We simulated imaging for the 23-antenna CARMA telescope in three compact configurations using a model image of Saturn. Simulated uv data sampled by the heterogeneous array of 10.4, 6.1, and 3.5 m antennas were used to make mosaic images. Three different Maximum Entropy deconvolutions were used. The best image fidelity was obtained using a joint deconvolution of the interferometer and single dish data. Using the single dish data as a default image, provides a total flux estimate and low spatial frequencies unsampled by the interferometric mosaic. Both methods give higher image fidelity than just using the interferometer data with a total flux estimate. We analyze two problems: i) sampling the short spatial frequencies, ii) Nyquist sampling the large scale structure, and show that the image fidelity is improved by a higher uv sampling density. Using the interferometer spacings between 3.5 and 6.1 m sampled by the 3.5 m antennas further improves the image fidelity because the interferometer observations provide better quality data for these short spatial frequencies, than can be obtained from the joint deconvolution with the 10.4 m single dish data.

This memo covers the basics of BIMA polarimetry observing, including writing polarimetry proposals, templates of observing scripts (to include both continuum and line polarimetry together), and then data reduction using MIRIAD. Some knowledge of interferometry and MIRIAD is assumed.

This memo discusses temporal changes of up to 3mm in the positions of BIMA antennas observed at HCRO. The motions occur on timescales of days to months, and appear to be related to seasonal temperature changes. Position changes observed over a three month period in the spring of 2001 are shown for illustration, and possible sources of the instability are discussed. Ground stability at the 0.1mm level will be required at Cedar Flat in order to avoid regular monitoring and correction of initial pointing and baseline solutions for CARMA.

A new set of Java wrappers for the MIRIAD input-output routines provides access to MIRIAD data from Java and Jython (Jython is a dialect of Python). Originally developed as part of the correlator for the Allen Telescope Array, these wrappers are general purpose and are available for application in other software projects where Java or Python is used. In this article we describe the architecture of these wrappers and demonstrate their use with some examples.

We review the expected sensitivity and spatial dynamic range of the CARMA array during the next 10 years in comparison with the ALMA, ACA, SMA and IRAM arrays. The scientific capabilities of the CARMA array can be greatly enhanced at modest cost by adding a full polarization spectral line correlator.

This document presents the results of an evaluation of the GILDAS software Package for the purposes of ascertaining its compliance with the ALMA Offline Data Processing Requirements given in ALMA-SW Memo 18. This provides a useful index and description of the data processing currently used for millimeter wavelength arrays.

This document presents the results of an evaluation of the MIRIAD software Package for the purposes of ascertaining its compliance with the ALMA Offline Data Processing Requirements given in ALMA-SW Memo 18. This provides a useful index and description of the data processing currently used for millimeter wavelength arrays.

This memo briefly describes the existing implementation of the BIMA telemetry and discusses in detail a conversion to CANbus, necessary for the merging of BIMA into CARMA. This implementation (by no means final) uses a VME board equiped with a microcontroller Phillips XAC3 to replace the current telemetry receiver, keeping the rest of the hardware intact. I also discuss the possible structure of the embedded software, as well as the limitations impossed by the hardware and the data rates on the field bus.

In this memo, we characterize BIMA proposals for the period 2000-2003 into 8 categories: Science, Antenna Configuration, Help Requested, Special Needs, Observation Type, Polarization, Receiver Band, and Mosaicing. We present the data as tables and pie charts for each category. Most, ~90%, of the proposals requested (and observed) spectral line data. Most of the proposals requested antenna configurations, receiver bands, and observing modes in the regime where the BIMA telescope works well without and special requirements for instrument or weather.

In this memo, we review the imaging requirements, calibration procedures, and processing options for the Allen Telescope Array. The array has excellent uv coverage and a very high data rate which will rapidly saturate conventional human data reduction and imaging. The array should produce final, calibrated images as its normal output. Imaging with the ATA was simulated for a range of sources. The performance of MIRIAD running on modern PC's appears adequate for the ATA, but not by a large factor. Efficient software is required.

In this memo, we compare the imaging performance of three compact configurations for the CARMA telescope. We discuss ring and spiral configurations. MIRIAD scripts simulate imaging with heterogeneous arrays, and pipeline the results into tables. We present tables of beam FWHM, brightness sensitivity and sidelobe levels for source declinations from +60 to --40 degrees. Mosaic images for source diameters $32''$, and $40''$ are analyzed. The residual imaging errors are characterized by the recovered flux, peak flux density and the sidelobe levels. We present tables of the sidelobe levels and image fidelity. The most compact configurations provide the best brightness sensitivity, but suffer from antenna shadowing. Spiral configurations are more versatile and give better imaging fidelity for the source models considered.

