.  

All the same, choosing the five DX stations that were heard, the outcome was only a 1dB enhancement at the acid pool, compared to my delta at home.  Allowing for 'design' gain difference, this amounts to ~4dB real difference on receive.  Not insignificant, but certainly not the kind of result seen when the antenna was at the same mountain, but with a clear view of the sea horizon.  

On transmit, the result was similar: 3dB weaker at the acid bath to the DX stations.

Another reason that partly explains a poorer result at the acid bath is the lengthening effect of the watery environment; on applying the analyser to the antenna at the end of the run, I found the resonance was down at around 13.9MHz.  Not a huge effect, given WSPR is at 14.0956MHz, but it's certainly not helping.

Could have been good, but wasn't particularly.

Of course, the location with a view to the sea was also where I have detected anomalies at VHF.  So I will have to try another site on the mountain now that is not above the mine drainage channel coincident with these anomalies.  That way, I might be able to extract the effects of mine waters from sea horizon.  Maybe...

 

Upper HF goes on and on...

Quite amazing to see 12m and even 6m open way past 00UT over recent days.  

Terminator correct for reports.

Mini-bog experiment.

Three years ago, I wrote a post about an odd phenomenon that happens just down the hill from my house, where radio signals at commercial VHF are strangely interrupted as one passes over an underground acid mine drainage channel.

This evening, it was sunny and warm, so I decided to take a walk up onto the old copper mine, directly above the underground stream.

The mini-bog site on the copper mine hill.  I didn't know the wet ground existed until today!

Well, the ground here is very mineralised, and it has a better and slightly higher view of the sea than my house, even though it's only about half a kilometre away.  In practice, I will never really be able to figure out which element of the environment makes any difference, if any.  But it's a nice comparison, all the same.

Despite this being an almost universally rocky area, I found an unusual area which was like a mini-peat bog, and consequently quite wet.  As well as providing a rare place to screw my antenna mount into the ground easily, such a site was likely to be excellent ground.

Parys Mountain disused copper mine.  Test tonight from upper centre left, not far from the main road. 

 

I ran WSPR at 14MHz using a 1/4 wave vertical, two elevated radials and my FT818, principally for RX-only.  At home, as usual, I was using a vertical delta loop, typically having 3dB more 'design' gain than the 1/4 wave.

Results were pretty amazing.  The median enhancement of the signal, taking differences in 'design' gain into account, was 9dB.  The maximum enhancement, to the signal from DP0POL lying near Svalbard (an essentially all-sea path), was an astounding +16dB, which is the kind of enhancement I might find when the antenna is actually at the seawater's edge.  

This calls to mind, and surely confirms as correct, the comments about the Fresnel zone for radiation angles <1 degree being "pushed [far] out to sea" that Les Moxon discusses with some excitement in his enduringly-good book, HF Antennas for All Locations.  With some elevation as well, that Fresnel zone for ultra-low angles may well be pushed further out again.

Line of underground drainage channel (yellow).  Blue circle centre was test site location.

Map displayed at Parys Mountain, confirming drainage channel route (top left, crossing B5111 road).

The minimum enhancement was +4.5dB, though that was to a TA station lying to the east, with the hill intervening.  It's not really worth taking a lot of notice of that particular result, other than to note it is still significantly stronger than the delta, which has a clear view to the east.

A total of 5 stations heard on the mini-bog were not heard at all from home.  All but one were good DX stations, being AA6FT, R0AGL, VE2DPF and K7GCB.  M0GUC, heard at a median -27dB on the mini-bog, was also missed by the delta at home.  

Taking AA6FT as an example, heard at a median -19.5dB on the mini-bog, the enhancement must have been around 14dB for it not to be registered at the detection threshhold (about -34dB) at home.  R0AGL's results also suggest an almost identical implied enhancement.

DP0POL's reception of my single 1W test signal, compared to others at around the time, adjusted for power output, can be seen as being very favourable - at least 7dB better than any other UK station, of which there are anyway only two (I was 14dB stronger than the other UK station):

The other good DX station who heard my 1W was ND7M.  Again, hardly any other UK stations making it around that time, with my 1W being, when adjusted for the different power output, a whopping 17dB better than GI3VAF:

-18dB from my 1W to ND7M, compared to just -28dB from 5W from GI3VAF.

