31 October 2016

Nov 1 Far Side DX

[I've decided to change the way I blog about my amateur radio exploits, since I'm blogging more than I thought I would and the monthly pieces are getting too long.  So I'll start a fresh blog piece each day from now on, well each day I have something to say anyway.]

ZL7G was busy on 80/CW when I got up before dawn today. The op was having a hard time digging callers out of the EU pile, and so was I: the pile was quite unruly and sigs were mushy, about S7 with some QRN to make things still more challenging. The op was also failing to persist with some partials and made a few sending mistakes before eventually he went QRT. I guess he had been doing the long night shift and was exhausted. I can picture him slumped over the key catching some zzzzz's.

02 October 2016

Far side DXing: October 2016

Oct 2: the Oceania DX contest phone section got off to a slow start with poor LF conditions last night and not much better on HF this morning, but this afternoon I found a few stations lurking on 15m and this evening I had a little European run on 40m. In other words it was patchy.

Oct 7: LoTW confirmations arrived from TO5FP including our remarkable 80m RTTY QSO ...

Oct 8: today I completed a little job that has been on the go for about a year - I put in a base for a prop to support the tiltover tower when it is tilted. It took two attempts: yesterday I used some old premix concrete that was kicking around in the workshop for an indeterminate period, probably a couple of years or more. It didn't set overnight so I dig it out and today I bought 30 kg of new fast-set readymix. This stuff starts setting almost as soon as the water goes in, so it was a race between me and the concrete to fill the hole. I won!

Oct 9: I got up at 2:30am local for the CW leg of the Oceania DX contest only to find 80m unusable due to severe QRN from a storm front linking Brisbane to Auckland:

40m wasn't too bad though. During the day I messed around on 15m and chased after S9YY, a German group on Sao Tome and Principe Island, eventually catching them on 17m.  Back to 40m for ZL dusk, and something over 300 40m QSOs in the log before I got bored and wandered off ...

Oct 10: the Oceania DX contest website is QRV again today after exceeding its bandwidth limit last night (doh). I guess that means the website is popular!

I got up early again to find 80m in good shape today. Kamal 4S7AB called me for a new DXCC on 80. I hope this time he QSLs. 

I spent quite a while fruitlessly calling an Estonian on 80/CW. I don't think he heard me at all but someone tried to 'help' which just confused matters. Not to worry, I'll try again another day. 

Just as the band creaked open this morning, S9YY was on 20/CW with a wide split and few IDs. Not a bad long path signal though with lots of Europeans calling them.

Oct 11: I've had 2 more QSOs with T31T today, including a surprisingly difficult one on 15/CW. This morning, they were strong on 15/CW, working NA so beaming directly away from me. I called them half-heartedly while working, then gave up and left them to their pile. This afternoon, I tried again as they were working JAs, beaming away from me and not so strong. I heard them work a VK and I think a ZL, so I tried a bit harder and eventually they heard me. I had to repeat my call for them maybe 3 times, despite running about 800 watts out to a 5 ele Yagi. I caught them on 30/RTTY this evening too, much easier despite the crowded band full of loud EU callers.

Oct 12: I'm finding the bright, crisp little P3 panadapter display surprisingly useful. I occasionally used a FunCube USB dongle and panadapter software in much the same way previously but it took some setting up, the software was flaky and it was tricky to set aside enough monitor space for the display. It had one advantage over the P3, though, which was to display callsigns from DXcluster spots at the relevant places on the waterfall. The P3 is easier to run: drawing just a few hundred milliamps from the K3 accessory output socket, it springs to life when I turn on the K3 and just sits there quietly doing its thing. I use it mostly to monitor DX signals and look either for callers in QSO with them or for holes in their pileups, and to find signals on quiet bands.

