AMATEUR
RADIO ASTRONOMY
Western Australia
Hans Michlmayr (vk6zt)
Two 5-metre diameter stressed parabolic dishes along a 24-metre East-West baseline
are configured as a two element phase switched interferometer. Each antenna can be manually set in declination from - 90 deg. to + 50 deg. The Earth's rotation moves the antenna beams in Right Ascension (Drift Interferometer).
At 1410MHz (21.27cm wavelength) the beamwidth of each antenna is ~ 2.44 deg.
The East-West distance between the antennas is 24 metres (= 112.8 wavelengths).
This increases resolution ( 57.3deg./ 112.8) = 0.508deg in Right Ascension at 0 deg.Dec-
lination. Each interferometer fringe arrives with 2.032mins spacing = 8.2 milli Hertz
fringe frequency (fringe visibility). This of course slows down as we set the antennas
further away from +/- 0 deg. declination by a factor of 2.032/cos Declination.
This does not amount to much at 30deg. (2.346 mins), but increases sharply towards
the higher declinations, i.e. at 60 deg. we have 4.064 mins, and at 89 deg. 116.4 mins.
Provided the signal is strong enough to record clear fringes, the fringe frequency can
thus be used to confirm the declination of an observed radio source.
The choice of phase switching rather than full power interferometer was made because
it allows much greater back-end gain to be used, resulting in far greater sensitivity than
any full power set-up could ever afford. No attempt was made to temperature control or
compensate any of the RF stages, because phase switching inherently rejects slow
gain drifts, wide field sources ( =>0.5 deg.) also get suppressed. Ideally, one would like
to have variable antenna spacing, but this not easily realised....
Electronics Line - up :
The pick-up element at the focus point of each antenna consists of a three turn helical
antenna. This was found to be every bit as good as fully welded aluminium horn
elements. As the helical is much lighter and LNA mounts easier implemented than on
the horns, the helicals were retained.
The first stage LNAs (MGF 4319F PHEMT) connects directly to the helical feed point.
A single stage coaxial cavity 1/4 wave filter is placed between the first stage and the
second stage LNAs, mainly as a precaution against possible RFI from cell phones.
The second stage LNAs are combined PHEMT and GaAsFET broadband amplifiers.
The RF signal outputs from both antenna LNA systems are then routed to a central
control box. The two lines connect to a dual channel RF amp. and double balanced
mixers. Both mixers are fed from the same (onboard) oscillator, stable RF phasing is
essential in this application. After the mixers, the I.F signals go through two 150MHZ
LPF blocks, then another stage of amplification, then 70m underground back to the
house. DC power for the converter and LNAs is fed via the I.F. coax.cables.
The receiver is actually only a dual input 121MHz wide band I.F. amplifier.
Both inputs accept the I.F. signals from the remote converters; they pass first through
two independantly variable attenuators (a couple of double balanced mixers operated
in reverse with variable dc injection), to exactly match both signal levels.
Then the two I.F. signals go to two RF transformers where one of them gets continually
switched between zero and 180 deg. phasing compared to the other one. This way
the complications of having to switch at the 1400MHz front end are avoided.
.
In addition to normal phase switch operation, either antenna can be separately se -
lected, or switched in comparison mode switching (Dicke).
In the phase switch stage both I.F. channels are combined , filtered and passed
to the detector. Bandwidth is approx. 14 MHz, although other B/W were tried at times.
The "backend" is where the detected signal gets processed. First a stage of low
frequency amplification, then into the phase detector. Here the L.F. components are
switched by the same square wave drivers which switch the I.F. phase switch.
The recovered variable d.c. component is then amplified, differentiated (to remove
long term drift), integrated (to remove short term noise and fluctuations) and then
finally d.c. amplified for output to chart recorder or A/D converter, then to the P.C.
My original data processing package used to be done with a PICO TECHNOLOGY ADC-11 (10bit ,11 ch.) low cost device with PICO LOG software.
Since July 2003 most of the data collection is done with
Jim Sky's SKY PIPE. This employs a very low cost ADC
and inexpensive software. SkyPipe is especially designed
for Radio Astronomy .
1400 MHz phase switched interferometer
[ Previous 5 Sites | Previous | Next | Next 5 Sites | Random Site | List Sites ]
Radio Astronomy :
Last updated 7th Oct.2007
