A low-noise receiver covering the 40-52 GHz band has been on the GBT wishlist for several years. The scientific case for such a receiver has been presented by Al Wootten (GBT Memo 109) and others. In GBT Memo 141, E. Wollack and S. Srikanth proposed a specific configuration for the initial GBT receiver in Q-band. Because of budget and manpower limitations, progress on this receiver has been slow, but some progress has been made.
The mechanical layout of a four-beam receiver has been designed and the cooled microwave components for one beam have been assembled. The initial low-noise amplifiers were 5-stage units using an input InP HFET and four Rohm Research GaAs HFETS. Load tests of the cryogenics have shown that the system with one beam installed has adequate cooling margin with a 1020 refrigerator. Noise testing has been done and various materials evaluated for use in the vacuum window. A draft technical description of a beamsteering optics system to be used with the receiver is being developed.
Four-stage InP "MAP/VLA" low-noise amplifiers have been installed in the receiver for evaluation. It is expected that amplifiers of this type will go into the final system. Data from the initial noise tests with the new amps show an unexpected increase in system noise above 45 GHz. (A similar increase was observed in the March tests using different HFET amps.) The cause of the excess noise continues to be investigated. The receiver noise budget indicates the noise should be 28-32K throughout the receiver band.
Funding has been approved to complete a two-beam version of the receiver. Most long-lead components are on order, and work on the receiver packaging is beginning.
Feed pattern measurements have been made with the feed in free space, and then within the dewar. No significant differences were found with the feed operating in the dewar and looking through the vacuum window and Goretex weather cover.
The cryogenic RF assembly is shown sitting on the test dewar 15K station. The assembly consists of a profiled corrugated feedhorn, an Atlantic Microwave polarizer which separates the left and right polarizations, and two identical channels consisting of: noise cal couplers, an isolator, an HFET amplifier, an image-reject filter, another isolator, and a mixer. Also included are power splitters for the noise cal and mixer LO signals.
The noise cal and LO signals enter the dewar through stainless-steel WR22 waveguide for thermal isolation. The two 1-8 GHz IF signals exit through 0.085 stainless-steel coax.
A heatshield sits on the refrigerator 70K plate to reduce thermal loading to the 15K assembly.
The dewar outer cylinder and top plate is installed. The dewar layout allows for four of the cryogenic RF assemblies. The feedhorn positions lie on a 3.77 cm (1.485 inch) radius. The resulting beams on the sky will lie approximately on a circle with 40 arcsec radius. (The FWHM beamwidth of the GBT at 46 GHz will be approximately 16 arcseconds.) GTD simulations of the GBT indicate efficiency losses due to the feed offsets will be negligable, and predicts an aperture efficiency of approximately 68% (exclusive of surface efficiency). The feeds are held in position laterally by teflon collars bolted to the top plate and which contact the feed outer diameter at four points.
The vacuum window sits on the top plate. The 7.8 inch diameter window consists of 1 inch thick polystyrene foam (Eccofoam PS 1.04) and 5 mil thick Kapton. Hot/Cold noise measurements through a double window indicate the noise contribution of the window is less than 2K.
A view showing the assembled test dewar.
Acknowledgements: Engineers who have contributed to this receiver system include E. Wollack, S. Srikanth, W. Grammer, and R. Norrod. L. Beale is doing the assembly work on the dewar. The HFET amplifiers are designed and constructed by the NRAO Central Development Lab.
