Renovation of Seoul Radio Astronomy Observatory
and Its First Millimeter VLBI Observations

Since the trouble with the 4 K cryogenic system mentioned above was severe, we had to consider purchasing a new one. We finally decided to recycle a receiver that had once operated in the Combined Array for Research in Millimeter-wave Astronomy (CARMA). The main reasons for adopting the CARMA receiver are that it is paired with an economical 15 K cooling system, and its performance is still good. The only drawback may be that the CARMA receiver works in double sidebands (DSB) mode. It features dual-band operation in the 1 mm and the 3 mm bands, receiver temperatures of 60–70 K (DSB), and compactness (Hull & Plambeck, 2015).

The CARMA receiver is displayed in Figure 1. Millimeter waves from the antenna pass through two Teflon lenses without complicated beam-guiding optics and go into the dewar. The interior of the dewar is divided mainly into two parts—left for the 1 mm band and right for the 3 mm band, as shown in Figure 2. The part of the 1 mm band consists of a feed horn, a circular polarizer, an orthomode transducer (OMT), superconductor-insulator-superconductor (SIS) mixers, and low-noise amplifiers (LNA) (Hull & Plambeck, 2015). This configuration allows dual circular polarization observations. The 3 mm band receiving system is similar but does not have an OMT, allowing only the right-handed circular polarization (RCP). Table 1 summarizes the receiver parameters in the two bands. Local oscillator (LO) plates at both sides of the dewar generate the LO signals for the two bands, as shown in Figure 1. The mylar beam combiners reflect the LO signals in front of the dewar. Then the LO signals combined with waves from celestial sources propagate into the dewar.

After we took the receiver to the laboratory, we tested its components one by one and measured receiver temperatures representing its overall performance. Figure 3 displays the I-V curve of the 1 mm and the 3 mm mixers without LO power. One can see a steep current increase at bias voltages of around 10 mV. The measured receiver temperatures are similar to the ones in Hull & Plambeck (2015).

The cryogenic system consists of a cold head and a compressor. Within the dewar is a CTI1020 cryogenic cold head that consists of two cooling stages, where the second stage cools down to 15 K. It was modified to have an additional third stage by the University of California, Berkeley, which is shown in Figure 2. The cooling power of the third stage is only 50 mW, but enough to cool down mixers and feed horns to around 4 K. The 15 K cryogenic system is not so expensive but has gained the 4 K stage after modification, which is one of the reasons that we adopt the CARMA receiver.

The third stage can be further cooled down by lowering the pumping frequency. The cold head typically operates at 72 rpm, which seems the most efficient cycle frequency for the heat transfer of cryogenic regenerator (Ogawa et al., 1991). However, it is empirically found that if the pumping period is increased by a factor of two, then the temperature of the third stage further decreases below 3 K, probably because the cooling He gas spends more time inside the third stage (Plambeck et al., 1993). We usually set the pumping speed to 30 rpm using a frequency inverter, SV-is7, made by a local company, LS Electric Co. The typical temperatures during the regular operation are 47.4, 12.2, and 2.8 K, respectively, from the first to third stages. As for the compressor, we use the model M600 made by Trillion. Though a water-cooled compressor has a higher cooling capacity with a smaller volume, we adopt the air-cooled one since the air-cooled compressor requires fewer maintenance efforts.

The signal’s intermediate frequency (IF) bandwidth after the LNAs is rather wide but is narrowed down to 3.5–4.5 GHz through bandpass filters. The IF signal is then down-converted to 1–2 GHz in the IF processing box in the cabin. It goes to the observatory building and is converted to 0–1 GHz before being fed into the spectrometers and VLBI backends. The spectrometer covers 1 GHz bandwidth with $2^{14}$ channels resulting in a spectral resolution of 61 kHz. The signal flows for single-dish and VLBI operation modes are summarized in Figure 4.

Since the configuration changed near the secondary focus, we need to check whether the direction of the maximum gain of the feed horn points to the center of the subreflector. First, we point the telescope towards a mountain, which is seen at low elevations, i.e., the 300 K background. Then by moving an absorber immersed in liquid nitrogen at 80 K in front of the subreflector and by measuring the output power of the receiver, we can find the direction of the maximum response of the horn. Figure 5 indicates that the feed horn looks at the subreflector downward by $1^{\circ}$, while it is well aligned in the E-W direction. We corrected this misalignment by adjusting the heights of the receiver plate in four corners.

The change of the receiver position also affects the pointing offsets, which must be corrected. We established the pointing model by carrying out five-point observations for the standard stars in the Hipparcos catalog (Byun & Yun, 2002), using the optical telescope attached to the side of the antenna dish (Koo et al., 2003). The rms pointing errors are around $15^{\prime\prime}$ in both directions right after the model fitting. However, systematic offsets begin to appear, probably due to the differential solar heating of the antenna structure, which may affect the 1 mm band observation because of its smaller beam size.

In February 2019, SRAO detected the first light of the CO $J=2-1$ line at 230.538 GHz toward Orion KL with the CARMA receiver (Figure 7). The best system temperature for a 1 mm band is measured as 400 K (DSB). It is found that, contrary to expectations, the system temperature rises in the winter. The reduction of cooling capacity due to the hardening of oil in the compressor may cause this problem. We expect lower system temperatures by keeping the compressor warm in the winter.

The 3 mm band receiver was operated in the spring for recent two years. The lowest system temperature in the 3 mm band is 150 K (DSB) at 86 GHz. We have made single-dish observations of several bright spectral line sources, such as Orion KL and TX-Cam, to inspect the pointing accuracy and verify the overall system performance before the VLBI observation.