Society of Amateur Radio Astronomers   Search Site:

---

Educational Programs

Success at Last - Exploits of a Novice Radio Astronomer
By Jon Wallace, email fjwallace @ snet.net

Unfortunately, my knowledge base was still too low to understand the nuances of radio electronics design. Looking back I now realize that I was afraid to solder sensitive parts and probably had poor solder connections. I have also learned that extra wire is a big problem at these frequencies, becoming little antennas and adding tremendous noise to the system. I still find that the concept of ground, although theoretically simple is, and always has been, a problem for me. I experimented for a long time with shielding and special wire (9913 cable, etc.) but still continue to try to learn this crucial topic. I next met Paul Shuler. He lived nearby and had a working radio telescope which did amazing things. He helped me start from scratch, testing each piece before we went on. We worked for a couple of years and I had some success building amplifiers and simple devices but I started getting frustrated with the length of time it was taking. About this time I heard about Jeff Lichtman's 'ready-to-use' systems, so I bought one.

The excitement of seeing my first solar scan come by is indescribable! I was hooked. After a while, I noticed that heat was a big problem and day versus night scans of the same regions varied tremendously. I started experimenting with several temperature control systems (currently I use a 'Supercool' system with a fuzzy logic controller from CSE). I also wanted a way to standardize my scan charts so I got a calibrated noise source. I have tried numerous configurations of my equipment. Currently I have the signal from the horn fed through a length of 9913 cable to all my equipment which is housed (on the ground) in a temperature controlled cooler. The output is fed to George Jones' Datalog A/D and recorded on a 386 laptop. I feel I now have as stable a configuration as I can get.

Over the past three years I have been interested in doing a full sky scan. This required doing lots of 24-hour scans over various seasons. After looking at the data I decided that I would need about 10 scans of each altitude I could observe. I have a small backyard with a limited field of view - thus my scans would be from 25 - 70 degrees - every 5 degrees and my azimuth was required to be about 140 degrees. Needless to say, I didn't do my scans every day. I wanted to be around and therefore I limited my observations to vacations and weekends when there were no storms. After two and a half years I started putting the data together. I first adjusted all data by using a value for the noise source that I chose to be about average and multiplied all chart values by a correction value. I then contacted Malcolm Mallett and he told me I could use his Python and Gnu programs. I adapted my data to his format and was able to do averaging and spike removal quite easily. I then averaged the data in Excel and made my first graphs and combined them into one graph to view similarities and point out problem areas. I decided I needed a few more 24-hour scans at each altitude.

While I was performing the scans I decided to go back and analyze my antenna aperture and pointing. I learned from Paul Schuler that you could use the sun as a pretty good point source and plot the width of the solar plot (left and right of center) versus the number of degrees away from the actual/calculated declination. I found that my antenna is a little asymmetrical and my pointing is off by a couple of degrees (see chart).

This fall I finished my additional scans and was pleased with the improvement in the Excel graphs (see graph below), but I decided I wanted to make a 3-D representation of my data. I looked all over for a program and finally settled on Mathematica. The programming of Mathematica was pretty difficult (because I got the discount version with no manual), so I purchased 1 year of 'support'. The gentleman I was teamed with turned out to be an astronomer and was interested in my project and provided me with help in designing the programming to produce the graph.

It took nearly two months to get exactly what I was looking for but we ended up with a great picture of the sky (see below). During this entire process I was sharing what I learned with my students. I used my understanding to build about 20 demonstration devices for showing the physics of non-optical electromagnetics and I now have a radio picture of the sky for students to look at. It's a different view than what they typically see and I have years of interesting and frustrating experiences to share with them and others. In fact, I tell all my optical astronomy friends that their 10-minute picture of the galaxy from their backyard took me three years to make - 'I had to work for my pixels'!

Directions for Analyzing Data (As I Did It):

I) Import *.dat file into Excel - click Delimited then Next - on the next screen click Comma and Next - click Do Not Import for all columns except the '140:00:00' (azimuth) column and the '1420' (frequency) column. (See below two spreadsheet samples.).

(Note: The 1980 is actually 2000, my old computer wouldn't roll over to 2000 choosing 1980 instead.)

*.dat file

Imported Columns from the *.dat file above

Next, look over the chart and deleted any spurious signals (usually associated with the sun and usually killing about 6 hours of data!).

Next, lookup the calibration numbers for all the charts and pick a representative value (the 'standard'). Compare all charts to this and multiply by the correction factor (divide the chart's calibration number by the standard - this is the correction factor). The third column shows the data after multiplying by the correction factor - in this case, 0.98 (To do this, press the '=' and the click on the original column and then type *.98. Copy this cell and paste the formula down the third column. When you are sure the data is correct, click on the third column to highlight the whole column, then copy and paste special (as values) to change the formula values to real numbers. You can then delete the second column

Then cut and paste the data onto a spreadsheet for right ascension from 0 to 24 hours - each usable chart's data pasted next to the last. Add them and divide by the number of charts with data.

Sometimes you'll have to go back and 'adjust' the data so that they line up properly (smoothly) all the way along the chart. (Even with a calibration, the charts don't always line up exactly on different days and different temperatures and weather conditions). The chart below shows the right ascension, followed by 5 columns of data and then the totals column (adding the rows up for a total) and then the average column (dividing the total by the number of columns of data)

Cut and paste a sheet with just R.A. and data. Add the declination in a column as shown below.

Run this finalized chart through Malcolm Mallette's programs by copying the data column from the Excel spreadsheet and pasting into the Python 'template'. See below:

# Sampling: 1 seconds per sample
# Total Time: 2h -> 7200 seconds.
# Number of Samples: 7200.
# start time: Fri Feb 26 19:30:09 1999
# DEC = 22
# colltime = 2h
# comments = Taurus A with full RF gain and no attenuation. Partly cloudy.
# ext_gain = 25
# freq = 1230
# gain = 50X
# inttime = 10
# rate = 1
# receiver_gain = 10.5
# starttime = 7:30 pm

Malcolm Mallette's 'template'

Paste your data in the data column (29.3 column) and paste additional 9.2E+08, 0.449219, and seconds (the 0, 10, 20, ... column) to match the length of the data set.

Open the Python program and run the smoothing and spike removal parts of the program. Copy the data column back to the original spreadsheet in Excel. This completes one declination - (I did 10 altitudes from 25-70 degrees and did 12 - 24 hour scans for each one to get enough data to generate a decent 'picture').

After all declinations are complete, cut and paste them next to each other as shown below (R.A. and the ten declination data sets) to generate a chart of the 10 declinations overlapped.

When done, save as a *.txt file so that Mathematica can read it. Run the Mathematica program. You'll get a picture like this one:

---

[ Home ][ Admin ][ Educ ][ Meetings ][ Membership ][ Photos ][ Projects ][ Pubs ][ Search ][ Contact ] entire website copyright © Society of Amateur Radio Astronomers this page last updated 20 Jun 2009 email the webmaster

Top of Page