RTL-SDR NOAA Satelite Update
So the last time I wrote about my adventures into the world of software defined radio satellite reception I was dealing with some local FM radio stations really messing with my radio's ability to get a clear sat signal. This proved to be a difficult challenge to overcome as the same interference would be present, and sometimes stronger even if I went to the roof of a building to take the measurement. Nevertheless I got some nice images which are in prior posts.
One of the advantages of software defined radio is that the data is read through a computer and so the radio can be made easily accessible by connecting it to a server. Many hobbyists have set up publicly accessible radios by doing this and some of them are very good resources for those living in areas with poor reception or lots of interference. You can find many of these at http://websdr.org.
One radio in particular, http://k3fef.com:8902/ is equipped with a quadrifilar helicoidal antenna much like my own and is scanning the APT downlink band for NOAA sats. I think you can see where I'm going with this.
I plugged in the radio base station location into gpredict to determine at what time the satellite would be passing overhead and managed to capture the audio with the built in recording function of the web server. Once decoded in noaa-apt this was the resulting image:
Alright so it's not much better than what I could get here in Pasadena but it was taken without removing me from my room, a somewhat desirable quality. Some of the biggest issues with this method seems to be the sampling of the audio on the server end, those ellipsoid shaped striations in the image I believe are small sampling or compression issues but resampling the audio (again with noaa-apt) to 11025 kHz was not able to remove them.
Something that I've seen people in nice quiet areas do is to put the QFH antenna in their attic and then receive from indoors. This is something I would like to do because at the end of the day if I'm getting useful weather information from these satellites I would like it to pertain to where I am.
To do this I need to filter out all the other nonsense polluting the radio's input. This is traditionally done with some sort of filter and low noise amplifier configuration which is something I would like to make on my own if at all possible. After finding the schematic for a simple band pass filter (two coupled LC circuits) and then learning that for RF applications a capacitor can be constructed from two twisted wires I decided to give it a go at making one.
This is the one that I made, you can see three capacitors. Two are each in parallel to an inductor one side of which is sent to ground and the other side of which is connected directly to the input/output of the filter (I did not include the dc coupling caps because my radio already has dc coupling and with a twisted wire capacitor I was worried that inconsistent twisting or inadequate capacitance might have unwanted effects.
I'm currently attempting to tune the filter to the frequency and bandpass that would be appropriate for NOAA satellite reception but this is challenging. My approach involves using the 'relative' mode of the open source spektrum (SVmod v01.9a) spectrum analyzer software. I short the signal path and take a spectrum which is then 'zeroed' in the software after which I clear the signal path and look at how the baseline of the spectrum changes. This picture shows a very broad look at the full range of the radio but remember I want a narrow bandpass of ~3 MHz centered at 137 MHz.
The blue trace shows the 'baseline' without any filter (shorted in/out) and the yellow trace shows the difference spectrum with the filter connected. Clearly certain bands are attenuated rather well however I'm not super certain where to go from here as it would be nice to have a clearer change in baseline. I'm looking at the possibility of changing the parallel caps to some non polarized capacitors I found lying around that are in the microfarad range instead of the picofarad range. These would not be ideal but I'll mess around with an online calculator to see if an inductor can be made to filter the desired range with these.
Thanks,
Alex Krotz
One of the advantages of software defined radio is that the data is read through a computer and so the radio can be made easily accessible by connecting it to a server. Many hobbyists have set up publicly accessible radios by doing this and some of them are very good resources for those living in areas with poor reception or lots of interference. You can find many of these at http://websdr.org.
One radio in particular, http://k3fef.com:8902/ is equipped with a quadrifilar helicoidal antenna much like my own and is scanning the APT downlink band for NOAA sats. I think you can see where I'm going with this.
I plugged in the radio base station location into gpredict to determine at what time the satellite would be passing overhead and managed to capture the audio with the built in recording function of the web server. Once decoded in noaa-apt this was the resulting image:
Alright so it's not much better than what I could get here in Pasadena but it was taken without removing me from my room, a somewhat desirable quality. Some of the biggest issues with this method seems to be the sampling of the audio on the server end, those ellipsoid shaped striations in the image I believe are small sampling or compression issues but resampling the audio (again with noaa-apt) to 11025 kHz was not able to remove them.
Something that I've seen people in nice quiet areas do is to put the QFH antenna in their attic and then receive from indoors. This is something I would like to do because at the end of the day if I'm getting useful weather information from these satellites I would like it to pertain to where I am.
To do this I need to filter out all the other nonsense polluting the radio's input. This is traditionally done with some sort of filter and low noise amplifier configuration which is something I would like to make on my own if at all possible. After finding the schematic for a simple band pass filter (two coupled LC circuits) and then learning that for RF applications a capacitor can be constructed from two twisted wires I decided to give it a go at making one.
This is the one that I made, you can see three capacitors. Two are each in parallel to an inductor one side of which is sent to ground and the other side of which is connected directly to the input/output of the filter (I did not include the dc coupling caps because my radio already has dc coupling and with a twisted wire capacitor I was worried that inconsistent twisting or inadequate capacitance might have unwanted effects.
I'm currently attempting to tune the filter to the frequency and bandpass that would be appropriate for NOAA satellite reception but this is challenging. My approach involves using the 'relative' mode of the open source spektrum (SVmod v01.9a) spectrum analyzer software. I short the signal path and take a spectrum which is then 'zeroed' in the software after which I clear the signal path and look at how the baseline of the spectrum changes. This picture shows a very broad look at the full range of the radio but remember I want a narrow bandpass of ~3 MHz centered at 137 MHz.
The blue trace shows the 'baseline' without any filter (shorted in/out) and the yellow trace shows the difference spectrum with the filter connected. Clearly certain bands are attenuated rather well however I'm not super certain where to go from here as it would be nice to have a clearer change in baseline. I'm looking at the possibility of changing the parallel caps to some non polarized capacitors I found lying around that are in the microfarad range instead of the picofarad range. These would not be ideal but I'll mess around with an online calculator to see if an inductor can be made to filter the desired range with these.
Thanks,
Alex Krotz
I like how you refuse to look out the window to see if there's a cloud in the name of science.
ReplyDelete