This memo summarizes the antenna position calibration used at Hat Creek, and discusses its application to the CARMA and ALMA telescopes. Observations of quasars over wide range of HA and DEC provide data from which the antenna positions are determined. If the antenna positions are in error by many wavelengths, 2pi phase ambiguities may make this process difficult, especially when the atmospheric phase coherence is poor on long baselines. In this case the antenna positions can be first determined from the phase difference between the sidebands in a double sideband system, or across a wideband single sideband. In addition to the antenna positions, there may be axis offsets on the antennas which can be fitted from the quasar data or measured mechanically. These offsets should be stable and not change when the antennas are moved.

We summarize measurements of linear polarization leakage terms made with the BIMA array at 1.3 mm over a 2 year period beginning in 2000. The individual antenna leakage terms show a median rms of 1.5% over this period. If these errors are stochastic, then the error in measured polarization is 0.5%. There is no apparent change in the leakage terms with time. The occurrence of the occasional bad calibration solution argues for the use of a mean set of leakage terms rather than frequent recalibration. There is no apparent reason to calibrate more frequently than once per array. The polarization fraction and position angle measured for 3C 279 agrees with that found at 86 GHz with BIMA and with that found at 22 and 43 GHz with the VLA.

This memo presents the first steps toward a functional optical pointing system at BIMA. The specifications of the STV video camera mounted on the optical pointing telescopes of the Hat Creek interferometer are shown. The sensitivity of the system has been empirically determined (zeropoint = 15.9 mag) and simulations have shown that a signal to noise of approximately 15 is needed to determine a centroid accurate to a 1" with 95% confidence. This means that a pointing solution can be obtained in a random field (stellar density averaged over the celestial sphere) with an integration time of a few seconds.

This note describes tests made in March 2000 of a planar L-band orthomode transducer. The transducer consists of 4 probes symmetrically arranged in a circular waveguide and two wideband baluns, made from back-to-back slotline-to-microstrip transitions. We designed this device to model the performance of a millimeter wave transducer where probes and baluns are microfabricated with SiON thin-film dielectric deposited on a quartz substrate. In the L-band model, probes and baluns are printed on 0.010 3thick Duroid 6006, mounted on a 0.25 3 thick piece fiberglass and placed flush against the waveguide. Each pair of opposite probes is coupled through a balun to a 50 ohm microstrip output. Measurements show that, from 1.2 1.9 GHz, the orthomode transducer has an output return loss less than 18 dB and crosspolarization less than 35 dB. It should be possible to scale this design to cover the 80-115 GHz or 210-270 GHz bands.

Here are some very preliminary thoughts on the design of compact and intermediate configurations for CARMA. I propose that we should design 15- and 23-element configurations for D array, and 15-element configurations for the larger arrays. The most compact configuration can reasonably expect to be $\ga$50m in diameter, in accord with Lee's previous assumptions. I show a sample, non-optimized 23-element configuration that utilizes all minimum separations available among the six different pairs of primary elements. The uniformly weighted beams for this configuration are $\sim$11\arcsec. Given the 0.3--0.4\arcsec\ beams for A array, this suggests that we should aim for intermediate configurations that yield uniformly weighted beams of $\sim$3.5\arcsec\ and $\sim$1.1\arcsec\ if we design four main configurations. I suggest that we do design four (rather than five) configurations, and try to share 3--5 pads among adjacent arrays where possible. Using resources that we save by sharing pads (and by eliminating a configuration), I also suggest that we design hybrid DC, CB, and BA configurations that utilize e.g. 10-12 fixed stations from each of the D, C, B, and A arrays, for observing low-declination sources. F. Boone is currently modifying his configuration optimization code to allow for heterogeneous arrays; this may be a very useful tool for CARMA array design.

The WASP and APHID spectrometers contain Intel 80C251SB microcontrollers to manage the spectrometer's real-time and other hardware tasks. The microcontroller accumulates data from the spectrometer hardware, sends signals to the telescope's chopping secondary and beamswitch (``nodding'') inputs, and sets the spectrometer's instrument state. It communicates with a host computer or terminal using simple commands and a serial interface. This memorandum describes the microcontroller commands.

