Amateur SETI: Project BAMBI

The Project BAMBI Team:

(left to right)
Mike F., Bob Lash, and Mike Fremont.

Example BAMBI Waterfall Drifting CW Detection (Terrestrial Origin)

Can you find it? A faint 3.8 GHz drifting continuous wave (CW) narrow band signal is visible below. Although this signal was of terrestrial origin (you are looking at the 30th harmonic of an ovenized 128 MHz TTL oscillator soldered to a resistor lead, mounted two yards BEHIND the dish for high attenuation, mixed with actual sky signal at RA 19:46:05  DEC +40 43), we are searching for real signals that have a similar drifting CW characteristic due to possible doppler shift.

The horizontal axis of this display block corresponds to 600 channels (1 Hz per channel), and the vertical axis represents time (240 lines at 2 seconds per line). The slope of this signal is approximately + 3 Hz per second, with an inflection reversing to - 3 Hz per second. Unlike this test signal, a real extraterrestial source would be expected to maintain a constant slope.

BAMBI Sites:

SITE A: California. SITE B: Colorado.
Approximate Antenna Separation: 1,000 miles

BAMBI Overview:
 
 

The following article, complete with figures, was originally published
in the June/July 1994 issue of Radio Astronomy,
The Journal of the Society of Amateur Radio Astronomers, pp. 1-6.

---

Up and Running at 4 GHz:
The SETI-Capable BAMBI Radio Telescope
 By Bob Lash and Mike Fremont

Abstract

The design, construction, and initial observational results of a 
4 GHz amateur radio telescope are described in this first report from 
Project BAMBI (Bob And Mike's Big Investment). The system is now operating 
continuously. The planned extension of the BAMBI project to amateur SETI is 
also discussed.


Introduction

A number of efforts are underway in the Search for Extraterrestrial 
Intelligence (SETI). We have been deeply interested in the search for some 
time, and have concluded that amateurs can in fact construct affordable 
systems with sensitivities comparable to professional all-sky search 
strategies even with antennas of limited aperture. We have also concluded 
that we can achieve a reasonably respectable frequency coverage of a search 
spectrum as well. We hope this project will encourage other amateurs to join 
in the search. Project BAMBI is divided into two phases:

     Phase I: Standard Amateur Radio Astronomy:

We have initially operated BAMBI as a total power receiver for several 
crucial reasons. Most importantly, it has been a lot of fun. Many of you 
must have experienced this same indescribable, goose-bump raising feeling, 
when you realize you are actually detecting radio emissions from a source 
almost a billion light years away (e.g. Cygnus A). Also, we confirmed that 
the front end of the system actually worked, and calibrated our pointing 
accuracy. We are now looking for the lower limit of our sensitivity, 
which will allow us to better estimate our actual system temperature for 
purposes of Phase II.


     Phase II: Narrow Band SETI:

You will notice that BAMBI has a number of features that, while not required 
or even necessarily optimal for a total power receiver, are essential for 
SETI. The incorporation of a 25 degree K LNB front end, and an inexpensive 
spectrum analyzer in the IF pathway make the telescope "SETI-Capable".  
In Phase II, the spectrum analyzer will be used as a PC-controlled tuner, 
funneling a "block" of frequencies to BAMBI's analog-to-digital converter. 
An FFT will then be performed to break the block up into narrow channels on 
the order of 0.8 Hz each, and the results will be passed on to detection 
software to log possible "hits". The smaller aperture antenna actually 
enjoys a benefit for SETI use; the wider beam width means that sources 
will stay in the beam longer allowing a wider spectrum to be examined. 
When promising "hits" occur, one of the ways we can help distinguish them 
from terrestrial noise is by repeatedly scanning for the frequency drift 
expected due to the doppler shift caused by the earth's rotation and 
possibly that of the source.