That's a screen-grab from the P3 of Stan H40GC working a pileup of JAs on 15m SSB. The wiggly yellow line shows near-instantanteous signal strengths across a 20 kHz slice of the band (taken from the K3's IF output), with my noise floor at about -145 dB.  The blue area below is like a chart recorder, constantly streaming downwards and showing signals varying in strength over time by the brightness of the bluey-green marks. The red strip shows my subreceiver monitoring Stan on 21270 kHz, while the green strip shows my transmit VFO just over 5 kHz HF, at the bottom of his split range "listening 5 to 10 up". There were 3 or 4 JAs calling stan, evenly distributed in his receive range. [I've added some text - H40GC and Pileup - plus the white bar showing the pileup, in a graphics editor. Those aren't on the P3 display.] 

Aside from QSLing 100% via Logbook of the World, I am sending out about 500 QSL cards per year, on average, the majority via GlobalQSL:

GlobalQSL charges just over US$100 (about NZ$150) for 1000 cards, so those QSL cards are costing me about $75 per year, about 20 cents per day. I send some direct cards too, about one envelope every two months of my own plus one per week responding to incoming direct requests, and I contribute about $300 to DXpeditions per year in donations and OQRS fees.  I'm spending roughly $1,000 per year on equipment, but recoup about a third of that through second hand sales. I don't know how much the shack electricity costs: perhaps I should check.

By comparison, it would cost me $1,100 per year just to join the local Napier golf club, plus the cost of clubs, balls, checkered trousers, a trundler trolley and a daft cap, plus diesel for the 50km round trips. Amateur radio is in the same ballpark, and I know which one I'd rather do!

Oct 15: there are several strong DX sigs on 80/CW this morning but none of them can hear me. ER1SS is calling CQ OC with short 'listening' gaps. DK1WI and RW9JZ are also CQing with no hint of hearing me. Admittedly it is now about 20 mins after dawn though, so perhaps the D-layer is already fully absorbant today.

At lunchtime today, Logger32 started spewing out errors due to DXcluster spots, specifically RTTY spots from a Japanese Skimmer ...

The above message causes cluster processing to hang until I click the OK button, whereupon a little sequence of further errors appears, all of which also need to be individually OK'd before cluster processing resumes ... until the next spot from 7L4IOU arrives, more specifically "7L4IOU: -#".  It looks to me like maybe he has forgotten to set the call modifier number for his Skimmer node e.g. 7L4IOU-5 (or whatever), so it is sending out spots with his callsign followed by the hash character whereas Logger32 is expecting an integer. Possibly. Anyway, it's an annoying bug because of all those confirmation clicks needed to clear it and no obvious way to stop the errors within Logger32 other than turn off the cluster feed (shock! horror!). 

I tried filtering out 7L4IOU spots using the filter option in VE7CC Cluster User software, but that didn't work. Maybe I should be filtering on 7L4IOU-#? I don't know, since there is precious little help in or on Cluster User. Perhaps the filtering only works on spotted calls, not spotters? 

Next option was to apply cluster filtering within Logger32:

Success! That seems to have done the trick. Luckily, the originator blocking filter is evidently applied before module ClusterProcessing is called, and it looks like ClusterProcessing could do with additional input checks for invalid originator callsigns.

Inspired by a photo of a similar but much more elegantly crafted design by W0GJ, I've made my first ever whirligig today:

It's very rough-n-ready, unfinished, a prototype really just to figure out the mechanics and learn what works ... or doesn't. 

The propeller part worked out surprisingly well: six strips of thin steel pushed into 45-degree slits in a plywood wheel, each one held in place with glue and a wood screw. The drive shaft running the length of the crossbar is a scrap of tubing, not straight as it turned out (one of the learning points from today!), running in two short offcuts of aluminium tubing glue-gunned in place as simple sleeve bearings. The actuator at the rear is, now, a small plywood wheel with a bent metal rod linking it to the end of the arm. Originally I tried a wooden cam but immediately discovered that slack in the arm's 'axle' (bent rod again) allowed the arm to twist under pressure from the cam - so that's another learning point. The finger presses down on a crude Morse key, hand-fashioned from a scrap of treated wood and supported on a base with more bent rod. 

I added a tail made from two scraps of ply to turn it in the wind, the whole contraption spinning on a metal rod sitting in some tubing in the end of the green support now screwed to a fence post for testing.