The variation of the overlap of the astronomical and water-vapor radiometer beams with height in the atmosphere is investigated for the current BIMA setup, and is compared with similar calculations for OVRO and a proposed system for ALMA. The findings are that at BIMA, the 3-mm and 1.3-cm beams essentially do not overlap above a height of ~1 km. The linear separation of the beams is ~35 m at a zenith height of 4 km. The dominant factor in determining the beam overlap is the angular offset between the beams, with the overlap decreasing very rapidly with increasing offset. Under conditions where the dominant source of water-vapor fluctuations has a scale greater than the beam separation, the small beam overlap will generally not pose a serious problem to correcting the phase.

However, we do find that the current 28-arcmin beam offset at BIMA potentially places a limit on the measurement of path length differences achievable with the APHID water vapor radiometer, although this is dependent on the height of the dominant source of fluctuations in the water vapor content. For a number of reasons, the beam separation will not be a problem for an instrument which has recently been proposed for ALMA. Finally, we also suggest that, if possible, future instruments should have beams which are offset in azimuth to minimize the error introduced by looking through atmospheric lines of sight of unequal length.

We introduce the MirStone, a dimensionless measure of how well a standard set of MIRIAD programs perform on a given machine. It is simply computed from the elapsed time a benchmark script takes on an otherwise idle machine. It is normalized to be roughly unity on a typical late 20th century desktop (Ultra-300 or Pentium-400). A MirStone is defined as 5 divided by the elapsed time (in minutes) the benchmark takes to run. This paper introduces the benchmark, and compares Solaris and Linux configurations in the late 20th Century. This paper is expected to evolve as better hardware becomes available.

Fluctuating amounts of water vapor over an antenna in an interferometer add phase shifts to the instrument, shifting the fringe pattern on the sky to decrease signal correlation. Optically thin line radiation from the same water vapor that causes the decorrelation can be accurately measured, however, allowing an estimate of the column of water above individual antennas. Knowledge of the water column allows an equivalent but opposite phase to be inserted in the data processing; this is the radio equivalent of adaptive optics. This memo explores the instrumental requirements for a multi-channel radiometer capable of measuring pathlength differences to 35 µm, a goal for interferometry at 1 mm wavelength. An examination of linear and nonlinear error sources shows that residual amplifier or detector nonlinearity is likely to be the most significant instrumental limit for atmospheric phase correction.

When 1mm receivers were first installed on the BIMA antennas in October 1997, the aperture efficiencies were found to be unexpectedly poor. Lab measurements of the 1mm receiver optics finally identified the problem - two of the optical components in each dewar, a lens at the feedhorn aperture and a flat 50 K window used as an infrared filter, had been machined from defective rods of extruded Teflon (PTFE). These rods were slightly porous along their central axes, leading to a radial gradient in refractive index which corrupted the receiver beam pattern. When the defective components were replaced with lenses and windows fabricated from Teflon plate, the measured beam patterns closely matched those predicted by Gaussian beam analysis, and the aperture efficiencies improved dramatically. The new lenses were antireflection-coated by drilling a grid of holes into each surface, rather than by cutting circular grooves into the surface. In this way the refractive index of the antireflection layer is insensitive to polarization. This memo summarizes the optical layout of the BIMA antennas and receivers, gives a brief history of the 1mm aperture efficiency problem, and describes lab tests and theoretical modelling of the 1mm optical components. Section 7 discusses the antireflection coating on the new lenses.

We discuss the calibration and use of water vapor radiometers used for atmospheric phase correction on the BIMA array. Over intervals of a few minutes the WVR corrected data are consistent with the thermal noise of 100 microns. On longer timescales the calibration is currently limited by variations in the WVR gain and the atmospheric scale factor.

The requirements for greater noise immunity from outside sources, such as satellites, in radio astronomy require systems with greater dynamic range so that the interfering signal doesn't wash out the very quiet signals from deep space. One of the classic methods for bringing a high frequency RF signal down to a lower more manageable IF frequency range is the quadrature down-converter. While quadrature down-converters are relatively simple, they suffer from sideband separation problems. Loud signals in one sideband can leak into the other and the amount of this leakage depends upon how much time and money is spent upon the circuit. Typically one can expect sideband rejection levels of only 30 dB below the interference even with components of reasonably high precision. The errors in a quadrature down-converter are easily identified and their effects on the power spectrum can be eliminated, digitally, at the low frequency IF, resulting in a circuit with superior side-band separation. These errors can be identified using the signal being measured so that no separate time consuming calibration is required. As the astronomic measurement is taking place an error function can be determined and used to servo the very slowly changing corrections required as the circuit ages or changes with temperature. As the cost of digital electronics drops it becomes economic to follow a relatively inexpensive quadrature down-converter with digital electronics designed to compensate for it's short- comings.

