More ham DX tests on VLF this weekend
- Just seen this from Stefan DK7FC:
My next VLF kite experiment will be tomorrow, saturday, 2nd october :-) After my request to the german air traffic control i got a very kind telephone call at 01:30 AM and later i got the permission to go up to 200 m above ground. See attached mail if of interest for you.
So, if there will be enough wind and no rain, i will try to become QRV at arround 8 UTC.
- (after some resonating tests): Transmitting a short message on 8970 Hz, mode: DFCW-600, hoping to be visible on several receivers in the EU region and maybe Canada and USA (!?)
- Transmitting a CW message (in 12 wpm and in 25 wpm), hoping to be audible at Bernd/DF8ZR (16 km range) and my grabber (40 km)
- Maybe trying a xband QSO 9kHz/144 MHz to DF8ZR
- Transmitting a message on 6470 Hz, mode: DFCW-600, hoping to be visible at many RX, especially Renato Romero who has set up a grabber page at http://www.vlf.it/cumiana/livedata.html
- Transmitting a 30 minute carrier on 8970 Hz which allows long time integration. Paul Nicholson wants to try to record the VLF audio stream from Florida and maybe extract my signal.
- Being visible at PA3FWM at the university of Twente who set up a grabber page at http://websdr.ewi.utwente.nl:8901/grabber.html
- Being visible at SQ5BPF and maybe a russian station?
- Being visible at Paul Nicholson and Markus/DF6NM, of course ;-)
- Finally reaching the 1 A antenna current limit
- and all i have forgotten
I will record my signal of my grabber and will extract an audible signal, if possible. This will be filtered, frequency shifted to 800 Hz and published here as a mp3 file.
My grabber will show the 8970 Hz signal at http://www.iup.uni-heidelberg.de/schaefer_vlf/DK7FC_VLF_Grabber.html
The 6470 Hz will be visible at http://www.iup.uni-heidelberg.de/schaefer_vlf/DK7FC_VLF_Grabber2.html
It is now on the Canadian and US RX stations to make possible what still seems to be inpossible: A transatlantic transmission/reception. The first time my signal can reach a 50 mW ERP limit (wind forecast announces strong wind, so a high kite angle, so a good efficiency!) on 8970 Hz. It is still not sure how the propagation behaves on this frequency at this time and locations so we may see a surprise. Of course any report is very much appreciated, no matter of what distance or country. Every clear reception above 859 km is a new world record by amateurs on VLF!
So, hope the sun will shine at good wind conds and low QRN and high interest :-)
Best 73, CU, Stefan/DK7FC
PS: Markus/DF6NM, will you be on the phone again to make QSP? :-)"
> ... Paul Nicholson wants to try to record the VLF audio stream fromFlorida and maybe extract my signal.
Paul, if you will be watching the Cape feed, do you want more gain? It
looks like it'll be very quiet in Florida this weekend, so I can crank
it up if you'd like.
- Hi Tom,
> It looks like it'll be very quiet in Florida this weekend,No thanks Tom, the gain can stay as it is.
> ...do you want more gain?
I'll timestamp and record the signal at the stream server,
along with Sebring too. Frankfurt to Florida is a long shot
but worth the small effort it takes to analyse the signals.
I'll be applying filtering and sferic blanking in the post-
processing. The forecast of a quiet background is encouraging.
I hope that at some point not too far off, we can switch over
to streaming accurately timestamped signals up to the server.
Then Cape Coral and Sebring can make a 2 element beam.
I have a little more software to put together before the test
tomorrow, so I'd better get on...
- Activated narrow band spectrograms on Cape Coral, Sebring,
and Cumiana streams, for both 6kHz and 9kHz bands.
I see you're testing with a marker near 8970,
This isn't showing up on my spectrogram, is the marker
outside the range 8969 to 8971? Or have we got that sample
rate error back again on the stream?
- Everything is ready here.
Cape Coral and Sebring are recording. Joel has tested
the frequency calibration of the uplink - a test signal came
through this morning at 6470.00 Hz.
I think the uplink sample rate from Cumiana might be wrong but
Renato is plotting/recording locally anyway.
