A Simple Method to Predict the Performance of your small 21cm H Line Radio Telescope.

[Alex P]

Here is a simple method to predict the performance of a small radio telescope system by   measuring the differential between the 10K Cold Sky and 290K Earth Noise.   

      a)  Point  the Antenna overhead and recordi a background reading off H Line freqs

      b) Point  the Antenna Towards the Ground ( or using a 50 ohm SMA load into the LNA ) and record the value

      c) Compare the differential between the two.

       This is a simplified method of calculating Tsys Equivalent System Noise Temperature.

      What is useful is that this differential is ~ 3-4 times greater than the   Brightest regions of the Milky Way H Line regions …

            Restated, A measurement of  the Hot_Earth vs Cold_Sky can be used to        Predict the Performance of your 21cm H Line Radio Telescope.

      AND, If you make a change to your System which improves the Earth<>ColdSky difference,    The H Line Performance will improve proportionally

       ( the plots illustrate this )

       =================================================================

 

I have verified AirSpy SDR# Display and Cursor Readout “dB” readout units is Correctly Scaled    ( not in absolute power level, but that a 20 dB change in signal level == 20 dB change in readings )

 

If you use another acquisition software, Test It.

 Verify the Software Units.  Use a few SMA Attenuator steps to  Certify the “dB” values are actually Calibrated in dB.

 

Perform a Cold_Sky<>Hot_Earth measurement prior to a test series   or at least write down the Cold_Sky value before each test .. If it is not within a dB of normal, identify the issue.

      Ideally, you want each data set to be similar so that they can be compared or potentially combined.

 Be Meticulous. Take Your Time. 

Document the parameters for each test for future reference.

 Once you have established a good routine :

Set up the hardware and software the same way each time : Be Consistent

 Alex Pettit

Project HLine 3D

[b alex pettit jr]

Tsys Calculation

[b alex pettit jr]

dB values = H Line pk @  Dec + 40 Dg, RA 20:30 Hrs  referenced to Cold_Sky

[Stephen Arbogast]

Hi  Alex,

I have been exploring   your  Java  software some more  and  now a problem....  When I run JA_IFAvg2csv_varMXB_intLogo_11Jan2501  against a new collection of  raw  data  the  app hangs  and no error  messages.   Is there  a log file  I can check for  error messages? 

Tnx,

Stephen

[b alex pettit jr]

Stephen,

Zip up the file set, email it to me & let me see what the issue may be ..

I ran over a hundred data sets to validate the integrity of the software.... & this included  data from 12-15 people.

and with  IFavg files that use a comma to delineate the decimal point. 

Alex

=================================================================

[b alex pettit jr]

The "Background.txt" file does not follow the IFavg format ..

Delete it from the data set and all will process fine.

-

[b alex pettit jr]

I am uncertain of your familiarity with the software ...

This is the Raw data set, unmodified

This is the "Drift" error .. but there are issues with the last several files ..

Delete them from the data set and try reprocessing

229 files processed

Some instability in the amplitude of the data began ~ 10 hours into the acq set

This is the mx+b "corrected" data set at 25% & 75%    ( not typical data )

[Stephen Arbogast]

Some  constructive  criticism.....   would  like  to    some  logging  of  errors  in future  releases of  your  software ............   

[Stephen Arbogast]

Back in my  day  we  used Log4j  for logging  in  Java.   I have not  used  Java  for many years  now.  I  prefer   Python..   fine  for  post  processing  data.......    Just  a thought..  a  good  idea  to include  logging   in any  language......  just  a comment!

Stephen

[b alex pettit jr]

Here is the Data 2D&3D

What was the Dec/RA of this  slice ?

Pk HLine above Cold_Sky = 1.0 dB

Large RFI spike at 1420.00MHz 

Noise across the spectrum

[b alex pettit jr]

Performance Comparison Plots re-post