Early results from eclipse experiments

Discussion in 'Radio Propagation Questions' started by Steve Nichols G0KYA, Mar 20, 2015.

  1. Steve Nichols G0KYA

    Steve Nichols G0KYA Administrator

    BBC_Radio_Scotland_810kHz.jpg Medium Wave
    Early results showed a marked improvement on signals from BBC Radio Scotland (near Falkirk) on 810kHz as received in Norfolk at 09:30hrs - see below. I have a whole three-hour medium wave Perseus SDR recording to wade through at some point. Some of the members of the Medium Wave Circle have also recorded the whole of MW for later analysis.

    I have now reviewed some of my Perseus SDR recordings and they are amazing.

    BBC Radio Scotland 810kHz goes from -74dBm at 09:19 to -64dBm at 09:30, and back to -77dBm by 09:32.

    BBC Radio Wales 882 kHz goes from -72dBm at 09:19 to -63dBm at 09:28, and back to -77dB by 09:34.

    Laurie (G3UML) in West London reported that reception of BBC Radio Wales on 882kHz improved dramatically too. Chris G4IFX, in Surrey, saw a similar effect with BBC Radio Scotland on 810kHz.

    Mike (GU4EON) reports: "On my 40mtr inv Vee BBC Scotland 810Khz came up from a very noisy S4 to just over S9."

    Wolverhampton Grammar School ran the experiment using radio station “France Info” on 711kHz (see attachment below). From around 09:20 music could clearly be heard above the noise.Reception improved slowly rising to a maximum of S8 or S9 around 9:30am.Between 9:30 and 9:35 it was easily listenable but not as strong as the previous night. It disappeared more quickly than it appeared, taking <10 minutes to drop below the noise level.

    St Thomas More Catholic school in Buxton postcode listened to France Bleu on 864khz. Their signal peaked signal 0935 and vanished at 0955.

    Peter Duffett-Smith (Cambridge) reports: "Hi Steve – my recordings show a clear eclipse signature. Please see sample attached (below) at around 09.36. I have recordings for all of the 9 KHz channels from 540 to 1600 KHz; these are just three. The purple curve is for a local radio station (1026 Radio Cambridgeshire). The blue and orange curves are both for Lisnagarvey (Northern Ireland)."

    Geoff G4AFJ, Leicestershire, reports BBC Radio Scotland 810kHz coming out of the noise at 09:41 to S-3 before going back into the noise shortly after.

    Iceland/Faroe Islands
    Lots of RBN and WSPR activity, but it doesn't look like stations Thor TF4M in Iceland or OY6FRA, the club station in the Faroes, were received via WSPR on 160m anywhere in Europe.

    OY6FRA was received all over the UK and Europe on 80m though - that will need closer analysis once the day's full WSPR results are available to download.

    Don G3BJ reports "On 160m, good signals from eastern Eu". Lots of 80 and 40m WSPR stations too, which is good.

    First 80m Reverse Beacon Network reports from Dave G3YMC:

    EI6BT G3YMC 3529.1 CW CQ [LoTW] 8 dB 25 wpm 0820z 20 Mar
    EI6BT G3YMC 3529.0 CW CQ [LoTW] 8 dB 22 wpm 0851z 20 Mar
    EI6BT G3YMC 3529.0 CW CQ [LoTW] 11 dB 22 wpm 0906z 20 Mar
    EI6BT G3YMC 3528.9 CW CQ [LoTW] 16 dB 22 wpm 0922z 20 Mar
    EI6BT G3YMC 3528.9 CW CQ [LoTW] 19 dB 23 wpm 0928z 20 Mar
    EI6BT G3YMC 3529.7 CW CQ [LoTW] 9 dB 23 wpm 0944z 20 Mar

    Note how the signal strength improves to 19dB at 09:28.

    Chris G3SJJ reports: "I was on 3525kHz from 0800 to 1000. Initially there were a few southern UK stations worked, then at around 0900 it suddenly opened up to DL and LA. I heard two very weak DXy callers but couldn't get callsigns. The eclipse passed over here at 0920. The band returned to normal at around 0945."

    Tom PC5D (Netherlands) reports: " I was active on PSK63 on 7041 +/- QRM during the eclipse. Rig 10W + indoor dipole. The few UK stations I heard during the eclipse, suddenly disappeared when the critical frequency (FoF2) dropped. I noticed similar effects during the UBA eclipse experiment in 1999. Today it also resulted in a big dead zone around so only a few stations where heard."

