Monday, November 30, 2009

LOTW vs. eQSL: Lies, Damned Lies, and Statistics

Came across some interesting observations out in the blogosphere re: LOTW, whether its worth the hassle and such, by VE3OIJ and KB6NU. I shared some thoughts a few months ago not long after I got LOTW working. Some time has passed so I did some logbook analysis on my contacts since getting back on the air (16 June through the 30 November) to see if things got better or worse.

Every QSO in my logbook for this time period (379 total) has been uploaded to both eQSL.cc and LOTW. Here's what I got for my trouble:
  • Confirmations via eQSL: 118 (31.1% return rate)
  • Confirmations via LOTW: 113 (29.8% return rate)
  • QSOs confirmed via both eQSL and LOTW: 55
  • QSOs confirmed via eQSL but not LOTW: 25
  • QSOs confirmed via LOTW but not eQSL: 58
  • DXCC entities confirmed by both eQSL and LOTW: 16
  • DXCC entities confirmed via eQSL but not LOTW: 10
  • DXCC entities confirmed via LOTW but not eQSL: 14
So what does this prove? Nothing, really, except that more than two thirds of the stations I've worked in the past 6 months don't use either method of electronic QSLing, or simply haven't bothered to upload their logs yet. As for those who do QSL electronically, just under half use both LOTW and eQSL. Unless and until more logging software integrates automatic uploading, I don't see how electronic QSLing will be a viable replacement for the analog method.

I'm sure more electronic QSLs will trickle in over the next few months, but in my experience I've found that most people who use eQSL and/or LOTW will upload their logs pretty soon after the QSO. Therefore, I don't expect the percentages to skew drastically either way. Whether my experience is in any way representative of reality, I can't say. I'm curious to know how other hams are faring. Is my 30% return rate typical?

Meanwhile, I'm sitting on a pile of cards ready to go out to DX stations but I haven't had a burning desire to cough up the $150+ for postage (including the return postage that most DX stations ask for), nor have I sent off any cards to the outgoing bureau. I've been happy thus far to sit back and watch the LOTW verifications trickle in and my DXCC and WAS totals slowly increase.

In a perfect world, the ARRL would accept eQSL as a legitimate confirmation method. But in reality, this would net me very few "new ones" since many of those 10 eQSL-only countries were already verified when I was QRV from New Jersey, years before either LOTW or eQSL.cc were a gleam in their programmers' eyes.

November 2009 Wrap-Up

The hunt for WAS and DXCC continues...: Got back to the shack this month (actually, late October) to work a few contests and return to my WAS quest. I worked Maine three times, leaving only Delaware unworked; QSLs still needed from AL, DE, ME and NV.

Now I'm turning my attention to filling in the grids on individual bands with an eye towards an eventual 5-Band WAS. As of now I have 42 states confirmed on 20m and 16 states on 40m. On the 80, 30m, 15m and 17m bands my counts are in the single digits, though the cards are trickling in and the number of states worked is increasing as I spend more time on 40m at night, 17m and 15m during the day, and avoid the lure of 20m as best I can unless something good pops up on the cluster. Even made a few QSOs on 80m the other night -- the Tarheel doesn't deliver an SWR better than 2.5:1 but my signal made it into the midwest for KS, MO and IL. Unfortunately 12m and 10m operations will have to wait as my Tarheel won't tune those bands (need to get it closer to the ground, I'm told).

FWIW, I now have 39 states confirmed on eQSL.cc, so their eWAS won't be far behind the "official" WAS award from ARRL. As for other eQSL awards, I have 17 zones and 36 countries confirmed.

On the DXCC front, I worked 7 new countries in Oct/Nov and received confirmations from 8 new ones. Now just 17 QSLs away from all-band DXCC...

Misc. QRM: I had my first MT63 QSO with N1GKE in Rhode Island on 17m -- saw his RSID pop up in DM780 and gave it a try. Neat mode -- not sure if it's as error-free and fade-resistant as Olivia, but close.

October/November 2009 Stats (To-Date Confirmed/Worked):
27 DXCC Entities (83/124)
18 CQ Zones (27/30)
26 States (46/49)
160 Total QSOs

Friday, November 20, 2009

Waiting for Baudot

I just submitted my meager log from last weekend's WAE RTTY test -- just 45 QSOs and a whopping claimed score of 1,530. I only operated for a few morning hours (between 1125-1345 on Saturday and 1245-1700 on Sunday) in order to give DM780 a try at good ol' fashioned 170/45 Baudot, a mode I haven't worked since days of yore with the trusty old KAM and a terminal program. High time to give the new technologies a try, said I.

Some random thoughts and observations about RTTY operation with the K3 and DM780 follow...

The DM780 + HRD Logbook combination did fairly well, considering HB9DRV himself says "HRD/DM780 is not contest software." As such, there is no easy provision for sending or receiving QTC info for extra multipliers, and it wasn't clear at first how to get DM780 to increment serial numbers in the exchange field (put them in [brackets], I finally discovered). Using a fresh database file for the log (as I do for every contest) lets me use the logbook's Awards Tracking and Worked Status functions to keep an eye on what stations and countries I've worked on each band, but I have to be careful to individually set the other databases (previous contest logs, plus my master logbook) to not figure into the worked status lookups (this is one in the Logbook Databases control panel).