This test again shows just how poor a comfortable, at-home station can be, even with a proven, good quality antenna.  The difference is not down to any RFI, because I've already proven that's not the case with a comparison of noise at home and seaside, where there was no difference.

Acid mine water found within Parys Mountain. It's red due to a very high level of dissolved iron, kept in solution by a pH of 1 or less (at above ~pH 5, iron falls out of solution as a red floc).

 

These kinds of amazing enhancements are at or beyond what we would achieve with a world-class tower-and-Yagi system when operated at a typical, non-coastal environment.  The environments that provide them really should receive more study.  I plan to do just that in the future. 

Even more importantly, the majority who have poor domestic radio prospects can easily get out into the outdoors and enjoy extremely effective amateur radio with only very modest equipment and low power; an enhancement of 14dB takes a typical QRP output of 5W to an effective 125W!

 

 

 

Rig Expert arrives!

Oooh!  The delivery man just appeared with my new AA-230 Zoom!  

From the outset, I'm impressed with this unit.  Good, strong packaging, containing a set of 'AAA' batteries (not such a great power option, but it also works off USB), a plastic case and strap, a USB cable, and - very usefully - a N-type to SO239 adapter.  

Even better, it has a real, physical user's manual!  That really does make the difference.  Just open the booklet and get going with large images and well-written instructions (if you really need them).

There are a fair few things to learn yet, but it's easy enough to calibrate and get going; using a N-to-SMA adapter for Amphenol calibration caps.   

Here's the first test: my vertical 14MHz delta, 300 Ohm twin to a 6:1 homebrewed voltage balun (yes, I know current baluns are better, but this one gave the best 'what I had available' match), and a short length of RG-213 direct to the rig.  Note: the rig is within its plastic protective cover:

Good old delta! Even better, when I measured the system without an antenna switch in place, the match fell to 1.03:1, or a return loss of 38dB, equiavalent to 0.001dB mismatch loss, or 0.015% of the input power!

That meter-style SWR output looked gimmicky to me before I bought the analyser.  But now I have it in my hand, it's really very good indeed - much better and easier to read than just a graphical plot that you have to scroll a cursor or index line along, which is of course also available.  You can just press frequency up or down, and see the SWR changing.  Very nice touch, that!

So it's a big thumbs up from me!  Oh, and yes, it does show the sign of the complex impedance!

2m Transatlantic?

This is my second year of being active on 144MHz.  I was unbelievably lucky that, within just a few weeks of getting a SSB rig for the band, I managed to cross to Cape Verde with only a 3-ele beam and ~50W on a 3m pole (accepting I'm already 100m up).  That's a QSO people have waited a lifetime to experience.  Had I made the QSO a week earlier, I would - very briefly as it turned out - have held the world record at 2m.  I repeated the exercise three times within a few days.

Hah hah!  3 elements to D4.  The mere idea that it happened upset a lot of people. 

To be honest, if I never made another 2m QSO again, I would walk away, happy with that D4 contact.  I started out to see if there was any chance, given my very good aspect and height in the direction of North America, that I could get a signal across there.

Of course, I realised this was always going to be a very long shot indeed.  But with radio, a lot is possible, if only one can be in the right place, at the right time.  That means, in practice, being in the right place at the wrong time very often indeed, so that the moment when that undoubtedly short duration when 2m might support propagation over the Atlantic is, ultimately, caught.

So, being on a hilltop in view of the sea, and with the comfort of a kitchen shack, my reasoning was that I could have the rig on most of the day and keep a listen out.  There's supposed to be a beacon on 144.285MHz on the Canadian side, which is a lot easier on my rig than endlessly sending out 'CQ'.  

I've listened and transmitted a lot on 2m last summer, the D4 QSO telling me it was indeed worth having a go, and upgrading to an 8-ele Yagi.  I got precisely nowhere, and never heard the Canadian beacon.  There's a beacon on the west coast of Ireland, too.  I hear that most days, but it goes from inaudible to meter-bending, according to tropo conditions.