    Steve G0KYA

    Attached Files:

    Last edited: Mar 20, 2015
  2. James Hollander

    James Hollander New Member

    630m and Solar Eclipse: PA0A into GM3YXM and G3XKR
    Today’s triple play of spring equinox, new moon and solar eclipse featured some suggestive 630m activity in UK/N.EU on the WSPR database. More details below.
    Others may recognize info I missed, so other posts are most welcome.
    73, Jim Hollander W5EST, Little Rock, Arkansas, USA.
    Netherlands’ PA0A 2w sent 50% TxPct WSPR (every other 2 minute time slots) 705km into Scotland-based GM3YXM, who decoded a W-shaped PA0A SNR sequence of [-2, -2, -2, -2, -4, -4, +1, -4, -1, -1, -2, -2] in the time interval 0918-1002Z. Solar eclipse maximum occurred in Glasgow 0934Z and Amsterdam 0938Z. http://www.timeanddate.com/eclipse/in/uk/glasgow
    0936Z was about the time of solar eclipse maximum at the single-hop midpath center point over the North Sea. The midpath would also be midway between points on either side of it where single-hop PA0A signal would have penetrated the D-layer on ascent and descent. The SNR peak +1dB of PA0A WSPR SNR at GM3YXM occurred in the 0942Z time slot, about 6 minutes later than 0936. No change in PA0A frequency 475.718 registered at GM3YXM.
    G3XKR at 763km in Devon also decoded a W-shaped PA0A SNR sequence of [-12, -12, -11, -13, -13, -15, -9, -14, -14, -13, -11, -12, -11,-15, -19, -11, -10] between 0914-0946 in the listing time interval 0906-1010Z. Solar eclipse maximum occurred in Plymouth 0925Z and Amsterdam 0938Z. http://www.timeanddate.com/eclipse/in/uk/plymouth
    0932Z was about the time of the solar eclipse maximum at the PA0A/G3XKR single-hop midpath center point over the North Sea near the English Channel. The SNR peak -9dB of PA0A WSPR SNR at G3XKR occurred in the 0930Z time slot. No change in PA0A frequency 475.715 registered at G3XKR.
    Compare today's info with 2-6dB-high W-shapes observed during a 1999 solar eclipse involving 5-minute moving average signal at 75KHz and 183KHz LF reported for 1 TX and 4 RX stations at 470-762km distances away. See PA3BSH (2001) http://misan.home.xs4all.nl/eclipse.htm (scroll 20% for observations, scroll half-way for simulation).
    A possibly contrary observation: G3YXM 555km 630m decodes of PA0A 0906-0950Z were steady at -3dB WSPR SNR through the 0930Z eclipse maximum at Worcester, except for -2dB at 0918Z. Possibly, weak ground wave favored by the salt-water North Sea path may have obscured the eclipse effect at the modest 555 km distance.
    On 80m, the WSPR database showed epiphanies of numerous 80m signals at 800km and longer distances from UK. I didn't check WSPR info at shorter distances.
    This 630m eclipse-suggestive info has no counterpart on the LF WSPR database nor 160m WSPR database. LF had too little activity. 160m had some singletons around eclipse time but nothing really eclipse-convincing. 40m and up showed nothing easily discernible to me as eclipse-related, given the continual daytime WSPR activity anyway.

    Not-pertinent: The 630m WSPR database today shows 3/20/15 0850-1138Z mid-morning reception of the distinct Netherlands station 0.5 watt PA0O into GM3YXM 723km and G3XKR 787km today. No peak around the solar eclipse shows up on the database for that reception mostly because the TxPct duty cycle was low and the solar eclipse occurred in a 20 minute gap between 0934Z to 0956Z. G8AGN’s decodes of PA0A ended 0934 prior to the eclipse max there.
    PE1RKT reception around the solar eclipse shows nothing unusual. Receptions by F1AFJ and PA3ABK/2 ended before, or show no spots, on the database around the solar eclipse.
  3. Tim K

    Tim K Member

    How are you planning to analyse the WSPR results? I was running on 80m as MM0TKE using a far less than ideal antenna setup (a 40m inverted V with an ATU) and 5W. A few stations received me but for a considerable period the local noise was so high there was no receive. I suspect it's due to some major building work going on next door and electrical power tool use.

    Looking at the reports from others, I seem to have been received by more stations during the eclipse but the signal to noise didn't improve. It would have been interesting if both figures were reported separately
  4. Steve Nichols G0KYA

    Steve Nichols G0KYA Administrator

    Hi Tim,

    I will have the full WSPR and RBN downloads for the day - the plan is to pick a number of stations and look at the S/N ratio of a number other stations as received over a given path. Then do the same for other paths. The first thing is to strip both the downloads down into usable files - that is, 160, 80 and 40m EU-only plots between, say, 0800 and 1130. Then an analysis can start using Excel and its filter function. What I might do is also make the stripped down files available on my g0kya.blogspot.co.uk site so that others can download them and do their own analysis. This was a process I used with RBN on a 6m Es study last year and, although it was hard work, it is doable. I think the exact process for WSPR will become apparent once I start.

