[CQ-Contest] Busted Reversed Beacon Network spots

[Brett Graham]

S56A said about RBN spots of HB9CVQ on frequencies other than where 
other skimmers were spotting the same call:

> HB9CVQ used 1 kW on 160 m with 2 x 36 m doublet. S50ARX uses QS1R with
> similar doublet on the top of the hill. HB9CVQ was certainly strong at
> S50ARX almost as at K1TTT with 84 dB S/N - HI. All that was required to
> translate HB9CVQ from 160 m was strong BC somewhere around 12 or 16 MHz.
> IMD products were just 10 dB above dead 20 m band noise. Ocassional CW
> Skimmer errors occur on harmonically related frequencies.

Yes, they must be intermods from b'cast.  Gaps between bogus spots are 
when modulation trips up skimmer.  Other strong signals causing these 
that aren't modulated are all intermittent, otherwise the bogus spots 
could happen all the time the intermod-causing carrier was present.  And 
there are strong signals everywhere, so there is nowhere to go where an 
intermod product will not fall in some amateur band.  Gee, why didn't I 
think of that - it's so obvious & it explains everything.  Mystery solved!

Eventual admission from N4ZR that a subset of just wrong-freq-same-band 
spots were seen at a tad over one-per-minute (during a contest) got me 
looking at this again, using non-weekend RBN data for 2013-03-08 (also 
with all spots of DXpedition/special event calls omitted, so hopefully 
missing majority of stations deliberately QRV on >1 band or >1 frequency 
within same band simultaneously, or deliberately transmitting a fecal 
signal as no contest that day).

S50ARX spotted the same call as other skimmers did within one minute - 
but on another band - 33 times that day.  And another 176 times S50ARX 
spotted the same call within one minute as another skimmer, but on a 
frequency >1 kc away on the same band as the other skimmer spotted it.  
Together, I believe that is almost 3.8% of all of the described subset 
of S50ARX's RBN spots that day.

I never finished a tally of the nearly 6k same-band spots that day 
(equivalent to 4-per-minute), so can only share these points about the 
wrong-band ones:

69 skimmers wrong-band spotted at least once.
7 skimmers wrong-band spotted >1% of what they spotted.
55 of the 69 skimmers wrong-band spotted >=0.5% of what they spotted.
15 of the 69 skimmers made 50% of all the wrong-band spots.
As a group, 0.52% of what the 69 skimmers spotted was a wrong-band spot.

In the end, it looks like all SDR architectures have spurious responses 
- in practice I believe far worse than those of conventional receivers & 
thanks to RBN this can be seen, ultimately preventing disappointment 
here from a planned upgrade to direct-sampling.

73, ex-VR2BG/p.