Introduction

Current status

Construction of A3

The essential changes to get to A3 from the B3i platform were a new (tunerless) mixer, and the replacement of the manually-adjusted quadrupler by a tunerless tripler. The mixers were built by the HIA/JCMT group using a design obtained from NRAO's Central Development Laboratory in Charlottesville, and the main challenge was to improve the noise performance at the high end of the band, above 255 GHz. Investigations were carried out at HIA comparing a six-element mixer design with one using four elements, and the latter was chosen as the one giving the better performance. To illustrate the construction of the mixer, images are shown here of the overall configuration (image size about 2mm) and a close-up of the junction and tuner (image about 0.6mm in total). In the interests of simplicity, and in spite of its age, B3i's microcomputer control was retained, and only minor changes to the VMS software were required.

Performance characteristics

• Receiver Temperature: In the laboratory before leaving Victoria A3 achieved a DSB receiver temperature of 70-80 K across the useable band. This represented a significant improvement over the performance of A2 above 255 GHz (see Trec plot with A2 and A3). The performance found on arrival at the mountain showed a degradation in noise temperature between about 245 and 255 GHz, as shown in the performance comparison. The cause of this is unknown, but may indicate a partial separation in the junction layers in the device. In general the elevated receiver temperatures do not affect commonly-used spectral lines. As a result further improvement to A3's mixer is being pursued in Victoria.

• Frequency agility: Tuning is fully remote; there are only two moving parts (Gunn backshort and tuner) which tune from one frequency to another normally within 10 seconds, and without having to visit the receiver cabin. This offers a significant advantage to programs requiring the observation of many lines. Already we have made use of automated frequency sweeps for calibration purposes. One of the major gains here is to be able to point while tuned to one frequency and then observe at any number of others without further pointing being required, since no visits to the cabin are normally necessary.

• Instantaneous bandwidth: The "native" bandwidth of A3 exceeds the maximum useable by the spectrometer; that is, up to 1.8 GHz. A 1.8-GHz filter is installed in the IF. The band is essentially flat across this frequency range, and thus offers an excellent extragalactic observing capability. A CO 2-1 spectrum of Arp 220 obtained during commissioning illustrates this point.

• Intermediate frequency: this was been increased to 4 GHz, bringing the new A-band receiver in line with all other JCMT heterodyne receivers. Since the receiver has no sideband filter it is necessary for observers to be aware of this difference between A3 and A2 (where the IF was 1.5 GHz). In A3 the sidebands are 8 GHz apart.

• Frequency-switching; "fast" frequency-switching (at 0.25 to 1 Hz) is offered with A3 and may provide better sky cancellation compared with normal "slow" frequency-switching (also available), which runs at a rate of typically 0.03 Hz. Fast frequency switching is also a more efficient observing mode than slow frequency switching.

• Historical performance; now that A3 has been around a while, it is possible to give a meaningful overview of the historical behavior of the receiver in terms of Trec - see this plot for an overview, and also see further discussion on the background information pages.

Pictures of A3

• A3 at HIA just prior to shipping
• Front view receiver assembly
• Beam path - detail
• Receiver assembly; view from right rear
• Electronics rack

• JCMT receiver cabin layout from above (membrane would be off the bottom of this picture). A3 receiver rack is in bay 9; A2 was in bay 8 (and is now at HIA, Victoria).

The JCMT group at HIA