If there's enough wind tomorrow, I'll grab some video showing the hand tapping out a series of dots. It does need a fair blow of wind to run - with hindsight, the propeller should probably have used wider vanes, although there's room to add another 6 between the first 6. The metal link at the rear needs more work too: currently it is just a press fit into the actuator wheel and the rear of the arm but it works its way loose. I somehow need to stop it sliding out horizontally backwards while still allowing rotational movement - maybe a circlip and washer in a void cut into the arm or a nut and washer on the front face of the wheel.

I must keep an eye out for a cheapo secondhand Morse key for the Mark II version, rather than that crude wooden monstrosity. And a better axle for the arm, plus a more anatomically-correct arm, maybe rescued from a shop mannekin, would be good too.

But, for now, the prototype is done. It took me 5 hours messing around in the workshop, though it seemed half as long so I guess I was in flow. All in all, a fun way to spend a damp Saturday afternoon.

Oct 17: CQ CQ CQ ...


I think I've finally figured out a way to get Logger32 to QSY the radio so that a spotted RTTY station comes up at the default receive frequencies in GRITTY: the trick is to set the MMVARI "Preset Audio frequencies" to 2285 Hz in the MMVARI settings:

MMVARI is still offset (it puts itself 85 Hz above the spotted frequency, as if spotters are reporting the mid-point frequency instead of the mark), but GRITTY is on frequency. Normally I only use MMVARI for transmitting, generally twiddling around to find a hole in someone's pileup, so the MMVARI offset is a minor annoyance.

The menu buttons in GRITTY aren't entirely obvious with no pop-up tooltips and a not-exactly-helpful help file, so here's a crib for the important bits:

9N7XW is on 17/RTTY this evening, weak, working simplex and, of course, he is being mobbed by JA and UA callers. If he would operate split across a reasonable range (up 1-2 or 1-3 would be enough), I might stand a chance but not simplex.  After calling for some while, I saw 9N7ZT spotted on 20/CW, working split, so I QSYd, tuned-in to his callers and worked him with 2 calls, easy peasy.  I've retuned to 17/RTTY now to find 9N7XW and a never-ending stream of JAs still slogging it out simplex. I wonder if they will be in Nepal long enough to work out the JA seam?

Oct 18: the planets aligned, for once. Franz OE7FMH emailed me yesterday, saying that he'd heard me on 40/RTTY calling S9YY yesterday, and could we perhaps make a QSO for a new slot.  I agreed and arranged to be on the band by 0530z today. I got there about 30 mins early to set things up and check out MMVARI and soundcard settings - good thing too as I'd messed up the configs earlier today. Anyway, it was soon resolved and I was on the air, when Franz emailed me to say he was up early and raring to go. I replied with a frequency and started calling him. Moments later, he called me and we completed a QSO, for probably the most straightforward and painless sked I've had.

I signed and uploaded the QSO to LoTW at 0522z and emailed Franz my sincere thanks for a very efficient exchange - quite a contrast to the usual kerfuffle when people ask me for skeds but can't hear me or haven't dampened their string, leading to days or weeks of torment as we try to establish contact. We usually make it, in the end, but the fun has drained out by about the fifth attempt: it's just a slog from there (unless it's a new slot for me too!!).

Oct 19: it looks as if, within the next few days, we will exceed last year's total of 1,169 logs submitted for the OCDX contest:

I've stumbled across a bug in Rick N2AMG's LOTW/EQSL synchroniser add-on for Logger32.  I accidentally clicked the Download LOTW QSL button twice. The second click, while the info was still downloading, brought up a run-time error message as the add-on crashed:

It's no big deal to restart the add-on, or to avoid double-clicking the download button in the first place. Thinking perhaps I ought to report the bug to Rick, I decided to check whether I am using the latest version. According to the program I am ...

... but according to Rick's website, the latest version is not 1.4.99 but 1.5.01:

I dutifully downloaded the updater but when I went to install it, WinZip told me it has the same size and date/time stamp as the file I am current using after all:

... at which point my goodwill and spare time almost ran out (I have sent Rick info about the bugs etc. since it appears he is working on a new version of the utility).