Listening across the Atlantic last evening.

I hear and see a lot of chat about NA operators listening "often" for transatlantic signals.  But that seems to be a very considerable exaggeration.  Even when 6m is going strong, which might encourage more 2m listening, it's often the case that PSKreporter shows absolutely nobody active in more NE parts of Canada.  OK, PSKreporter isn't perfect, but it's certainly an indication that 2m transatlantic is a very niche pursuit that has very few, and perhaps no sufficiently active participants to stumble across that magic, brief moment.  

Past 24 hours' activity, all stations, on 2m.  It's nearly midsummer, but where are all those NA operators who are supposed to be listening "often" across the pond?

 

There have also been 'expeditions' to try and make the transatlantic crossing.  The latest I know of took place in late July, when peak Es season was past.  In practice, just turning up for a week and hoping for the best isn't remotely likely to cut it, anyway.

Tropo, it seems to me, is how a lot of people believe a 2m transatlantic crossing will be made. That may well be correct for qualifying European stations in tropical and equatorial regions, where conditions are often strong.  

But the 144.285MHz beacon, for example, perpetuates the idea that the crossing will be from the NE coast of North America to Ireland or the UK.  That's the direction it is beaming in.  Similarly, a lot of operators on the NA side are listening in that same direction.  And the very capable, EI2DKH beacon, aims due west.

Added to this, I think misguided, belief in the NA-EI/G path is the lack of understanding I see a lot of operators exhibit, relating to spherical geometry represented on flat maps.  The real, as opposed to imagined path goes very far to the north, where the required tropo almost never forms.  

Indeed, thinking about it this morning, the weakening of the difference in temperature between polar and tropical latitudes, due to climate change, is likely to make tropo even less likely, as I see it.  You can see the general effect on the latest F5LEN forecast, where colder air moving deep into southern latitudes is disrupting any potential development of tropo in more northerly latitudes. 

Colder, polar air sinks south over the Atlantic.  Image (C) F5LEN.

 

Everything, therefore, is in place for everybody to be looking in exactly the wrong direction for this propagation to actually occur.  Of course, there is also Es.  The chances of that providing the sought-after link, though still very low, seem to me to be much higher than the correct tropo conditions occurring in the north Atlantic. 

I agree entirely with the likes of John, EI7GL, who regularly review what is going on in the 2m world, and correctly assess that the Brendan Prize rules are almost certainly going to be satisfied not by Canadians contacting Irish, Welsh or English men, but by Americans contacting a much more southerly part of Europe (anywhere on land or non-tidal areas of the European continental shelf, as defined by the Times Atlas).

Ultimately, it's a question of what's important: a technical demonstration that 2m across the Atlantic is possible?  Or a human being claiming a prize?  The reason I put it that way is because, it seems to me, the best way to address this problem is not through humans laboriously sitting it out, wasting years of their lives for the right moment to arrive, if it ever does, to make a two-way, traditional QSO.  Instead, it would be so much better to put an automated system into place that can listen, second-by-second, day in, day out, year on year, until the stars, or some other mechanism, lines up just right.

 

 

 

Bye, bye, SARK-110

It's been a lovely antenna analyser.  But today, after five years of service, it was time to say 'goodbye' to the SARK-110 analyser.  This is not to be confused with other analysers going under a SARK name.

I've had doubts over the reliability of the 50MHz results from the SARK-110 for a while now.  The cause is most likely to be the remarkably stupid use of a MCX connector.  This has long been identified as a problem with the SARK, and one that, so far as I know, years later, still hasn't been rectified. 

A great analyser, let down by a simple design failure.

 

MCX connectors have a very low connect/disconnect rating - it's only about 200 or so.  Even when we add a pigtail extension to avoid any disconnections, the connector type is still extraordinarily prone to stresses and intermittent connection failure, usually manifest as noise on the trace.  When you look at the male connector side, you wonder how this ever became a connector for anything, let alone sensitive RF measurements; it literally relies on some weak springy leaves of metal and weaker friction along a very narrow ring at the end to maintain the connection.  It's a connector type that should be consigned to the great design failure bin of history.