A 'grabber' is the term used by radio amateurs for a spectrogram!
- There is now a signal showing on Stefan's spectrogram,
> ... quiet conditions in Florida...Indeed! I monitor the 289.0 kHz DGPS beacon on Cape Canaveral and see a
very unusual 0.00% peak error rate over the last 14 hours. The path to
it is across the noisiest part of the penninsula. There is nothing out
there. A perfect day for a weak signal.
> > ... quiet conditions in Florida...I've tried 1 hour integrations at 0.5mHz resolution.
Nothing but the grass I'm afraid. Those quiet conditions
must be just too flat. I'll try 5mHz now, in case the soundcard
clocks are wandering a bit.
Stefan is transmitting again, looks like a stronger and
more constant level now. I think his kite must have gone down
earlier, judging by the way the signal petered out.
- Hi Paul,
The problem Stefan is actually facing is the generator ran out of fuel, and at the moment he trying (obviously w/o luck) to get it running again.
The last "over" was made from the car battery (400 watts), but the battery also went down, causing a slow, 18-millihertz "sweep". That's why your 0.5 mHz integration cannot catch it.
Now keep fingers crossed for the rest of the afternoon !
Wolf (watching VLF on a remote site, i-net access too slow for streaming)
--- In VLF_Group@yahoogroups.com, Paul <paul@...> wrote:
> > > ... quiet conditions in Florida...
> I've tried 1 hour integrations at 0.5mHz resolution.
> Nothing but the grass I'm afraid. Those quiet conditions
> must be just too flat. I'll try 5mHz now, in case the soundcard
> clocks are wandering a bit.
> Stefan is transmitting again, looks like a stronger and
> more constant level now. I think his kite must have gone down
> earlier, judging by the way the signal petered out.
> Paul Nicholson
- Just received this info from the RSGB reflector, about DK7FC's 8970 / 6470 Hz attempt..
Have a nice weekend,
Original message from Markus, DF6NM :
16:45: Unfortunately the attempts to restart the generator have failed, so no possibility to transmit on 6470 Hz. Now Stefan is finally packing up and calling it a day. He is really looking forward to come home and go through the reception reports!
Many thanks to all involved, and also my congratulations to Jacec for his excellent reception at 902 km.
Greetings from Nuernberg,
- Thanks for the update Wolf.
> The problem Stefan is actually facing is the generatorHmm, that really should be a checklist item.
> ran out of fuel,
The 1 hour integration was over the 1st transmission where the
frequency seemed constant.
The decaying battery accounts for the large drift on the 2nd
transmission. Actually I though the signal was stronger on
that one. I've tried a half hour integration with a drift
correction of 38mHz/hour but still nothing.
While you're on Wolf, I have a Spectrum Lab question. I got
a bit of a signal from Stefan's test on 25th, where he had about
250W into an earth loop. I'm using Stefan's spectrogram to
get the timing of the transmission,
and I'm getting the times by counting pixels on the spectrogram.
But SL is using a 4.5mHz = 222 second FT size - are the pixels
timestamped with the start time of the FT buffer, or the end time,
or maybe the middle?
If I just integrate from 09:45 to 11:00 I get this,
with the two peaks 50mHz apart. The transmit frequency was
changing a little and I can sharpen each peak separately if
I put in some drift correction and restrict the integration
period. It would help if we can pin down the transmission
times more accurately.
- Hello Paul,
> While you're on Wolf, I have a Spectrum Lab question. I gotThe timestamps are for the 1st sample involved in the FT.
> a bit of a signal from Stefan's test on 25th, where he had about
> 250W into an earth loop. I'm using Stefan's spectrogram to
> get the timing of the transmission,
> and I'm getting the times by counting pixels on the spectrogram.
> But SL is using a 4.5mHz = 222 second FT size - are the pixels
> timestamped with the start time of the FT buffer, or the end time,
> or maybe the middle?
Actually, the thread which draws the spectrogram pulls samples out of a buffer, and the buffer-object returns the timestamp of the first sample (along with the samples). It is not aware of the FFT windowing (cos^2 or similar) so the timestamps are, strictly said, off by half the FFT length.
Have a nice sunday everyone,