Another limitation from a contesting perspective is HRD Logbook's inability to output Cabrillo files for log submission after the test, a feature that was available in HRD v4. So I have to use another app (SP7DQR'S nice ADIF2CABR freeware app) to convert an ADIF export file into Cabrillo format, and that only after doing a search-and-replace on the ADIF file to change the tag to that the conversion app is looking for. No biggie, I mud-wrestle data for a living, so this is just another day at the office...

That said... I'm familiar and comfortable with DM780 and HRD Logbook so it all worked just fine for me.

I worked the first day with the K3 in DATA A mode before remembering that AFSK A mode allows DUAL PB filtering to peak the mark and space tones. The DM780 waterfall screenshots below illustrate the difference far better than words:

K3 set to DATA A mode. BW = 400 Hz, Fc = 1530 Hz

K3 set to AFSK A mode. BW = 400 Hz, Fc = 1530 Hz, DUAL PB enabled

Note that the overall bandwidth in AFSK A mode is quite a bit narrower, even though in both cases the K3 was set to 400 Hz, and the distinct notch between mark and space tones is indicative of how effective this filtering mode works. Back in the pre-DSP days with the TS-930S and JST-135/245 transceivers and NRD-525/535 receivers, I used to use a Datong FL-3 audio filter which had a RTTY mode that accomplished the same thing, albeit at the AF stage.

DATA A and AFSK A each have their advantages. In DATA A (or AFSK A with DUAL PB turned off), I can open the bandwidth up and see a good portion of the band (I generally set DM780 to display 3 kHz on the waterfall and set the K3 bandwidth to match) and all the signals on the air, then select the desired signals with a point-and-click like I do in PSK31 mode. If QRM is a problem, I can crank down the bandwidth and shift the passband center frequency to pass only the station I'm working; once the QSO is complete, a quick tap-twist of the K3's shift control recalls my wide settings and I'm back on the hunt. I rarely touch the VFO dial, all tuning is done with the mouse.

In AFSK A mode with DUAL PB enabled, however, the K3's center frequency is fixed at 1530 Hz so all tuning must be done with the VFO. Also, the bandwidth is limited to 500 Hz max (which as shown in the image above is a bit less in practice, more like 250 Hz or so) making VFO tuning very touchy and slow (the 1 Hz fine steps must be used) and renders the waterfall useless for spotting other signals. But the filtering advantage is huge, especially in a contest scenario.

For me, it's a no brainer -- in the latter stages of the WAE contest I found myself using DUAL PB almost exclusively, occasionally switching it off and opening up the bandwidth if the band was quiet or if I've already worked the majority of the stations I tune across, since clicking on a waterfall makes it far easier to hunt and pounce.

DM780 facilitates the switch from narrow DUAL PB to wideband waterfall tuning easily: I first activate the center frequency marker (Tools>Program Options>Waterfall menu, or F8) and set it to match the K3 DUAL PB center frequency (1530 Hz). After finding a signal on the waterfall and clicking on it, I can then click the C button just above the waterfall to center it at 1530 Hz (HRD offsets the K3 VFO frequency to do this), and then activate the DUAL PB (press/hold the #6 key on the K3 keypad). The bandwidth is narrowed to 500 Hz, and the mark and space tones are perfectly positioned for decoding. To switch back to wideband, press/hold DUAL PB, and tap XFIL a couple of times to select the 2.7 kHz filter or use one of the filter presets to select my standard wide data setting of 3 kHz. This can perhaps be simplified to a one-button process using the new macro feature Elecraft just added to the latest firmware version; I need to check into that...

Note to K3 users: when working RTTY in AFSK A mode, either the radio or DM780 needs to be set to reverse, as AFSK A demodulates the lower sideband while DM780 looks for the upper. DATA A, however, works in the upper sideband.

Friday, November 13, 2009

CQ WWDX SSB 1996: Lost Log Discovered!

Almost as much fun as finding a $20 bill in an old coat is finding an old contest log on an even older hard drive! Today I discovered my log data from the 1996 CQ WWDX SSB contest, which I must have exported from Log Windows before my erstwhile Toshiba laptop screwed the pooch back in the late 90's.

After some data mud wrestling, I was able to convert to ADIF and import into HRD and saw my total DXCC worked count jump from 119 to 123 (the "new" ones are 3DA Swaziland; FS Saint Martin; V4 St. Kitts & Nevis; and GU Guernsey).

I then uploaded the new QSOs to LoTW and immediately had 13 new QSLs credited to my account, with a handful of new DXCC/bands confirmed: VP5 Turks & Caicos (80m); P4 Aruba (15m); 8P Barbados (15m); V2 Antigua & Barbuda (15m & 20m); and GI Northern Ireland (20m) -- the latter two being all-time new ones confirmed, bringing my DXCC confirmed count to 80, and DXCC Challenge totals to 239 worked/123 confirmed.

God only knows how many of my other contest logs are gone forever -- I must have worked at least another dozen contests back around that time, and all these logs got nuked along with the Toshiba. This is why I now keep separate paper logs as backup.

Also found this photo of the old N2HIE shack in Closter, New Jersey circa 1998.


Complete with that utterly dreadful JRC NVT-56 desk mic. Wish I held onto that, seeing how one just sold for over $1,300 on eBay. Please, stop laughing... I'm not kidding:

WTF is wrong with people? Clearly some hams have more money than sense...