Recently Eric Keto of the Smithsonian Astrophysical Observatory ( Technical Memo Number 134 ) suggested using four level, shifted m-sequences as an alternative to two level Walsh functions used for phase switching in interferometer arrays. Phase switching is used to reduce DC level shifts, to separate the sidebands, and it is used to provide cross-talk and interference rejection. In order for cross-talk and interference rejection to work, it is necessary for all of the phase demodulation functions for each of the baselines to be orthogonal. It is also necessary for all of the antenna phase modulators to be orthogonal. As bigger arrays are designed and the need continues for shorter integrations, there is a need for more efficient switching schemes. The requirement for baseline and antenna orthogonality places severe constraints on the selection of switching functions. The crucial question that must be answered in judging any modulation scheme is how efficient it is in supporting the largest antenna array in the fewest steps. We examine m-sequences, normal two level Walsh functions, and two kinds of complex Walsh functions. It appears that, for large systems, even an ideal solution may produce a set of phase modulation functions that are too long for short integrations.

This memo describes measurements made of a scale-model ortho-mode transducer to evaluate its suitability for use at millimeter wavelengths. The device consists of four probes placed in a single cross-sectional plane of a cylindrical waveguide. Each pair of probes is driven in anti-phase with a 180 degree hybrid. A good 50 ohm match with a return loss of 20 dB and polarization isolation of 40 dB was achieved with a fixed backshort in a 40% bandwidth. An implementation in the 3-mm band is suggested.

This memo discusses the imaging properties of the combined BIMA \& OVRO arrays. The 15-antenna array with 8 GHz bandwidth and a system temperature 100 K will have a sensitivity 1 mJy min$^{-0.5}$. For extended sources the image fidelity depends on the source brightness distribution, and data sampling. For mosaic observations the heterogeneous array comprised of 10.4 and 6.1 m antennas will have three different primary beam types. We compare the image fidelity obtained as a function of the $uv$ sampling, the pointing sampling, and the image complexity, and compare these results with imaging by a homogeneous array of 8 m antennas.

Oliver Lay, May 13, 1999
available as: HTML

Water vapor in the Earth's troposphere introduces extra electrical path in the propagation of radio signals through the atmosphere. The distribution of water vapor is irregular and distorts the wavefronts of incoming radio waves, limiting the angular resolution that can be achieved with ground-based telescopes. The level of fluctuations depends both on the location of the site and on the prevailing atmospheric conditions. The ability to measure the fluctuations is therefore important when choosing a site for a new instrument, and for scheduling observations of existing telescopes.

Existing phase monitors are radio interferometers that monitor monochromatic beacon tones from geostationary communications satellites at a frequency of ~12 GHz. They have a classical heterodyne design based on two satellite receiving antennas; each has a frontend for amplifying and downconverting the incoming signals using a local oscillator that is phase-locked to a common reference frequency. In addition to multiple phase-locked loops these instruments require expensive phase-stable cabling to reduce the effects of thermal drift.

The new system uses two consumer 18" digital satellite TV dishes to monitor satellite TV broadcast signals over a bandwidth of 500 MHz (12.2 to 12.7 GHz). The novel design eliminates the need for phase-locked loops and thermally stable components, and uses a pair of Gilbert Cell multipliers to perform the broadband correlation. A phase monitor has been been built and deployed at the site of the Berkeley-Illinois-Maryland Association Millimeter Array in Northern California, and has been operating successfully since June 1998, measuring the difference in electrical path length for parallel lines of sight to the satellite separated by a baseline of 100 m. With a hardware cost of approximately $4000, it is much cheaper than previous instruments, and the low power requirements and high reliability make the system suitable for site testing in remote locations.

Plots showing current conditions at the site and a detailed description of the instrument can be found at http://bima2.hcro.berkeley.edu/phasemon/phasemon.html.

