Its all in the timing - Spectrum Lab

[Hawksmoor]

Please would one of you help me with a puzzle. For a couple of years, I have been using Spectrum Lab with a Funcube Dongle and a Yagi antenna to monitor radar reflections off meteor plasma. Having got a fair bit of data on various laptops - I thought I might, on cold cloudy winter nights, try some simple analysis using Excel . Not being the most observant bunny, (quite ironic when your hobby is astronomy ), I have only noticed recently that the duration recorded for such events, which I always believed was in seconds, does not seem to reflect the actual duration in seconds.  Is this a calibration issue, a software error or is the read out in decisecs ? If it is a calibration issue how can you do it? Your assistance would be much appreciated.

Puzzled of Lowestoft

[johnb]

Hi

Im no expert but by default the duration is in Cycles - if you divide by 6 you will get seconds - if you are a regular collector of data you may be interested in this: http://meteor.m81.co.uk/welcome.php

 

John B

[The Admiral]

John B is correct, what is counted is the number of cycles that the FFT goes through during the event. Now I'm having to think back a while here so might not be correct in all the details, but what determines when an event starts and when it finishes will affect the duration measured, and that will depend on the script being used by SL. It will most likely depend on when the signal reaches a certain threshold and when it drops below a threshold, and SL will count the number of cycles performed in that time. So it will to some extent depend on the thresholds chosen. Also, you've probably seen scatter traces that last tens of seconds, but during that time the signal strength can fluctuate wildly and may even go above and below the thresholds several times. SL will only count cycles when the signal is above the thresholds, and so may not indicate the 'total' time, but only what might be called the 'live' time.

I can't remember clearly now what the cycle length is, but I have a feeling it will depend on the sampling rate and 'decimate' settings used. You'd need to consult with the SL manual I think.

Ian

[Hawksmoor]

5 hours ago, johnb said:

Hi

Im no expert but by default the duration is in Cycles - if you divide by 6 you will get seconds - if you are a regular collector of data you may be interested in this: http://meteor.m81.co.uk/welcome.php

 

John B

Hi John

Thank you very much for the useful information.

Best Regards George

[Hawksmoor]

3 hours ago, The Admiral said:

John B is correct, what is counted is the number of cycles that the FFT goes through during the event. Now I'm having to think back a while here so might not be correct in all the details, but what determines when an event starts and when it finishes will affect the duration measured, and that will depend on the script being used by SL. It will most likely depend on when the signal reaches a certain threshold and when it drops below a threshold, and SL will count the number of cycles performed in that time. So it will to some extent depend on the thresholds chosen. Also, you've probably seen scatter traces that last tens of seconds, but during that time the signal strength can fluctuate wildly and may even go above and below the thresholds several times. SL will only count cycles when the signal is above the thresholds, and so may not indicate the 'total' time, but only what might be called the 'live' time.

I can't remember clearly now what the cycle length is, but I have a feeling it will depend on the sampling rate and 'decimate' settings used. You'd need to consult with the SL manual I think.

Ian

Hi Ian

Thank you for the information - very helpful.

Best regards George