Background: Amateur Radio Astronomy at 4 GHz


The radio sky at 4 GHz is an interesting place. While sources emitting by 
the synchrotron process (e.g. supernova remnants and radio galaxies) are 
less intense than at lower frequencies, thermal sources such as hot ionized 
gas clouds (H II regions where new stars are forming) and other thermal 
emissions from the plane of the galaxy are actually stronger. A big benefit 
to the amateur radio astronomer at 4 GHz is the narrower half-power beam 
width obtainable with even a modest sized antenna (1.9 degrees for our 8.5 
foot parabolic dish) than at lower observing frequencies, and the ready 
availability of inexpensive state-of-the-art low noise amplifiers, feeds, 
antennas, and positioners mass produced by the TVRO industry for satellite 
TV reception.


Design

An overview of the design is illustrated in figure 1. An off-the-shelf 8.5 
foot parabolic antenna with a built-in horizon-to-horizon positioner was 
pole mounted and oriented to sweep in a north-south arc covering a usable 
declination range of +90 to -42 degrees. The telescope's IBM-PC clone moves 
the antenna to the desired declination automatically using a standard 
positioner control unit. A 25 degree K HEMT GaAsFET LNB (3.7 to 4.2 GHz 
broadband input, 62 dB gain, complete with built-in down converter producing 
950 to 1450 MHz output) with attached Chaparral style circular feed was 
mounted at the focus. Cabling was run indoors and the signal was down 
converted to the 225 to 725 MHz range using a mixer driven by a 725 MHz 
local oscillator. The signal was then supplied to an inexpensive 0.1 to 
810 MHz spectrum analyzer plug-in card for IBM-PC's (built from a kit 
featured in the August 1991 issue of Radio Electronics Magazine and 
available from DKD Instruments). This card consists of a CATV type tuner and 
two NE615 receiver / IF Amplifier stages providing about 55 dB of gain in 
our application. The output of the 1st IF stage (now at a bandwidth of 280 
KHz centered at 455 KHz) was then coupled to two wide band op-amp gain 
stages (for 20X and 12X voltage gain) and then sent to a square law 
detector circuit. This was then buffered by a unity gain voltage follower, 
and sent to an integrator with a 70X gain and a 30 second integration time. 
Baseline adjustment is accomplished automatically by means of a computer 
controlled 8-bit D/A converter driving an offset adjust circuit (it 
algebraically subtracts the D/A output voltage from the signal). A DC 
amplifier stage delivering 50X gain provides the final signal to an 
8-bit A/D converter attached to an IBM-PC. Total power data is logged at 
30 second intervals. The log files are time averaged over a series of scan 
nights and the results are shown graphically on a PC.

Construction

If you have some familiarity with analog and digital electronics as well as 
programming, you should be able to build a system similar to BAMBI. We are 
both electronics and computer hobbyists (yes, we did later get degrees in 
electrical engineering and computer science, but that sort of academic 
background is not necessary to build a working system).  All of the front 
end electronics (from the antenna up to the IBM-PC spectrum analyzer card) 
are pre-made items that were patched together using F and N type jumper 
cables. Even the spectrum analyzer can be obtained pre-built if desired. 
The kit construction proceeded flawlessly, but we cracked one of the IF 

transformer cores with a metal screwdriver during the initial tuning and 
had to replace it (be sure to invest $1.25 for the recommended TOKO plastic 
adjusting tool!). The remaining electronics were assembled on a small 
breadboard, and the layout was not critical because the signal at this 
stage is very low frequency. The indoor electronics were placed in two 
rack mount bays (see photos) and attached to a dedicated IBM-PC. A second 
IBM-PC (on the right side of the photo) is used for time averaging analysis, 
display, and other development work, so that the telescope can continue 
observations without interruption.


Initial Observational Results


After checking out the system on a TVRO satellite (ANIK-E2) and the sun, 
we then realized we still needed some more gain, and added the two above-
mentioned IF amp stages that follow the spectrum analyzer. We then 
reobserved the sun (fig. 2), and as you can see the signal was so strong 
that it pegged off scale and illuminated the antenna sidelobes.