A big problem in correlator design is the large number of signal wires that must be distributed to the correlators. A 10 antenna system has 45 baselines. Each antenna must supply signal to 9 correlators for a total of 90 signal paths. If each antenna has a single digital sampler that samples to two bits then there are 180 signal pairs that must be carefully designed to arrive at the proper place at exactly the correct time. If an extra antenna is added there are 220 signal pairs to deal with and the whole signal distribution system must be redesigned. By a simple strategy it is possible to limit the signal paths from each antenna to two. It is also possible to expand the system indefinitely without the need to redesign the distribution of signals.

Fast switching cycle times to minimize atmospheric phase fluctuations are calculated for conditions applicable to BIMA A array observing. The optimal cycle time is defined as having the lowest total noise from thermal and atmospheric contributions. It is assumed that decorrelation from the atmosphere can be partially corrected using observations of a test calibrator. Guidelines are given for cycle times for various array, source and atmospheric parameters.

This memo presents suggested new antenna configurations for the 10-antenna array at Hat Creek. We use 4 antenna configurations with angular resolutions 18'', 6'', 2'', and 0.4'' at 100 Ghz. We have re-designed the 6'' and 2'' arrays to have more uniform uv-coverage and lower sidelobes. The new arrays have better instantaneous uv-coverage. Two antennas do not need to be moved when re-configuring the array.

This memo analyzes data collected at Hat Creek on the atmospheric phase conditions during the 1997-98 observing season. The data come from observations of quasars during normal observing and from the dedicated phase monitor. Percentiles of rms phase are calculated as a function of time of year and time of day. The length of time over which the phase remains below a given threshold is measured and discussed with regards to dynamic scheduling. The effect of phase noise on observations in different arrays is briefly discussed.

We present results from atmospheric model calculations for the design of an atmospheric phase correlation radiometer for the Berkeley-Illinois-Maryland Millimeter Array (BIMA). The radiometer will monitor the atmospheric path delay by observing the fluctuations in the emission from tropospheric water vapor which causes de-correlation of astronomical signals which are observed along different lines of sight. We discuss the applicability of monitoring the optically thick 183 GHz water line and the optically thin 22 GHz water line. We conclude that for the BIMA site, which is not extremely dry, optical depth effects make observations of the 183 GHz line unfavorable. We discuss possible observing schemes and conclude that a multi-channel radiometer provides the highest achievable accuracy, as it provides the possibility to fit a line shape model to the observed channels. Systematic errors due to the unknown altitude distribution of water vapor and optical depth effects can be significantly reduced by this scheme, compared to methods which only monitor the change of the peak brightness temperature of the water vapor line.

This memo briefly compares radial, circular and irregular antenna configurations for aperture synthesis arrays. The best array configuration depends on the source structure. Combining arrays with different antenna sizes is discussed.

We built fixed-tuned SIS mixers for use between 70 and 115 GHz on the Berkeley-Illinois-Maryland Association (BIMA) array. The mixers are similar to 215 GHz SIS devices designed by R. Blundell et al. Our 1.4 X 1.4 micron Nb/Al-Al_2O_3/Nb junctions have extremely sharp I-V characteristics, with R_{sg}/R_{n} ~ 40, and \omega RC ~ 2. Heterodyne tests were made with a temperature-regulated vane mounted in the input waveguide to the mixer block. With vane temperatures of 20 and 35 K, we measure double sideband receiver temperatures T_rcvr of 17 K from 80 to 115 GHz. Some gain compression occurs for input load temperatures greater than 50 K. Hence, with vane temperatures of 85 K and 280 K we measure T_rcvr ~ 35 K. Similar noise temperatures are measured with 77 and 295 K loads mounted outside the dewar. Because the mixers are partially saturated, we bias the junctions approximately 0.1 mV above or below the gain maximum, where the receiver response is a nearly linear function of the input signal level. This yields T_rcvr of 35-40 K and makes it possible to calibrate the receivers on the telescopes using the standard chopper wheel method.

With no LO power applied to the mixer, we observe a small IF signal which is proportional to the square of the input load temperature; we believe this is due to self-mixing of blackbody photons in the junction.

A study has been made of the BIMA Cassegrain optics to assess alignment tolerances, depth of focus, and aberrations, and the present degree of correction of the system. This has been done both by ray-tracing the design specification for the Cassegrain system, and also by investigating holograms made by correlating one antenna with another. The surface accuracy was assessed by RMS and efficiency computations. Residual aberrations are expressed as coefficients of Zernike polynomials, and these can be used to characterize the system.

