G8JNJ

About

Username
G8JNJ
Joined
Visits
3,123
Last Active
Roles
Member
Points
400

Activity

  • benson

    Seems formatting disappeared in that last message. A more readable version is at http://sparks.epizy.com/SNR/Wideband_SNR.txt

    January 28
  • benson

    Hi Martin,

    Included the Python code that plots SNR based on linkz's microkiwi waterfall. Standby for that...it seems that this type of file can not be posted on the forum and I will park it somewhere else and let you know later.

    I am not sure how familiar you are with Python. It is possible when trying to run this it will alert you some libraries are missing. Usually this can be remedied by command line "pip install" instructions.

    You can change some entries in the code like the server name you want to monitor, start frequency and zoom level, waiting_period, time_span_minutes.

    The current set-up uses values that I used for looking at effect of filters etc. and covers 7 - 22 Mhz range which you can change of course.

    You can remove the plotting of MinSig by commenting out the line with ax2.plot(xdata, y1data,'g-'). MinSig which is a proxy for the noise level over this wide frequency range.

    Currently the time span is 40 min with the idea that you can switch through your 8 antennas at g8jnj.zapto.org and record 5 minutes of each and show it in one plot. Actually this switching could futher be automated by using /?ext=ant,n and when re-evaluating every hour the antenna switch could be held at the antenna with best overall SNR over the frequency band of interest. This may not correspond to what a listener would like since if not browsing one could be mainly interested in reception in a certain direction.

    The plotted values are noisy not in the least because signals over such a large bandwidth are resulting from reception of signals coming from many directions with constantly varying propagation and possibly switching on and off.

    I tried various filters when plotting but this did not help much and of course slows down the response. So this version is unfiltered.

    Kindly do not share the code at this time. In case you have remarks or installation problems let me know what Python version and which IDE you are using. If really necessary I can probably re-pack the code so it can be run as a windows exe file, but that would require adding a GUI first in order to be able to alter parameters. To much work frankly...

    Best regards,

    Ben

    PS: Switching around it turns out that this good old Datong 370 dipole shows nice performance in terms of SNR.

    PS2: Ptyhon code is at http://sparks.epizy.com/SNR/Wideband_SNR_Martin.py

    January 28