Fig. 2 - Solar Transit

After kicking up the post-detector amplitude to about 70 mV, we then successfully 
observed Cassiopeia A (figs. 3 & 4), Taurus A (fig. 5), Cygnus A (fig. 6), 
and Virgo A (fig. 7). The published strength of these sources at 4 GHz are 
listed below:

Fig 3. - Cassiopeia A, six separate scan night plotted

Fig. 4 - Cassiopeia A, time average of six scan nights

Fig. 5 - Taurus A (Crab Nebula), average of six scans

Fig. 6 - Cygnus A, average of seven scans

Fig. 7 - Virgo A, average of ten scans

The observations for Cass A in figure 3 show the individual plots obtained 
over 6 separate nights. Figure 4 shows the dramatic reduction in noise that 
occurs when the 6 nights are time averaged together. The remaining objects 
shown are also time averaged. In the Cygnus A plot (fig. 6), it is 
interesting to note that the Cygnus arm of the galaxy (the spiral arm we 
are in) is also clearly seen. The transients seen prior to and following 
the Virgo A detection (fig. 7) are due to baseline adjustment artifact. 
This source was close to the sun causing more frequent baseline compensation 
due to thermal changes. We are now beginning to search for weaker sources to 
find the practical limit of our sensitivity.


Planned Extension to Amateur SETI


As mentioned earlier, and quite remarkably, amateurs can obtain 
sensitivities comparable to NASA's all-sky search (which we were saddened 
to learn was recently cancelled due to budget constraints), even using small 
antennas, by means of applying the FFT to break the signal up into channels 
significantly narrower than what NASA was using (e.g. using 0.6 to 1 Hz vs. 
20 to 32 Hz channels). Man-made radio signals such as the strongest military 
RADAR systems radiate 14,000 megawatts into a bandwidth of 0.1 Hz. Even 
ordinary TV carriers concentrate about half their power in a very narrow 
band of 0.1 to 1 Hz (ref. 3). This narrow band detection approach provides 
a dramatic boost in the signal-to-noise ratio which significantly 
compensates for the smaller aperture dish. The next phase of project BAMBI 
will be to run a high resolution FFT spectral analysis on the output of the 
spectrum analyzer in a search for narrow band carriers not found from 
natural radio sources. This will initially be carried out in software, 
and later will be implemented using a dedicated FFT processor chip which 
was donated to the project. We will submit updates as the project 
continues.


Resources

You can download the BAMBI Software Collection right here.

Other resources:

DKD Instruments (810A Spectrum Analyzer IBM-PC Plug-In)
750 Amber Way
Nipomo Ca. 93444
(805) 929-2285
Fax (805) 929-5983

Orbitron (SX-8.5 TVRO Antenna) (Available from Skyvision)
351 S. Peterson St., Spring Green, WI  53588  
(608) 588-2923

Chaparral Communications (Polarotor Feedhorn) (Available from Skyvision)
2450 North First St., San Jose, CA 95131
(408) 435-1530

California Amplifier (25 Deg. K LNB) (Available from Skyvision)
460 Calle San Pablo, Camarillo, CA  93012
(805) 987-9000

Vigar Electronics (Surplus source of Cal. Microwave L.O.)
242-01 Alameda Ave., Douglaston, N.Y. 11363
(718) 423-1906

California Microwave (Very helpful with info on their PC7PL T.C.X.O. 
Local Oscillator)
(408) 732-4000

Mini-Circuits (ZFM-150 Mixer)
P.O. Box 350166, Brooklyn, NY, 11235-0003
(718) 934-4500


References

1. Sickels, R.M., Radio Astronomy Handbook, Catalog of 400 Radio Sources, 
pp. 296-299, 1989

2. Baars, J., et. al., The Absolute Spectrum of Cas A;  An Accurate Flux 
Density Scale and a Set of Secondary Calibrators, Astronomy and 
Astrophysics, 61:99-106, 1977

3. Papagiannis, M. (ed.), The Search for Extraterrestrial Life;  
Recent Developments, Proceedings IAU Symposium 112, pp. 263-270, 1984

Bob Lash's personal homepage is here.

See the BAMBI Observations of Comet Shoemaker/Levy-9's Impact on Jupiter!

BAMBI Home | What's New | Software | Jupiter Observations | SARA / Amateur Radio Astronomy