Welcome to Issue #7 of the Radio-Sky Journal
September 2002
Copyright 2002 by Radio-Sky Publishing
All rights reserved.

Estimating Signal Flux at Other Frequencies 
Jupiter Season is Here
Radio-Sky News
Amateur Tip #9 - Baseline Jump Culprits

Estimating Signal Flux at Other Frequencies 

A recent post to the SARA listserver asked the question, "If I know the signal
strength of a cosmic source at one frequency, how can I calculate the strength
at another frequency?" In past issues of this publication, we have discussed how
cosmic radio sources emit a variety of different spectra. Some sources are best
characterized as "thermal" sources which emit radio waves which increase in
intensity in proportion to frequency ( at least up to some point ). Others display
the opposite spectra, where the signal is stronger at lower frequencies. These are
often "synchrotron" radio sources. There are other sources which display very
strange spectra which do not fit these simple models. So how do you make an
intelligent guess about the signal strength (flux density) for a frequency for which
you have no published survey?

First, lets clarify that this discussion is about cosmic sources which remain
rather steady in the amount of energy that they emit. Sources like the Sun, and
Jupiter are excluded. We are talking about sources such as supernova remanats,
quasars, and active galaxies which stream radio waves rather steadily. While some
pulsars do produce occasional "burps" of extraordinary power, for the most part,
their average peak emissions are rather predictable and so this discussion could
also apply to them.

The jansky is the unit of measurement that we generally apply to cosmic radio
sources to indicate the amount of power that we receive from them. The unit is
named after the acknowledged father of radio astronomy, Karl Jansky, and refers 
to the integral of the "brightness" over any "discrete" radio source. A discrete
source is one that has well defined boundaries and brightness is defined as a
measure of the power recieved per unit area of our antenna, per unit area solid
angle that the source subtends in the sky, per unit of bandwidth. Wow.. That
gets a little mind boggling, (at least for me), so I tend to think of the jansky more
simply as measurement of how much energy I can expect to receive from a source.
If I need to convert from janskys to some other measurement, then I put on the
thinking cap, and look at it more analytically. For this discussion, the simple
conceptualization will do. You will also hear the jansky refered to a unit of "flux
density", and that is the general term we will use here.


Not long after World War II, scientists armed with surplus radio receivers and
antennas began looking at the variations in strength of radio signals received
from different portions of the sky. It soon became obvious that there were many 
sources of radio waves that were quite small in angular extent. As these things
go, competition soon arose as to which scientific group could map out the sky
in greatest detail, mapping and measuring these small radio sources. Catalogs
were created which listed the position and strength of the sources. Surveys
were carried out at different frequencies, and a picture began to emerge of how
the indivdual sources varied in intensity when viewed at different wavelengths.

Most of these survey catalogs are available today on the internet. In them, you
will find tables that list objects determined to be discrete sources. You will
typically find tables with rows similar to this one from the 4C (Cambridge)

-05.55 12 53 34.5 -05 28.6 20.9 1 305.1 +57.1 a 3C279 

In a separate file you will find an explanation of what appears in each column:

Byte-by-byte Description of file: radio4c.dat
Bytes Format Units Label Explanations
1- 8 A8 --- ID *The 4C number
11- 12 I2 h RAh *Right ascension equinox 1950.0
14- 15 I2 min RAm R.A. in minutes
17- 20 F4.1 s RAs R.A. in seconds
23 A1 --- DE- Sign of declination equinox 1950.0
24- 25 I2 deg DEd Dec.
27- 30 F4.1 arcmin DEm Dec.
34- 40 F7.1 W/m2/Hz FluxDen *Flux density (1.0E-26 MKS) at 178 MHz
43- 44 I2 --- ErrClas1 *Position and flux error class (in number)
48- 52 F5.1 deg GLON Galactic longitude, System II.
55- 59 F5.1 deg GLAT Galactic latitude, System II.
61- 63 A3 --- ErrClas2 *Position & flux error class (character form)
65- 72 A8 --- com *Remarks

I know these columns are not going to line up properly in the 
table above, but I wanted you to be aware of the type of format
issues you will be contending with, so I just cut and pasted this from
the original... Oh, I did add DOS type carriage returns. All of the
catalogs use UNIX type text formatting and so when you import them
to many word processors, you are going have to convert the carriage
returns so that everything doesn't appear on a single line. 