    I want to look at north /south and east/west paths and also look to see if there was a delay between any build up (160m) or attenuation (more likely on 40m due to the FoF2 critical frequency dropping) as the eclipse moved west to east. We'll also look at potential paths to Iceland and the Faroes. I have lots of medium wave plots from more than 50 contributors, the Fairford and Chilton Ionosonde data, 5MHz beacon data and much more. Other ideas will no doubt come up in due course. The first stage is to get the data in a format that can be analysed easily.
  5. James Hollander

    James Hollander New Member

    Regarding Markus DF6NM's suggestion that ionospheric lensing accounts for the "W" shaped sequence of LF SNRs in the 1999 solar eclipse: I can provide math to determine the positions and times of maximum spatial rate of change of solar radiation during a partial solar eclipse, relative to partial eclipse maximum. This could help one consider the lensing concept more deeply. If interested in a copy, request it directly to me at mrsocion@aol.com.
  6. Roger Coley

    Roger Coley New Member

    I've tried to present a simple analysis of the WSPR 160m data collected on 20th March 2015 over the period 08:00 to 11:30. The original data used comprised 9140 spots mainly taken from WSPRnet as provided by Steve Nichols, G0YKA, plus my own meagre results that didn't get loaded to WSPRnet.

    The spots were made from 68 stations across Europe and the top 8 stations G4JNT (1041 ), G3ZJG (521), G1MFG (513), G4BOO (480), G8LIJ (435), G0FCU (419), G3TKN (419) and GJ4CBQ (416) all had more than 400 spots each.

    My initial question was 'Were these stations spotted more often during the eclipse period?”. A graph of the spots for the top 4 stations is given below


    and the next 4


    Generally speaking there is a peaking effect from 09:15 to 09:45 during the eclipse period, although the high spots recorded for G3TKN outside this period need further investigation.

    Although this is not a scientific analysis I hope it gives a helpful visual picture of what happened with the 'top' stations on 160m with WSPR.
  7. Bill G4GHB

    Bill G4GHB New Member

    Interesting stuff and I look forward to seeing the full results and analysis.
    I listened to Radio Deutschlandfunk in the Manchester area at 06.30 with a good signal strength and going down and only just detectable at 07.45 and then going out completely. It then started coming in at 09.10 and getting stronger. At 09.14 onwards it was good for several minutes with deep fading in between until 09.27 with rapid flutter and getting weaker, at 09.31 just detectable and by 09.33 no signal.
    Interesting to note other people reporting their stations were peaking around 09.30 when mine was well on it's way out.
    I used a Pye valve set type P131MBQ, no meter so just audio sound levels used.
  8. James Hollander

    James Hollander New Member

    Bill, yes, interesting timing. What Deutschlandfunk DLF freq or its location did you monitor? That can help indicate path midpoint from DE to Manchester, if reflection was near midpoint. Wiki DLF lists Nordkirchen (NW DE), Thurnau (SE) 549; Cremlingen (N.Central), Ravensburg (far S) 756; Ehndorf (far N) 1269; Heusweiler (SW) 1422 kHz. Your description reads like multipath, with rising signal on a first path at 0910. Then...as if a strengthening second path arose 0914-0927, phasing and fading the signal by relative 1st-2nd path length changes as eclipse moved. Then followed by weakening of both paths after 0927 until 0933. Do you have any more detail about the 0914-0927 epoch? (since the path strengths could have peaked after 0920 while interfering.) Maybe Markus' DF6NM lensing and skew propagation could also help explain why various reported peak times vary so much from eclipse max times of respective path midpoints to reporting stations. 73.
  9. Bill G4GHB

    Bill G4GHB New Member

    Hi James,
    Yes, it was Heusweiler on 1422 kHz.
    I can't give more details between 09.14-09.27 as I was only going off signal strengths by ear and I can only say it didn't appear to get a lot stronger at any one point.
    The only radio I have with a meter is a Racal RA17L which needs repairing, I have replaced an R.F. choke and screen grid resistor so that stage is now oscillating again but there's still another fault to be found so I was not able to use that.
    Even my homebrew H.F. rig has no meter, no sidetone either and I've kept saying it would be nice to have sidetone for about 14 years now, just one of those things I never get round to doing.
    The results look as though they will be interesting when everything has been analysed and if more reports come in with the same effect I noticed as well as the ones already saying theirs peaked around 09.30 when times and locations are taken into account as the eclipse moved over.
    73, Bill.

Share This Page