Spectrum Lab recording far too many false positives

[jdsteve64]

Hi all,

I am a complete beginner to meteor scatter work and have just joined this group. 

The following equipment is being used: 2m/70cm dual band Yagi on the roof of my house. This antenna is pointed almost vertically (I live in Lyon, about 200 km due south of the GRAVES transmitter). Quick question: should I be orienting the antenna in this way? I am doing it in view of my relative close proximity to GRAVES. Would appreciate any comments on that one.

I am using the RTL-SDR.com dongle and had no issues setting up Spectrum Lab. I am using SL version 2.93 b8 and have downloaded Ian Lauwerys CA script "MetScat_Comprehensive_V7.usr" which if I understand correctly is adapted from Paul Hyde's work and avoids the different processor thread issues between the FFT calculations and the CA loop as mentioned elsewhere.

My current problem (and I have just been testing my system as we come off the April Lyrids maximum) is that I am getting far too many false positives: even after only 30 min tracking the system captures well over 300 events. The SNR_Threshold value in the CA script, which was originally set at 15dB , has been progressively increased, in steps of 2, to 25dB in attempts to reduce the number of false positives, however this has had totally no effect of the number of false events detected.
I have attached a screen capture. I greatly suspect I'm making a Newbie error here so any assistance would be greatly appreciated.
Thanks
Steve

 

[wxsatuser]

Looks to me your receiving GRAVES direct, you can see the pulses, these are the trigger.

I would lower your beam and point it south and see if it nulls off the back.

[morinmau]

Hello Steve & Blue Stragger, first i wanted to start a new topic, but it would have been a clone of this one: I have exactly the same issue you Steve experienced.
I live in Northern Italy, about 80km south of Milan (44.988211N, 9.124843E). Initial antenna (yagi 5 elements) bearing was toward GRAVES radar, so azimuth 314° and about 25° elevation (in order to cross GRAVES beam at 100km height), horizontal polarisation. Distance from GRAVES is 384km.

GRAVES beacon is always present, I would say 70-80% of the time on 24h basis, from strong to medium weak. Furthermore, the signal drifts in frequency from its main location, and often doubles. My waterfall is very similar to your Steve. Some examples below.

 

As you can imagine, when the signal is strong enough, i can get as much as 50% false positives on the total hourly captures. I implemented several test in SpecLab CA to filter the beacon out, but essentially without success. I even set up two frequency bands for trigger (each with different thresholds), but since the signal is drifting, the 'rejection' band spell should be too large to be effective, so removing GRAVES but also (most of) genuine captures.

I move antenna orientation toward west, azimuth about 270° and lower elevation, around 10°. In the map below you see original antenna setting (in green) and this 'west' position (yellow line). I thought that is the most extreme position you can pick up in order to be still in GRAVES radiation area.

The situation slightly improved (GRAVES a bit weaker in average) but nothing changed in term of false positives.

Now .. Steve, did you solve the problem ? In case, what did you do to achieve that ?
My impression is that we are 'too close' to the radar, but honestly i don't understand (technically) how the signal (that comes from tropo reflection) could be so strong and persistent, since I don't see the TX directly, and tropospheric conditions are far from ideal this period.

Any suggestion is welcomed.
Maurizio IU2JWG