I am trying to make the point that the survey tables are not user friendly
by today's standards, but the info is free and out there on the internet.
The clumsiness of working with these tables several years ago prompted
me to try to import some of them into a more user friendly format. The
result was a Microsoft Access database of several catalog tables which
I include on the Radio-Sky CDROM I. This makes for much less effort
when researching a particular source. The CDROM also contains a 
number of the surveys formatted into Excel spreadsheets.

Back to the Problem

So let us say you have looked through the catalogs and found no
reference that supplies you with what flux density you should expect
for Cass A at 12 GHz. What do you do? You can use the "spectral
index" of the source to estimate the flux density of Cass A at the
frequency you want. Basically, the spectral index (a) describes the
proptionality (:) of the variation of flux density (S) with wavelength (L)


Estimating the flux density at given frequency can be done if you know
the flux at another frequency and the spectral index of the source. Take
a look at section 8-3 of Kraus's Radio Astronomy 2nd Edition. (I have this
book for sale on the website, if you don't have a copy or don't have a copy
in your local library.) Non-thermal sources generally have a positive
spectral index, meaning that their flux increases as you view them at
lower frequencies. The spectral index is simply :

a = log (S1/S2) / log (F2/F1)

where S1 is the flux at some frequency F1 and S2 is the flux at some
higher frequency F2.

Unfortunately, not all cosmic sources have flat spectra and so it is common
practice develop a curve which better defines the relationship between
frequency and flux density. This can be done with as few as three widely
separated flux measurements. 

As a practical matter though, if you have the spectral index and a flux
measurement of the source at some frequency NOT TOO far removed 
from the frequency for which you wish to calculate the flux, you can plug
the values into the equation above and solve for S1 or S2 to get your answer.
Cass A has a relatively flat spectra and a spectral index of about 0.75. 

I cheated a little here. I used my old DOS program Radio-Sky Planetarium
to get the spectral index and to get a flux value for a not too distant
frequency of 3200 Mhz. You could also find flux values by searching through
the various radio source catalogs on the internet. Some multi-frequency
catalogs also publish a spectral index for the source. The Radio-Sky
Planetarium program computes a high and low range spectral index if it
has enough data available for the source. If only a single frequency is
available it assumes a spectral index of 0.8 as this is the most common
value for the synchrotron sources that amateurs are likely to look at. This
assumption can lead to a wildly wrong estimated flux so in important
situations, you should try to verify the spectral index by searching the
catalogs and calculating the index yourself. OK, given:

F1 = 3200 MHz
F2 = 12000 Mhz
S1 = 1300 janskys at F1 (determined from table or program)
S2 = unknown flux in janskys at F2
a = spectral index of 0.75

0.75 = log(1300/S2) / log(12000/3200)
0.75 = log(1300/S2) / log(3.75)
0.75 = log(1300/S2) / 0.574
0.43 = log(1300/S2)
10^0.43 = 1300/S2
2.7 = 1300/S2
S2 = approx 480 janskys

Of course, we need to be really careful when we get up into the microwave
region to account for things like atmospheric effects, but the method does
yield a rough estimate that we can work with.

Note: The Radio-Sky Planetarium is also available on the same CDROM 
along with the reformatted sky catalogs.


Jupiter Season is Here

Time to break out the old shortwave receiver. Jupiter is back in business as it
swings far enough away from the Sun to make observing practical again. We are
hoping that this will be an exciting year. Hopefully, the receding sunspot numbers
will allow for a better ionospheric conditions. At the same time, from Jupiter's
point of view, the declination of the Earth is moving us away from the cone of
signals she emits. So who knows what will happen? 

This year brings us expanded global coverage of Jupiter, with new observatories
in Nice, France ( Ruggero Ulivastro ) and Pforzheim,Germany (George Lauffer,
DC1GEL ). Rob Carver should be serving data from Australia, and there are 
several new observers on the US mainland. Here in Hawaii we have an exciting
new instrument at the Windward Community College Radio Observatory on
Oahu. This new spectrograph will scan the 10 Mhz wide region from 18 to 28
MHz, 10 times each second. The instrument is still in development, but we hope
to release viewing software for the real time spectrograph in the near future. 

There is a big gap in our Jupiter coverage that hopefully someone will be able to
fill. Despite huge numbers of capable people in Japan, India, and cental Asia
no one has of yet been sharing data via Radio-SkyPipe from those regions. If
you know of anyone who might be interested, please give the a nudge!


