I think propagation was pretty good but it was noisy during the evening here and there were still excessive crackles this morning. Its confusing because storms in Mexico and Caribbean don’t look organized or as aggressive as they sound and the storms in the Great Lakes region and Ontario don’t seem extremely impressive either, but perhaps I am wrong about that. The bottom line is that it is too noisy for October!
The geomagnetic field has calmed to unsettled levels with periods of quiet. The Bz has been variable but is currently at unity as solar wind velocities continue at over 600 km/s. DST values are stable but continue below normal. These numbers probably should be a lot worse than they are.
DK7FC was the centerpiece of trans-Atlantic activity, receiving reports from WD2XSH/17 and WE2XGR/3:
Doug, K4LY / WH2XZO, reported that QRN was high, limiting his reception of WSPR stations to five. He was decoded by thirty unique stations.
Roger, VK4YB, reported that the band was very much “one-way” and issued a “Code-5” on the basis that it was a quiet night in Queensland. Roger’s reports and statistics follow:
“Rx 10*wh2xxm (-23) 18*wh2xgp (-13) 34*wh2xxp (-10) 3*ve7bdq (-22) 11*wi2xbq (-26) 26*wh2xcr (-20)
Tx 6*wh2xgp (-22) 5*ve7sl (-25) 1*wg2xsv (-27) 10*kl7l (-25) 11*we2xpq (-24) 25*wh2xcr (-12) 2*jh1inm (-24) 12*zl2bcg (0)”
Ken, K5DNL / WG2XXM, received decodes from VK4YB:
Ward, K7PO / WH2XXP, was decoded by 53 unique stations, including six VK’s, ZL2BCG, JE1JDL and JH3XCU. He adds that there was a “sunrise bump” at 1336z where he received decodes from four VK stations:
Larry, W7IUV / WH2XGP, reported that he decoded by fourteen WSPR stations including VK4YB and several eastern stations using the Omni receive antenna. He was decoded by 41 unique stations including VK4YB plus four other VK’s, ZL2BCG, ZF1EJ, and several eastern stations. Larry indicates that high latitude absorption continues to impact northern stations:
Neil, W0YSE/7 / WG2XSV, experienced another big night, with decodes of VK4YB. He provided the following comments and statistics:
John, VE7DBQ, was decoded by four VK’s during this session:
Joe, NU6O / WI2XBQ, received reports from three VK’s during this session:
Rick, W7RNB / WI2XJQ, reports that the band seemed a bit subdued during this session. He decoded nine WSPR stations and was decoded by nineteen unique stations.
Phil, VK3ELV, continues to have nice reports into Japan, receiving WSPR decodes from 7L1RLL4 and JH1INM late in the previous session. He also received reports from this session from JH3XCU:
It was nice to return to the morning 472 CW sked and hear stories from my friend who just returned from a hiking adventure. CW is perfect for story telling! I received and provided numerous CW level reports through the night from a wide array of stations so the band seemed to very open. Given the nature of the noise I have a hard time believing that these receiving stations were unaffected so its likely that reports could have been even stronger on a quieter night. I can’t wait for winter! My WSPR transmission reports can be found here and my WSPR reception reports can be found here.
84 MF WSPR stations were observed on the WSPRnet activity page at 0200z. A review of the map suggests that there were more stations active later in the evening and even with 84 stations on MF, that was still more than 20 than on 80 or 160-meters. Activity is very good and its only October! VE5DLD was observed as a new receiving station but he was reporting the old frequency of 503.9 kHz. If anyone has contact with this stations, please ask him to look at this link in order to get details on where we are actually operating and how to reconfigure the software.
Regional and continental WSPR breakdowns follow:
There were no reports from the trans-African path.
Eden, ZF1EJ, reported several North American stations including WH2XGP, who has returned to air after a few days of work at his station:
Laurence, KL7L / WE2XPQ, reports damage to the transmit loop and noise from arcing power lines. High winds had subsided some by morning, but aurora was seem to the North. Laurence concurs with Larry, W7IUV / WH2XGP, that the path to VK is depressed from these latitudes:
Merv, K9FD/KH6 / WH2XCR, shared two-way report with JA and VK and was decoded well into the eastern portions of North America and the Caribbean. Merv is both hearing and being heard well:
Jim, W5EST, presents, “NUMBERS OF DECODES PER NIGHT VS. PEAK SNR”:
“Yesterday’s blog discussed diminishing returns in noise canceller performance and limitations at low SNR levels. Today, whether you use a noise canceller or not, let’s talk about how the ability to copy or the average chance of decode varies with nightly peak SNR.
For WSPR, JT9 and CW, the decode threshold or readability threshold is about -28, -25, and -13 dB respectively. * Your experience may vary.** Remember that receiving a TX at -28dB SNR by your RX station using a directive RX antenna enjoying both absence of local QRN and no storm noise is far superior to the experience of decoding a station at -28dB SNR under lots of local QRN and storm noise. All the dB thresholds remain valid, but the reception experience is so different because of the much greater signal strengths needed to deliver same SNR in a higher noise level environment. Far fewer WSPR slots per hour, if any, will be decoded in the high noise case from any one same TX station because it will have far lower median SNR in high noise.