This evening, I tamed an unruly EU pileup on 30m - or at least retained what remains of my sanity - by spliting down 1 for a change. I had been troubled by a couple of VKs chatting away on SSB just HF of me (yes, on 30m!) and the usual, expected, anticipated and feared antisocial, not to say ignorant and obnoxious EU zoo, so the down-split made sense. None - not one -  of the Europeans who had been clambering over the top of each other and calling through my QSOs evidently copied or understood "CQ DN1" ... but OT and FB CW op W6LFB did, so we had a very pleasant QSO - me on my new-fangled Bencher and Winkeyer, he on a WW2 Navy J38 straight key. I'm somewhat ashamed to admit that his fist was not only nicer sounding but more accurate than mine  :-)  Tnx Jim in Dallas, it was a real pleasure, again. I am looking forward to many more QSOs OC, and patiently hoping that the EU rabble will eventually twig that their antisocial on-air me-me-me antics simply aren't productive. Life's too short.

Here's another bug or design flaw in the LoTW/Logger32 exchange, as demonstrated by one of my two QSOs yesterday with Franz OE7FMH. The 40m RTTY QSO is the problem: yesterday  I completed a 40/RTTY QSO, on sked, with Franz as I reported earlier and signed and uploaded the QSO info to LoTW shortly thereafter. Franz duly signed and uploaded his QSO info also, and LoTW matched the two records. So far so good.  However, somewhere along the line, the mode has been variously recorded or reported as "DIG" and "RTTY" ... and it appears Logger32 and/or N2AMG's syncher can't handle that discrepancy. L32 refused to synchronize the inbound LoTW confirmation for our RTTY QSO, unceremoniously dumping the QSO info to the Bad.ADI file. I checked the file manually, deleted the header and saved it as "Fixed.ADI" (since I think L32 refuses to import a [fixed] bad.ADI file), then imported it successfully into L32 ... only to find I now have 3 successive QSOs with OE7FMH in my log:

The orange background shows the two LoTW-matched QSOs but, according to L32, the phantom "DIG" QSO remains unmatched. Ho hum.

My money is on the bug being in L32, which sometimes has trouble handling digi-mode QSOs made using MMVARI (e.g. allocating spurious DXCC countries, which Club Log points out for me ...). 

Both of those were RTTY QSOs. Seems to be something sporadically going wrong either in MMVARI or when the QSO info is passed to L32 to be logged.

Oct 21: I caught VA1AXC/CY0 on Sable Island on 20/SSB this morning, via the long path. I've been stalking him for a few days now. Today his signal peaked right on cue at 20:30z and I was finally able to break through his US callers. He has a characteristic Southern drawl, fairly slow and steady and, thankfully, patient enough to dig me out of the noise. CY0 is an awkward path for us, polar both ways. I guess the A index has fallen just enough after the last couple of days.

The only other excitement today has been H44GC - Stan LZ1GC now back on Solomon Island after his trip to H40 Temotu - and V6Z - a cheery couple of Scots winding up for CQ WW SSB.

Oct 22: I'm having a few antenna issues at the moment, including the 15m beam. It is set up for the CW end (naturally) but lately it seems mysteriously to have moved itself HF, up to mid-band. This afternoon while I was calling someone on 15/RTTY, the amp retuned itself a couple of times then packed it in with an error message. I tried the usual amp reset procedure but it wouldn't play, then it went weird, half-lighting the LEDs, doing a Night Runner display on the LEDs, and then cycling itself on and off without me even touching it. A power-cycle hard reset calmed things down, and cleared the fault - and just in time. TL0A was spotted on 30m, quite early for us so I wasn't surprised to hear him just above the noise floor. His US and EU callers were much stronger than him, including those calling simplex over the top of him several minutes after he had split. I tried a few speculative calls, listening for his QSOs and finding a hole in the pile nearby but not expecting to make it, when I caught "FB". A couple more calls later, we had synchronized and although I never did get my complete call from him, I heard all the parts of it separately and we swapped reports. He immediately started calling CQ VK, or possibly CQ VK/ZL/OC (he was back below the noise floor and QRM level already). Nice one!