Radio-Sky News

Radio-SkyPipe Version 1.2.7 Available

This is a big leap forward. This release adds ADC supoort for Win2000 and Xp.
Pro users will get up to 9 separate client Windows. ( I was watching charts 
simultaneously from seven observers the other day during a coordinated
observation). You can open and review up to 5 data files at the same time in
separate windows. Greatly expanded observer log functionality. There is just
too much to list here. 

Of course, even while people are getting upgraded to the 1.2.7 version, another
version is in the works. In this version we have Sidereal Time logging, integer
file saves for reduced file size, and an exciting new concept called the User 
Data Source (UDS). The UDS is an open source code project that allows
anyone with a little programming skill to contribute to a pool of modules which
will allow SkyPipe to use virtually any ADC or other data source available. I 
have already written a couple of UDS modules for Radio Shack and Metex 
computer enabled DMMs. One interesting aspect of this concept is that the
module that collects the data does not have to be on the same PC as SkyPipe.
Thus, over a LAN or the internet, you can establish connection to a UDS
module collecting data at some remote location. The UDS program can be
very small, and it is hoped that future Java, Delphi, and C versions will allow
data collection via a PC using any operating system. Conceptually, the UDS
allows for hardware generated timestamping. Thus, it may be possible to
get highly accurate timing with the right hardware. 

Goto: http://radiosky.com/skypipeishere.html


MAX186 8 Channel ADC Kits / Units Available Soon

This same headline was in the last issue of the RS Journal and indeed many kits
and units were sold. This is just a heads up that we are currently out of
MAX186 chips, but Maxim has promised us a shipment in mid November
of 2002. A recent article in Oct. 2002 Nuts&Volts magazine is partially to blame
for the depletion. You can still place an order now if you don't mind waiting a
couple of weeks. You won't be charged until it ships.

Amateur Tip #9

Jumpy Baseline Culprits.

It seems that more often than not when I set up a radiotelescope experiment, there
will be places on the chart where the baseline makes sudden jumps. It can jump
down or up and the duration can last from seconds to many hours. If you see 
sudden jumps in your chart you can be pretty sure its not something good. Real
celestial object signatures will provide smooth transitions in signal strength as
they enter your antenna's beam. Here are some things that I have found over the 
years which can cause the jumpy baseline phenomena.

1. Broadband interference from motors, light dimmers, televisions, and computers.
This is not usually too hard to find unless its in a neighbor's house. Some types
are hard to hear through a speaker and sound much like natural cosmic hiss.
If the broadband interference is coming through the antenna you must either
null it out with antenna phasing or pointing, eliminate the source of the
interference or shield it, or move to a new location.

2. Out of band inteference. This can be a interfering source of most any limited
bandwidth. What makes it so terrible to find is that it may be almost anywhere
in the spectrum but just be powerful enough to overpower your bandpass
mechanisms and sneak into your receiver. Consider the fact that it may sneak
in over paths other than the antenna and transmission line. I have seen this
type of interference manifested as both jumps up and down on the chart. The
downward jumps are probably caused by the signal desensitizing the front
end components. Fight this with strong bandpass filtering and careful bypassing
of power and control leads. Also make sure you have a clean local oscillator
signal and a strong mixer stage when possible.

3. Bad connections. RF connectors are notorious beasts. You must be sure to
check them carefully. Sometimes a temperature change over the period of a
day can switch in and out a connection as sharply as a relay. Tiny 
movements caused by wind and room vibrations can have the same effect if
the connector is barely making contact. Solder them if possible. Solderless
connectors can be quite good, but test them if you are having problems. 

4. Oscillating stages. Amplifiers can turn into oscillators quite easily. Its a
common problem even at the NRAO. One thing that they do besides the
extraordinary design engineering and layout is to make sure each stage is
properly terminated in the right impedances. They think nothing of throwing
away a little gain if it is going to be made up for by the added stability and
so often you will see amplifiers chained together with rather strong 
attenuators between them. The attenuators assure that amplifier sees a
50 ohm load on its input and ouput. You might have to add extra stages of
amplification to make up for the "padding". Don't try to get too much gain
out of your DC op amps either. Use extra small gain stages if you have to.

5. Powerline voltage fluctuations. Here in the backwoods of Hawaii I can tell in
my observatory if someone uses the toaster in house! Double regulate your
power supplies, or even triple regulate them! Don't use them near the limits
of their specified current handling capability. Battery power can be a real
help in some situations. It can also free you from interference traveling in
over the AC mains.

Well those are the jumpy baseline culprits I have found. If you have any more 
let me know and we will post them next time.