Rubbish.  MCX male (the SARK end is female).

The first sign of problems came a long time ago at 50MHz, when I tried to cut a replacement matching 75 Ohm stub for my 2-ele quad feed using the SARK's facility for that process.  I never found the correct length using the SARK, which kept reporting the same results, regardless of length.  I shrugged my shoulders and just copied what I already had in place, which had corroded.  But I ought to have looked more closely at what was going on.

I don't know for sure, but either the SARK has a firmware problem, or the MCX connector isn't reliable at higher frequencies.  Whatever the cause, I couldn't get a reliable output for 50MHz from the SARK, no matter what.  This was most recently evident when trying to make a 6m magnetic loop; I couldn't ever get a match with it according to the SARK.  But I've made dozens of loops, and I know it isn't my lack of skill in this case.

As a final attempt, I tried soldering a pigtail extension directly to the MCX connector body.  The SARK is very small indeed, and getting all the plastic push-buttons, especially the power switch button, to line-up again is a massive headache.  When I thought I'd succeeded, there was no change - except the display had now gone south, despite wearing a static-draining wristband whilst working on it.

Well, the only destination now was the bin!  

Was the SARK-110 a good unit?  Yes, but not at 50MHz.  It might have been peculiar to my unit, but that MCX connector - and the extremely small size of the SARK - were always two big problems, especially when the average age of a ham operator is well into retirement territory, where fingers and eyes no longer work as they once did.

There also has to be a question - and I put it no stronger than that - about the generic look and some components of the SARK; it has similarities to other Chinese vector analysers that cost only about £50, which also have accompanying PC-based software, just like the SARK.  

I wasn't going to chance the sometimes poor quality control and unknown effectiveness of such branded units this time, when I need a known, good quality analyser to continue antenna work.  But I might well try one in the not-too-distant future, just to see if our expectation of having to fork-out £400 for a antenna analyser is just misguided.

My analyser for coming years.

Of course, Seeed Studio - the maker of the SARK - is itself Chinese, based in Shenzhen, where so much of the world's electronic gizmos are created.  Their stated mission is to provide "open technology", which may explain the similarity of the SARK to the E-bay VNAs.

Was the SARK good value?  Well, I go on the calculation of the price, £350, divided by the number of years it worked reasonably well.  That comes out at £70 per year, or about £6 per month to have the use of an antenna analyser.  I think that is not too bad an outcome, when I think of the considerable time, frustration and wire that has saved me.  It's fair to say it has paid for itself.

What next?  Well, I was very close to buying a RigExpert - a fine Ukrainian product -  last time, and so I've given them a chance now and have a AA-230 Zoom unit in the post.  We'll see how well that does over time.  At least I'll no longer have the world's most idiotic connector and tiny buttons to contend with!

Tweaking the half sloper.

For several years now, I've been using half-slopers cut for the 40 and 30m bands off a common 1:1 balun feedpoint where the 'missing half of the antenna' is provided by my tower, which has a 12m LFA on top, in this case, acting as a capacity hat.  When I go onto 60m, I just add a disconnectable 5m length of wire to the 40m leg.

Foggy morning!  40m half-sloper in kevlar wire.

Whilst 40m was a very good match, needing no ATU tweak, the 30m leg was not as good as it should have been.  Because putting new antenna ideas up often happens during the boredom of the dark winter months, at least here, I guess I just stuck this leg up on a wet, windy night, found it a reasonable match, easily tweaked with a very small amount of ATU involvement, and just never perfected it.  I also don't really use 30m much, so there was no pressure to change things, either.

7MHz half-sloper: no need for adjustment there!

Anyhow, I decided I would have a look and try to improve the 30m leg this morning.  I had plenty of spare wire wrapped around the end nearest the ground, so it was easy to adjust things.  The wire was quite a bit too long, and there were a few small tree branches and leaves touching the wire, which would also tend to make things look longer.  

So I cut the tree growth first.  Still a bit too long.  Just unwrapped the wire, released the tensioner, and wrapped some more wire up to make it shorter.  After a quick check on the analyser, I could see the wire was now pretty much as good as it could be.