Oct 23: my DXCC total for 2016 has passed 265 and, with 40 CQ zones too, my CQ DX Marathon count has matched my Oceania-leading score from 2014, with 2 months' DXin still to go. The CQ country list includes Shetland Islands as a separate country and I think maybe I worked someone up there too.

Next on my hit-list are ZD8W and TY2AC, both of whom are quite active but I keep on missing. ZD8W will be there until December 1 and has a good signal here (e.g. during the recent WAG contest, working only Germans!) so I'm quietly confident of bagging him, but the TY always seems to turn up on bands that are completely closed here.

According to Club Log, the JX2016A I worked in April was a pirate, not the real JX2016A. That has been my only JX QSO this year and given that it is Winter there now, I'm unlikely to get another chance to bag Jan Mayen in 2016. It's one that got away. Mind you, if a QSL or LoTW confirmation for that QSO magically materializes before 2017 I shall reinstate my claim!

Oct 24: this morning, HV0A was busy working Europeans, T33 something was sending CQ then QSYing to another band, repeatedly, and 4U1UN was spotted. i didn't catch them, but it's OK as I'm not sure any of them were genuine. I gather (from circulating emails allegedly from the station manager) that the real HV0A was QRV on 15m SSB last weekend but not this week.

Oct 26: up well before ZL dawn this morning, I found myself chasing a weak ER1SS, fruitlessly again, on 80/CW at about 16:45z. After he disappeared, I CQ'd on the frequency and (thanks to the wonderful K3 QSK) noticed in the headphones that my echo was markedly LF. I double-checked the K3 settings to make sure I wasn't using split or RIT/XIT - nope, that's not it. I glanced over at the P3 after sending a brief message to see the echo of my transmitted signal on the waterfall to the low side of my TX frequency:

The echo was about 140 Hz LF, with a spread of about +/- 25 Hz. The spread looks normal to me (partly due to the width of my transmitted signal, partly due to the ionosphere I guess) but the LF shift is unusual. 

As dawn approached, I picked up a fair few callers from EU and AS so the 80m band must have been in reasonable shape. Someone spotted me with a comment "Good sig despite aurora" and I noticed the A-index had shot back up to about 20 after gradually settling down to about 5 over the past few days. A short while later I found an AuroraWatch UK aurora warning from Lancaster Uni sitting in my inbox with a link to this pretty graph ...

So, I think there's a fair chance the 140 Hz LF-shifted echo was the result of the aurora, perhaps some sort of Doppler shift? Interesting phenomenon anyway, whatever the cause.

The aurora warning (now upgraded to a red alert) probably explains the dearth of HF signals today.  The ionosphere is well and truly fluxed ...

Oh well, not to worry, I should be working anyway to clear the decks before maybe having a play in CQ WW SSB at the weekend. I just hope condx improve for the DXpeditioners and contest teams that will be doing the contest in earnest.

My 2016 DXCC scores are coming along nicely. According to Club Log I just hit the century on 80m and a double-century on 15m ...

Oct 31: for once I had a semi-serious go at CQ WW SSB, just 15m though SOSB(A) HP. I've claimed 350k points from 1,000 QSOs and 26 hours in the chair. I've almost doubled the ZL record of 188k points set by Wes ZL3TE in 2005. 
My QSO rates were OK at times, according to Cabrillo Statistics by K5KA & N6TV.  My best hour was 169 QSOs.  Holger at ZM4T peaked at 193/hour and soundly thrashed me overall.

Something went wrong towards the end. I could hear Holger still running the Americas on 15m when I struggled to be heard by anyone. I guess I ought to check the ~130m long run of coax to the 15m antenna in case the goats, sheep or cattle have chewed on it. And when I finally get the 5 band quad built, I really must install open wire feeder ... or build a contest shack under the tower!