SkyPipe Calibration
at the
NJ3B Radio Observatory

Jim Brown, NJ3B

Last modified 12/31/2009

Why calibrate?

Data collected in the Radio JOVE program can be submitted to the JOVE data archive.  This data is freely available to students and researches alike. 

Uncalibrated data is useful in that it shows the Jupiter or Solar burst relative to the background noise but it doesn't convey the true intensity of the burst.  Uncalibrated data between, for example, two JOVE sites can't be compared. The lack of comparability between these uncalibrated sites reduces the true value of  information contained in the data sets.

Calibrated data sets show a true value of the intensity of the burst at your location.   With calibrated data, values of bursts can then be used in research.  With known values to work with, a researcher can then use YOUR data in their research programs.   Just think about that for a second.  With the calibrated data sets we submit to the JOVE Archive, we are making a REAL contribution to radio astronomy!

Such studies include, but are not limited to:

• The study of decametric radio emission from the planet Jupiter.
• Studying very-fine structure in the dynamic spectra of the Jovian decametric emission.
• Developing models explaining certain types of lane features present in the decametric radiation from Jupiter.
• Determining a more precise value of the Jovian rotation period.
• Determining the shape of the emission beams from Jupiter by studying how the declination of Earth affects the observations.
• Earth ionospheric study; the propagation of radio emissions at several locations based on ionospheric transparency or translucency.

Calibration

SkyPipe software allows the user to take data from a receiver and manipulate it so that the data plotted is the true value of the intensity of the burst.  This value is, by convention, is displayed in "Degrees Kelvin".  This is also called the "Equivalent Antenna Temperature".  A more detailed description of antenna temperature can be found at: http://en.wikipedia.org/wiki/Noise_temperature.

The Procedure

The following steps can be made with either a sound source going through your sound card in the computer, or, as in my case, an A/D converter and a sound source from the receiver straight to the A/D converter that is then connected to my computer via a printer port.  Systems can vary and it's not in the scope of this tutorial to cover them all.   A more complete description and information can be found at: http://www.radiosky.com/skypipehelp/skypipehelpindex.html.

In order to calibrate your data, you need a known value, or a series of known values displayed on SkyPipe so that you can compare the known value to what is actually being displayed on your SkyPipe chart.

One such known value noise source is the RF2020 noise source generator manufactured by Richard Flagg (rf_at_hawaii_dot_rr_dot_com).

• Connect the noise source generator to the antenna input on your receiver.  If you're using a JOVE receiver, you may need an adapter to accommodate the "F" connector on the JOVE receiver.
• Make sure the noise source generator is set to the lowest value, then turn on the noise source generator and let it and your receiver warm up for a few minutes. 
• Start SkyPipe charting.
• Allow the trace from the noise source to run for about one minute, then switch to the next position on the noise generator.  Repeat this for all the settings. 
• When you're finished, stop SkyPipe and save the file.  Call it anything you like such as "calibrate".  It should automatically be saved in your SkyPipe directory.
• Load the "calibrate" .spd file into SkyPipe and you should see something like this:

 

 

• With the file loaded into SkyPipe each temperature level (step in noise level) is averaged using a special feature in SkyPipe. 
• Use the zoom feature (a dashed line box with a Z in it) and drag a box around the first step. The first step will now fill the entire screen of SkyPipe. 

 

 

• Put your pointer on the data line (anywhere) and right click and select "Get Avg for View". 

 

 

• The average Y value for the visible portion of the chart will appear in the status bar. 

 

 

• Repeat this for all the steps in your chart. 
• Those numbers are then entered into a spreadsheet.  It does not matter if you use Excel or, in my case, KaleidaGraph, the procedure is the same.

The values produced for this demonstration are as follows (your values will differ):

• When entered into a spreadsheet and plotted, you would end up with a graph like the one below (it is beyond the scope of this tutorial to walk through the procedures.  Just remember that this is all part of the graphing function of either spreadsheet):

 

 

• Using a feature of KaleidaGraph called Curve Fit (in Excel as well), I applied a 3rd order polynomial to it and it produced the equations where X is the value for the current data in SkyPipe:
  

20 MHz = 8.5948+.35016*X+.0015232*X^2+.000000627X^3

• Next, I went to the SkyPipe formula entry section (under Options, Data Source then Equations) and set up the equations as follows:

For Channel 1 (20 MHz), A=X;B=X^2,C=X^3;8.5848+(.35016*A)+(.0015232*B)+(.000000627*C)

• The following table shows the same invormation as above, but with the calculations in place in SkyPipe.

As you can see from the table, with the calculations applied, the results are very close to the actual values and the difference is negligible.

Accounting for loss

• The last step is to figure in the loss in db in your system.  There will always be some loss of signal due to the length of coax and anything between the antenna and the receiver.  Such items as a bandpass filter, lightning arrestor, etc.  It's important to measure all this to determine an accurate measurement of signals received.
  

Example:

• If you have access to an MFJ antenna analyzer, you can measure the loss of the coax used (if 50 ohms) directly as well as other components you might have between the receiver and antenna.  If not, you can calculate the loss from this table.

For argument sake, let's say that I've accounted for all the loss in my system:

• To account for this loss in my system, I modify the calculation (above) to take in to account the measured loss.  My new equation looks like this:

A=X;B=X^2,C=X^3;((-8.5116)+(.91888*A)+(.0041002*B)+(.0000024432*C))*(Log(4.07).    (Remember this is a log scale, hence the Log(4.07)

• With this addition, your output trace in Skypipe should be calibrated to accurately display the equivalent antenna temperature in degrees Kelvin.

Return to the previous page

Web design by JLB Consulting