10MHz, after adjusting (in this case, for the WSPR and FT8 frequency areas).

Half slopers have proven very effective, if not the best possible antenna type here.  In particular, they have made it possible to keep working in all but the worst hurricane-force winds, and even then, it's the Yagi that forces the whole tower system to be lowered for protection.  Some say they are difficult to match, but that seems to only be true when a tower or other metal support acting as the missing half of the antenna has no Yagi or similar antenna as a capacity hat.  

 

 

 

Chinese (lack of) quality.

Linear power supplies are something many of us use in the ham shack.  As well as being RFI-quiet, albeit very inefficient PSUs for our radios, smaller units are also useful for quietly powering other equipment, such as Raspberry Pi computers and the like.

Badged linear PSUs like this are widely available online.

 

I bought a generic linear PSU for computer use recently, but it never actually provided any output.  The seller on E-bay was decent enough to refund the price in full, and didn't want it back.  

I had a look at the PSU before dismantling and recycling its component parts last evening.  There was no immediately-obvious failure of anything.  The soldering quality was pretty dire, so I had a go at resoldering several contacts.  

Still no output.

I then dismantled the front controls panel board, where I immediately found at least one cause of the problem:

Hmm!  No wonder there was no output!

 

As you can see, two of the four potentiometers had snapped clean in two!  I'm not entirely sure how this happened during manufacture, but it may be because the board is first secured using the controls at either end, then tightening the centre two controls.  Even so, there should be more than enough 'give' in the solder legs, and even enough strength in the plastic body of the pot to take the strain.

Well, that is the trouble with Chinese manufacturing quite often: poor quality control.

Midsummer approaches...

Nice weather over Wales at the moment.  Lots of upper HF propagation, too!

Not only that, but the noctilucent cloud season is well underway, with a nice display last (04/06/2021) evening.

Radar PMSE returns from 70 degrees north, in Norway, have been showing very strong ice formation at mesospheric height for several days, and it was very strong indeed yesterday.  Due to the sun-NLC-observer geometry, that is the kind of overhead latitude that provide us with the beautiful 'space cloud' displays down here, in the UK, at night.

Spectacular ice formation and consequent MST radar return strength (Andoya, Norway).

The pattern of Es propagation at 50MHz yesterday suggested a good possibility of NLC during the evening.  Transatlantic propagation was very good in the morning, which then suddenly died and translated into a strong propagation to the east and northeast.  With Es lying to the east of the UK, that meant the westerly flow would be bringing any attendant NLC over our way a few hours later.

And that is indeed what transpired!  A beautiful and bright display that was strongest in the pre-local midnight sector:

'Space clouds' grace the northern sky, 04/06/2021, 23:32UT.

It proved to be a busy night: very strong propagation on 50MHz to the US just kept on going.  I had to give up working an endless pile-up - I only have a two-element quad - at around 00UT, as I had NLC-watching duties to keep up!  

Last evening's comings and goings on 6m (21-22UT, approx.)

Just 30 minutes in the hard life of my 12m LFA:

10m FT8 shoot-out: Delta vs. Magloop.

For some time, when needed, I've used my 14MHz vertical delta on its first harmonic, for 28MHz operation.  With an excellent match for both bands, this cut down the need for yet another, separate antenna for 10m.  Feed to the delta is by 300 Ohm twin to a 6:1 balun, then direct via ~4m coax to the rig, no matching box then necessary.

I was well aware that the harmonic use meant a less-than-ideal radiation pattern.  Less than ideal, but by no means 'useless'.  

The 14MHz vertical delta used on 28MHz for this test.  Apex height ~8m.

 

Here's the predicted pattern for the 14MHz vertical loop used at 28MHz, using real ground values (remember, I am on the flanks of an old copper mine).  Lots of horizontally polarised cloud warming, but also useful, vertically polarised lobes at low angles:

MMANA-GAL prediction for 14MHz vertical delta at 28MHz.