The goal of the wide-band digital proposal is to build a system using readily available, off the shelf parts. There are four elements to the wide-band digital correlator proposal. First, a sampler of unique design is proposed to digitize an 8 GHz bandwidth signal. Second, an 80 channel digital correlator is proposed. Third, a digital delay line is proposed to replace the existing analog delay. And finally, a finite impulse response filter is proposed to replace the existing single-sideband mixer and third LO.

I present results of a study of the BIMA archiving system in an effort to measure the typical rates at which data is transferred from Hat Creek to NCSA for archiving. This was done by altering the archiving software at Hat Creek to measure data transfer times as the data is sent to the archive server. The average archiving rate over a two week period was found to be about 350 kilobits per second with large variations between 200 and 700 kilobits per second. Internet traffic was found to be the primary limiting factor to the archiving rate over software overhead and competition for local system resources. The results are used to make recommendations for limits to be placed on the data production rate to ensure that the archiving system can keep up.

We describe the design model and implementation of the data archiving system used by the Berkeley--Illinois--Maryland Association (BIMA) Millimeter Array Telescope. This system transmits data in real-time from the BIMA telescope at the Hat Creek Observatory in northern California via the Internet to an archive server at the National Center for Supercomputing Applications (NCSA) in Urbana, Illinois. Once at NCSA, the data undergo minor processing to make it available to astronomers via an HTML interface: (1) the data are checked for successful transmission, (2) metadata are extracted from the data and inserted into a searchable database, and (3) the data are sent to the NCSA Mass Storage System. When necessary, the system can carry out rollback operations which allow it to easily recover from errors, particularly those associated with the often unstable Internet. We also comment on some ways in which the system can be improved and expanded to adapt to changing observing strategies.

We plan to use the atmospheric path RMS as a parameter for a dynamic scheduling program at Hat Creek. The RMS path can be measured by a satellite phase monitor or we can use a short observation of a quasar to measure the atmospheric phase RMS every few hours. A power law fit to the quasar observation is used to estimate the RMS path on the longest baseline. The scheduling program can be script driven, providing a convenient way to accommodate the specific observing environment at Hat Creek. We have extended the LST time format used in observing scripts at Hat Creek to allow dynamic scheduling over multiple days.

This memo briefly reviews some considerations for imaging with aperture synthesis arrays containing mixed antenna sizes. At millimeter wavelengths, where mosaicing observations are important, the effect of the mixed primary beams is significant. For mosaicing observations, the sensitivity depends on the number and diameter of antennas with each primary beam type. The array configuration can be optimized to use the different antenna sizes, and depends quite strongly on the source structure. Data sampling rates are set by both the largest and smallest antenna diameter, and are larger than for an equivalent homogeneous array. The negative primary beam from mixed antenna baselines may be a problem in the mosaicing algorithms. Conceivably it might also serve to knit together the mosaic by providing different weightings of the overall image. The dynamic range of mosaiced observations is often limited by pointing errors. Pointing is more of a problem for larger antennas, but the mixed primary beam patterns may make it easier to implement a pointing self-calibration algorithm. Mosaicing with mixed antenna baselines needs detailed study.

We argue that a direct measurement of the atmospheric phase rms provides the best weather parameter for scheduling interferometers for high frequency observations. We evaluate the performance of a dynamic scheduling program under varying atmospheric phase conditions, and compare the simulated schedule with the results from actual observations.

We report on the current state of autocorrelation capabilities at BIMA. Spectral line observations with window bandwidths <= 50 MHz are reliable, but observations with 100 MHz windows and continuum measurements are not currently feasible. The relative calibration among the individual antennas is typically good to about 20%, and often the agreement among all but one or two of the antennas is better than 10%. The line shapes and strengths of BIMA observations of CO and HCN in IRC+10216 agree to within the noise with spectra from the Bell Labs 7-m and the NRAO 12-m telescopes; a 500" x 350" BIMA map of C18-O from NGC 2024 agrees well in structure with an NRAO 12-m map from the same region, though the emission measured at BIMA is about 20% weaker. Sensitive observations of a weak CS line in MBM12 show that there is no noise floor at the level of at least 16 mK per 24 kHz channel width. The measured rms noise in all the antennas is consistently about 8-15% lower than the theoretical noise; we are unsure of the explanation for this discrepancy. We describe new MIRIAD routines to process autocorrelation spectra and give template observation and data reduction scripts.