Turning to the subject of higher SNR levels now, here’s a rough rule of thumb from my 630m experience: For WSPR2, JT9 and CW, if the SNR is at least about -10dB, -8dB, +4 dB respectively, then you generally will miss only a few decodes or continually copy Q5 signal even if SNR varies a few dB down due to QSB. Of course, deep QSB will punch holes in anything.
I state those mode-specific high dB values -10dB, -8dB, +4 dB because those numbers stand a few dB above twice SIQ*** above the mode’s readability threshold. On 630m, SIQ is about 7-9 dB on one-hop paths.
As you consider nightly station cases lower and lower in peak SNR, the numbers of digimode decodes–or percentage of time you can read CW–declines rapidly. The log of the number of 630m WSPR2 decodes per night of weaker stations (adjusted for TxPct) is approximately related to their respective peak SNRs that night.**** No adjustment for TX power dBm nor type of RX antenna is included because both factors similarly affect the log of number of decodes and peak SNR dB, all else equal.
At some very low peak SNR level, the chances of receiving any decode from a given station become so remote that no decode from that station is received an entire night at your receiver. Perhaps only on the most favorable nights of a season do some such stations get a single decode or a handful of decodes on your 630m receiving system under those circumstances.
If you have your own data or wisdom to add on this topic, let us know so we can blog it. TU!
*For a table of approximate thresholds of these modes and other modes, see 2/15/16 blog: http://njdtechnologies.net/good-band-conditions-continue-as-storm-levels-never-materialize-we2xpq-and-wg2xsv-make-history-with-first-jt9-qso-from-alaska-on-630-meters-pa0a-ua0snv-jas-and-vks-for-wh2xcr-and-we2xpq/ SNR is conventionally stated in 2.5KHz noise bandwidth.
**The decode process has unavoidable statistical fluctuation http://njdtechnologies.net/100516/ (scroll 1/3). Near decode threshold, SNRs at which decodes occur do fluctuate. So you may very occasionally see successful WSPR2 decodes down below minus 30dB. Conversely, you may sometimes fail to decode a signal that on average was stronger than the mode’s decode threshold.
***SIQ, or SNR Interquartile Difference, means the number of dB that span the middle 50% of the night’s decodes (centered on median SNR) between two stations on a given path. For a map of 630m SIQ estimates, mostly 7-9 dB, on one-hop USA paths on some early November nights in 2015, see: HamCom 2016 W5EST slide 21: https://www.dropbox.com/s/4hsk6ltvi65z8en/PDF_HamCom_PPT_Presentation_Final_061116.pdf?dl=0 Accompanying presentation audio (33:48-35:30): https://www.dropbox.com/s/96es4o2p1etn4d8/HamCom_JimW5ESTAudio_Presentation_050316_061116.mp3?dl=0 .
For weaker signals and longer distances, 630m SIQ measurements may show less than 7dB because many weaker TX slots fail to decode. 630m propagation is a complicated subject, but for communication purposes 7-9dB SIQ is probably the “bottom line” for a 630m station operator.
**** I fitted equations for weaker stations subject to scatter and based on WSPR2 data at W5EST (AR) on WH2XGP (WA) and WG2XKA (VT). Both results individually point to one formula more generally for number of WSPR2 nightly decodes on 630m paths, in N. America 630m fall-winter season especially:
Nwspr2 ~~= [5 TxPct (TSRe – TSSw)/12hr] x 10^(0.11 ΔSNRpeak), and not to exceed Nmax.
ΔSNRpeak is amount the night’s peak SNR exceeded -31dB. Example: At 20% TxPct, median number of WSPR2 spots for a station peaking -20dB on a given night of 12 hour common darkness is about 16, (16=5×0.20x(12/12)x10^(0.11 x (-20 –(-31)) subject to scatter variation among nights with the same peak SNR. Compare with the log scatter plots and discussion at Feb. 2, 2016, blog: http://njdtechnologies.net/interesting-pre-sunset-propagation-yields-strong-reports-across-north-america-and-strong-domestic-conditions-overnight-ve3ot-cw-audible-in-texas-wspr15-activity-in-europe-and-alaska-ve1hf-g3xkr/ Tell us how to improve the work-in-progress formula.
When peak SNR is strong, limit predicted Nwspr2 value to maximum number of WSPR2 slots spanning the common darkness duration on the path on that night. Use this formula for RX-only:
Nmax= TxPct x 30/hr x (TSRe – TSSw) (UTC, hr).
TxPct refers to the remote TX station. If an RX station also has TX, ignore slots when both station-ends of path transmit at same time. Common darkness duration in hours is the time interval between UTC time of westward station’s sunset that night and subsequent eastward station sunrise irrespective of which station is TX or RX.
Average chance of decode overnight is estimated by: p(decode) = Nwspr2 / Nmax. Given 12 hours of common darkness on a path for the 16 spot example. Nmax= 72 = 0.20×30/hr x 12hr. p(decode) = 22% = 16/72.”
Additions, corrections, clarifications, etc? Send me a message on the Contact page or directly to KB5NJD gmail dot (com).