 

In any event, you can look at a wire loop and come round to thinking that, as it's so big, it must do pretty well, because of its capture area and excellent matching.  Surely, it must be better than a set of four 10mm copper rings, just 45cm in diameter and 2m above ground?  There are still plenty of angry white men who shout, loudly, that magloops are just 'dummy loads'.

4-loop array, with ovoid primary loop.  External diameter of loop = 45cm.

But is that kind of conclusion correct?  The radiation pattern for the loop to keep in mind is this:

Much better, vertically polarised pattern at low angles for the magloop, with moderate cloud warming also.  24MHz, single turn prediction.

I set up my parallel, 4-magloop array at about 2m base height, using a traditional primary loop arrangement, rather than a pseudo-gamma match this time.  I found the primary loop, fitted between the second and third secondary loops, yielded a quicker matching solution, and that pulling it into an oval, with a fair amount lying outside the secondary loops, brought fine matching control, eventually giving 1:1.

 

I did need to use different rigs to gather the data.  The delta used a FT-450, and the magloop a FT-818.  Both systems used ZLP interfaces and the latest WSJT-X.

I used FT8 to gather data quickly, and from a wider range of stations than WSPR at 10m typically provides.  On WSPR at the moment, most stations are in Germany, with only a small number elsewhere.

Whilst FT8 does indeed provide lots of quick data, processing it can require good spreadsheet skills, or hours with paper and pencil to time-match 15 minutes of captures.  I can tell you I regretted using the latter approach!

The summary is that the magloop was indeed generally better than the wire delta.  Indeed, there were notable, modest-distance DX stations (see last few plots at the end of this post) that the magloop heard repeatedly and consistently, whilst the delta either did not hear them at all, or only heard them occasionally.  4LQL and UA6XES were important DX cases in point.

A note on the plots: the x-axis denotes just the sequential number of simultaneous spots.  I haven't noted the time, because it's not really useful here, and the data analysis was already very hard-going!  The test was between 09:25 and 09:43UT on 31/05/2021.

Let's start with very close stations - 2I0UIR, which is close enough (roughly 200km) to be ground (actually, mostly sea) wave, plus a bit of Es-scattering, possibly:

2I0UIR received FT8 spots, 28MHz.  Median delta = -9.5dB; median magloop = -13dB

Moving out 1127km to Germany, DO5SB provides the first useful comparison, and an indication of better performance by the magloop array:

DO5SB. Median delta = -5db; median magloop = -2.5dB.

The greater number of spots from DO2SBS (1255km), convincingly confirms the magloop's better performance:

DO2SBS: median delta = +0.5dB; median magloop = +15.5dB

Moving 897km to the south, to EA7ALL, the magloop returns a slightly weaker response from this direction's horizontal polarisation:

EA7ALL: median delta = -4dB; median magloop = -7dB.

Italy, the bread-and-butter 'first hop' for British stations, with IO2LXV not too different in both antennas, at 1318km:

IO2LXV: median delta = +6dB; median magloop = +7dB.

Moving back to central Europe, SP9UPH at 1005km provides a good result for the magloop again:

SP9UPH: median delta = +3dB; median magloop = +9dB.

Now on to two Slovenian stations, first S50XX at 1528km...

S50XX: median delta: -9dB; median magloop = -1dB.

 

...and S56GD at 1554km:

S56GD: median delta = -6dB; median magloop = +3dB.

Now some modest DX, from RK4FF at 3213km, where the delta is better on average:

RK3FF: median delta = +6dB; median magloop = +1dB.

Same story for R3PK, at 2672km.  The dramatic change for the delta at spot 4, as in other cases, indicates a shift in arrival angle and/or polarisation for the signal in question:

R3PK: median delta = -7dB; median magloop = -10dB.

Now for the really interesting results, from reasonably good distances.  First, 4LQL at roughly 3700km, which was heard consistently, and at good strength with the magloop, but missed repeatedly by the delta (I don't know why the spreadsheet coloured this differently!):

4LQL.  No medians, as it's clear the delta isn't hearing the station most of the time!

Then UA6XES, at 3635km:

UA6XES.  Again, the delta almost always didn't hear him.

And finally, the same story with E73DN, at a closer 1966km: