Small Hydrogen Line Radio Telescope Background Drift/Shift Correction Techniques
Alex P
Small Hydrogen Line Radio Telescope Background Drift/Shift Correction Techniques
1) The Dicke-Switch Technique works by switching the input of the radio-telescope LNA between the Antenna and a Temperature Reference Load. Gain variations can have their effects corrected by measuring the difference in signals between the Antenna and the Noise Level of the Temperature Reference Resistor. The Gain Correction Process should be implemented by Division. It requires added components in front of the LNA (coax switch or internal switch) which degrades the system noise figure. Only Gain Variations can be detected / corrected by this technique.
2) 1423 MHz Frequency Band-Switching probably dates from the Spectra-Cyber Era when stepped frequency analysis was the norm such that a 256 point freq spectrum might have required 256 seconds, but it had poor S/N as each frequency was measured for only one Second. As a sample at 1423 MHz 'would be beyond the H Line ', another one second made little difference to the total acquisition time. However, when used with an SDR, this consumes 50% of the acquisition time. Since the variations are primarily background noise level shifts, the process should be implemented by Subtraction . This technique will allow correction for All Variations in Background Shift ( drift )
3) In-Band Cold Sky Normalization can be obtained by selecting a small spectral region Not influenced by Hydrogen emission spectra. That value can be obtained from within normal Non Frequency Shifted spectral data samples. It consumes no additional acquisition time. Since the variations are primarily background noise level shifts, the process should be implemented by Subtraction. This would typically be a Post-Acquisition Process in order that the lowest noise frame be determined and used for data set normalization.
This technique will allow correction for All Variations in Background Shift ( drift ).
Suggestion : A few Members with familiarity with GNU_Radio / Python might team up and try implementing an IF_avg type M curve gain correction for data acquisition in GNU_Radio and Cold_Sky_Normalization drift correction post test processing in Python. This way, a standardized process can be used to compare systems or adjustments.
As a reference, use SDR# > IF_avg > HL3D to validate your code: It Works.
Once a stable, consistent data set is obtained, further processing such as filtering, smoothing, or VLSR correction can be applied.
