3 Observations

We performed continuous six frequency 1-22 GHz spectra observations of compact extragalactic sources from 1st to 22nd December, 1997. We used the 600 meter ring radio telescope RATAN-600 (Korolkov & Parijskij 1979; Parijskij 1993) at the Russian Academy of Sciences' Special Astrophysical Observatory, located in Karachaevo-Cherkessia Republic (Russia) near Nizhny Arkhyz and Zelenchukskaya at the North Caucasus. The northern sector of the antenna (a part of the main ring reflector) was used together with the secondary mirror of cabin No. 1. Six broad band receivers are located at this movable cabin. The cabin moves along 150 meter long rails in order to be placed in the focus of the antenna system at different elevations. The main (meridional) transit method of observation was employed. Accuracy and reliability of these spectra measurements were essentially higher than earlier experiments conducted at the RATAN-600 due to the following improvements made to the receivers, antenna control system and the procedure of observations.

   

Table 2: Parameters of RATAN-600 broad band receivers in 1997, used in this work

$\lambda$,	nh	$\nu_0$,	$\Delta \nu$,	TphysLNA,	TLNA,	Tsys,	$\delta T_{sys}$,
cm		GHz	GHz	K	K	K	mK
1.4	2	21.65	2.5	15	23	77	3.5
2.7	2	11.2	1.4	15	18	70	3
3.9	2	7.70	1.0	15	14	62	3
7.7	1	3.90	0.6	15	8	37	2.5
13	1	2.30	0.4	310	35	95	8
31	1	0.96	0.12	310	21	105	15

   

Table 3: Measured and estimated beam widths and ratios r for the RATAN-600 northern sector with the secondary mirror No. 1 for different wavelengths $\lambda$ and elevations h

$\lambda$, cm	$\mathrm{HPBW}_{RA} \times \mathrm{HPBW}_{Dec}$
	$h=80^\circ$	$h=40^\circ$	$h=20^\circ$
	$r\approx5$	$r\approx10$	$r\approx20$
1.4	8 $.\!\!^{\prime\prime}$0 $\times$ 35 $^{\prime\prime}$	8 $.\!\!^{\prime\prime}$5 $\times$ 1 $.\mkern-4mu^\prime$4	13 $^{\prime\prime}$$\times$ 4 $.\mkern-4mu^\prime$3
2.7	16 $^{\prime\prime}$$\times$ 1 $.\mkern-4mu^\prime$3	17 $^{\prime\prime}$$\times$ 2 $.\mkern-4mu^\prime$8	24 $^{\prime\prime}$$\times$ 8 $.\mkern-4mu^\prime$0
3.9	23 $^{\prime\prime}$$\times$ 2 $.\mkern-4mu^\prime$0	26 $^{\prime\prime}$$\times$ 4 $.\mkern-4mu^\prime$3	39 $^{\prime\prime}$$\times$ 13$^\prime$
7.7	48 $^{\prime\prime}$$\times$ 4 $.\mkern-4mu^\prime$0	53 $^{\prime\prime}$$\times$ 8 $.\mkern-4mu^\prime$8	1 $.\mkern-4mu^\prime$4 $\times$ 27$^\prime$
13	1 $.\mkern-4mu^\prime$3 $\times$ 6 $.\mkern-4mu^\prime$5	1 $.\mkern-4mu^\prime$3 $\times$ 13$^\prime$	1 $.\mkern-4mu^\prime$8 $\times$ 37$^\prime$
31	3 $.\mkern-4mu^\prime$2 $\times$ 16$^\prime$	3 $.\mkern-4mu^\prime$3 $\times$ 33$^\prime$	5 $.\mkern-4mu^\prime$0 $\times$ 100$^\prime$

We used a new set of broad band receivers at the wavelengths of 1.4, 2.7, 3.9, 7.7, 13 and 31 cm with low noise HEMT amplifiers (LNA), cooled to a temperature of 15 K at the four shortest wavelengths (Berlin et al. 1997, 1993). Parameters of radiometers and antenna beams are given in Tables 2, 3.

Table 2 lists wavelengths $\lambda$; numbers of feedhorns n_h; exact central frequencies $\nu_0$; band widths $\Delta \nu$; physical temperatures of the LNA T^phys_LNA; noise temperatures of the LNA T_LNA; total noise temperatures of systems T_sys, including the antenna noise at middle elevations; rms noise temperature sensitivities of the system $\delta T_{sys}$ for a one second integration time. Dual-feedhorn receivers are beam-switched. Single-feedhorn receivers have a noise-added, gain-balanced mode of operation. Linearly polarized systems were available at all frequencies: horizontal at 7.7 cm and vertical at other wavelengths.

Table 3 gives half power beam widths (HPBW) in right ascensions HPBW_RA and declinations HPBW_Dec = rHPBW_RA, for various elevations. The values of HPBW_RA were obtained from our measurements. Ignoring the aberration effects, we estimated the factor r using theoretical simulations of the RATAN-600 beam by Esepkina et al. (1979) and their experimental testing by Temirova (1983). A map of the knife-like beam of the RATAN-600 northern sector is known to have different shapes of contours cross-sections at high and low power levels. In the absence of aberrations, the shapes can be described as ordinary elliptical contours (elongated on declinations) at half power level and higher levels. The contours are transformed to "the elongated eight" at lower levels or to "a dumb-bell" at 0.1 normalized power level (see Esepkina et al. 1979 for details).

The full permanent automatic control of 225 elements of the main ring reflector was achieved using the new control system of the antenna (Zhekanis 1997; Golubchin et al. 1995). Errors in position of each element of the main reflector, if present, were recorded in order to check the quality of the antenna surface for each observation. The positioning of cabin No. 1 with the secondary mirror was measured from one of the eight geodetic reference points, located every 20 meters along the rails. The accuracy of semi-automatic positioning of the cabin directed towards the focus was again checked by us some minutes before each observation. If an error of more than 2 mm with respect to a value given in the schedule was found, it was corrected.

Figure 1: Example of a full multifrequency scan for 4C 39.25, observed on 16 December 1997 with 0.1 second integration time

All horns of the radiometers are horizontally located and form a new configuration, which is an optimal one for decreasing transversal aberrations. Observations were carried out in the main meridian (transit mode). As a result, a response to an object is obtained due to its horizontal scanning by the antenna beam because of the daily rotation of the Earth (see an example of 4C 39.25 full scan on Fig. 1; the moment of culmination is at $2\hbox{$.\!\!^{\rm m}$ }5$ here). The total duration of each six frequency observations was usually about five minutes, and included also two sets of noise temperature calibration for 30-40 seconds before and after the passing of the source. The data acquisition system (Chernenkov & Tsibulev 1995) controls radiometers and records the output signals. After each observation, the main ring reflector and the cabin No. 1 with receivers and the secondary mirror were repositioned for observation of the next source on a new elevation.

We have optimized the observational schedule, using new software (Zhekanis & Zhekanis 1997). To increase the reliability of results by final averaging of the spectra, we endeavoured to include each source in the schedule two or more times during the set and each flux density calibrator in more than 70% of the days. The typical number of sources observed in a 24 hour observing session was about 80 in the optimized schedule. Several breaks in observations occurred because of weather conditions (snow-falls or unusually low temperatures $t<-15^\circ$C), nine hours of technical maintenance per week, etc. As a result, the total number of successful observations was about 1450 during 21 days, or about 69 observations per day on average and 2.6 spectra per source during the set (formal averaging). In 20% of the sources the spectra have been measured only once.