We describe the use of an automated clump-finding routine for analyzing spectral line data cubes that has been recently installed in the MIRIAD package. There are three programs: the main analysis program, clfind, that reads in the data cube and outputs a cube of the same dimensions with the clump assignments; a statistics program, clstats, that calculates clump positions, sizes, linewidths and masses; and a clump plotting program, clplot, based on velplot, that allows the user to isolate and produce maps of a user input list of clumps.

VLBI observations at short millimeter wavelengths are required to resolve the compact radio components associated with active galactic nuclei. These components are optically thick at longer wavelengths, or obscured by scattering, as in the case of SgrA*. Large new telescopes planned for completion in the next 10 years provide a powerful new platform for millimeter and submillimeter VLBI. The atmospheric phase correction required for successful operation of millimeter arrays on baselines of several km also enable phase coherent VLBI observation with $20 \mu {\rm arcsec}$ resolution. This is sufficient to resolve structure on the scale of an accretion disk around a massive black hole in the nearest quasars. The new telescopes can be outfitted for VLBI at modest cost and should be planned with this capability in mind.

The idea of using a minimum redundancy linar array for radio interferometers is familiar to radio astronomers. For East-West linear arrays the minimum redundancy array provides essentially an exact optimum solution for the choice of antenna separations for a 12 hour earth rotation sysnthesis. For two dimenional arrays, there is no corresponding exact solution. The authors discuss the application of the minimum redundancy concept to the 2-d "T-configuration" array.

In order to make the best use of varying weather conditions we propose to to implement dynamic scheduling for BIMA array observations. The schedule is determined in real time from a prioritized list of observations and the system status including the weather conditions. We briefly discuss the relative merits and problems of flexible scheduling.

This memo describes the current approach for using the Hat Creek array as an element in a VLBI array. Theoretical and practical questions are addressed. The principal elements of a VLBI station are the antenna, a phase stable receiver and IF system, a stable frequency standard, usually a hydrogen maser, and a backend for recording the IF (Figure 1). We consider each of these in Sections 2 through 5. We also discuss procedures for the setup and execution of a VLBI experiment at Hat Creek in Sections 6 through 8. Finally, we give a wish list for improvements and technical innovations in Section 9.

We propose 10-antenna arrays for the coming year, giving some discussion about the strategy of their design and possible alternatives. For the A and B arrays, we propose those that were worked out and described in the BIMA Array paper (1996, PASP, 108, 93). For the C array, we have a new design that better emphasizes sampling of short spacings in the uv-plane. We have also worked out a new H array for low declinations that has better uv-sampling of short spacings.
We discuss alternative strategies for observing low-declination sources at moderate resolution, and we compare the H array with a combination of C and B arrays. We find that sources at low declinations are as well or better observed with C and B arrays than in H array. Given the scheduling inefficiency of the H array, we propose that it be eliminated.

We briefly discuss the relative merits and problems of self-calibration and fast switching calibration for the BIMA a-array. Self-calibration is very successful for strong sources, but can lead to spurious results for small arrays of antennas if the source is weak or too complex. Fast switching between a source and a nearby calibrator can reduce the rms phase on baselines longer than few hundred meters. Switching times of a few seconds are desirable, but are inefficient.

The sampler used in the BIMA correlator quantizes the signal into four levels. This coarse quantization causes the correlation function to be distorted. The distortion may be removed by first normalizing the correlation function and then scaling each measured correlation coefficient by a correction function. The proper function to use and the normalization depends upon the input power level relative to the sampler clipping level. The normalization to use is the topic of this memo.

This memo defines the algebra and outlines the steps used for calibrating polarization switched data, and making polarization images for Stokes I, Q, U and V.

Presently the length of observing tracks at BIMA are based on several different criteria: user requests, arbitrary cutoff elevations, or standard track lengths. All of these methods can lead to the continuation of tracks for only minimal gains in signal-to-noise ratio (SNR) or UV coverage. In addition, some tracks are cut short when they are still productively adding to their UV coverage and SNR. To obtain the maximum productivity for the array it is important to determine the optimum track length and the optimum array configuration for each declination. The memo proposes a method of determining the optimum track length by looking at the SNR of a track compared to a horizon to horizon track weighting by the system temperature. We derive a table of minimum and maximum track lengths for different source declinations.

