VLF: Detections of 5 microwatt transmission

[Markus Vester <markusvester@aol.com>]

Dear LF,

 

for my VLF test transmissions on the last weekend (Oct 9: 8969.998 Hz, Oct. 10: 8989.997 Hz), I have received reports from three receiving stations:

 

- Walter DJ2LF near Erlangen (20.2 km) received the carrier again in good quality, using 0.95 mHz resolution. Radiated power and received SNR were quite similar to our two-way QSO on June 4th.

 

- Stefan DK7FC in Heidelberg (178.5 km) reported about 10 dB SNR, and both dashes are still visible in the QRN minima on his 47uHz grabber window

 

http://www.iup.uni-heidelberg.de/schaefer_vlf/DK7FC_VLF_Grabber.html

 

This was the intended purpose of the experiment, and a nice counterpart to the earlier detection of a 200 uW kite transmission on Aug 29th. Even though both of us were locking the samplerate to a 10 kHz GPS-derived reference, the dashes appeared about two pixels low - perhaps due to a very minute rounding error in SpecLab's frequency scale display.

 

- To my utter surprise, Paul Nicholson (Todmorden, 1030.5 km) produced two spectra, taken over the duration of the transmissions:

 

   http://abelian.org/vlf/df6nm/2010_10_09a.gif   (9:50 - 18:00, 34 uHz)
   http://abelian.org/vlf/df6nm/2010_10_10a.gif   (9:00 - 15:00, 46 uHz)

 

After taking a deep breath, we now have to deal with the question whether this is a significant positive detection. Except for a known central artifact on 8970, the highest peak appears in the correct frequency bin in both spectra. Naively, one could then propose that the probability of this happening at random would simply be the inverse of the number of displayed bins, ie. around 1:230 for Saturday and 1:170 for Sunday. Thus the combined probability of a false positive detection on both days would seem to be only 1 in 40000. Certainly there is a degree of arbitrariness in the choice of the display range (8 mHz).

 

Paul estimated that the signal was about 3 standard deviations on Saturday (0.3% false detection rate), and 2 sigma in the higher noise on Sunday (5%), giving a combined false positive probability of 1 in 6667.

 

We can also look at the plausibility of the absolute fieldstrength of the peaks (about 0.2 fT, equivalent to 0.06 uV/m). If I remember correctly, Paul's first detection of Stefan's kite signal on March 15 was at about 3 fT, and Stefan was then radiating approx. 1 mW EMRP. Scaling this down to my estimated 5 uW EMRP, and taking another dB for the slightly higher distance, would theoretically result in 24 dB less fieldstrength, or 0.19 fT - almost a perfect match.

 

So by these lines, it would seem at least very likely that Paul has indeed observed my feeble signal! We intend to do repeat the experiment in the near future for additional confirmation.

 

Very many thanks to all involved in this work!

 

73, Markus (DF6NM)