The BIMA 9-antenna array provides resolutions 5'', 2'', and 0.5'' in three standard array configurations at 100 GHz. Source structures more extended than a few times the synthesised beam width are attenuated and distorted in the synthesised images; structures larger than about 10 times the beam are not detected. We propose to add a more compact D array configuration with a resolution 17'' and better sensitivity to larger structures. Mosaicing observations using this D array configuration are discussed.

We have analysed data obtained with the Hat Creek 9-antenna array with antenna spacings between 80 and 1045m. In clear weather we find a correlation between the atmospheric phase fluctuations and the total power measurements. This memo provides a recipe for correcting the phase, and an analysis of some recent data.

The BIMA array receivers each have a single linear polarization. Polarization Switching is required in order to obtain polarization observations. This memo explores efficient switching cycles.

The Berkeley-Illinois-Maryland Association millimeter wave interferometer has a very clean and simple optical design. Each antenna holds one dewar which houses mixers for 4 frequency bands; the feed horns for the 4 frequency bands are located slightly off the telescope axis and look directly at the subreflector without any intervening transfer optics. To change frequency bands, the telescope pointing is offset slightly. The feed horn to antenna coupling is frequency independent; aperture efficiency is measured to be 75% at 100 GHz and is expected to be greater than 71% up to 270 GHz. Main beam efficiency is about 83% at 100 GHz. The total antenna surface error is 30 mu-m rms, corresponding to lambda 100 at 100 GHz.

By extrapolating a tipping curve to zero airmass, the residual telescope noise temperature is found to be 3 to 5 K, in good agreement with calculations. Teflon (PTFE) lenses, serving also as windows in 50 K and 10 K radiation shields, provide excellent blockage of infrared radiation but very low loss at millimeter wavelengths.

(paper presented at the European Workshop on Low-noise Quasi-optics MPIfR, Bonn, Germany, 12--13 September 1994)

This memo describes the optical design chosen for the new BIMA antennas (those manufactured by TIW). The coma due to the off-axis feed locations is calculated and the shape of the primary beam is calculated.

Gifford-McMahon refrigerators operating at ~4 K are currently used to cool SIS mixers on radio telescopes at the Berkeley-Illinois-Maryland array. The refrigerators are constructed by adding third stages onto standard 2-stage GM cryocoolers; Er3Ni spheres are used as the low temperature regenerators. The refrigeration capacity is 50 mW at 3.5 K.

Two refrigerators have been installed on telescopes thus far. The first has operated for 10000 hours. Its temperature fluctuates between 3.5 and 5.7 K, sometimes on time scales of minutes, because helium leaks intermittently past the seal on the third stage displacer. The temperature stability of the second refrigerator is much better. It uses a spring-energized lip seal (Bal Seal) which is mounted on the third stage cylinder, rather than on the displacer. This refrigerator has now operated on a telescope for 3000 hours. It maintains the SIS mixer at 3.0 K, stable to +-0.07 K over periods of days, even as it is tipped over an elevation angle range of 150 degrees.

The improved reproducibility afforded by the Bal Seal allows one to more reliably compare different regenerator materials. The refrigerator performance is almost identical with Er3Ni or neodymium spheres.

(paper presented at 8th International Cryocooler Conference, Vail, Colorado, 28-30 June 1994)

In this memo, we review the three basic schemes of obtaining low spatial frequency data for millimeter array observations: (1) use of a large single dish, (2) interferometry with a multi-size array, and (3) scanning with a homogeneous array. We describe an application of the homogeneous array concept using the three antenna Hat Creek array. In an appendix, we analyze the effects of random, uncorrelated pointing errors on the effective interferometer primary beam shape, generalizing an analysis originally due to Rachel Padman and since lost.

Here we review the sign conventions used by the ATCA, for data within an RPFITS file and within the MIRIAD system.

This memo describes the current state of the on-line passband calibration, and gives some examples of the astronomical passband calibration for data from the BIMA array. Polynomial fits to each interferometer baseline and correlator window currently provide the most reliable passband calibration.

Multiple pointing observations may be done either to get more uniform sensitivity over the primary beam for a source that nearly fills the beam, or it may be used to map a source that is larger than the primary beam. In either case, it is necessary to carry out a complete set of all the pointings while still within each uv cell of the source visibility function, in order that the source visibility be properly sampled. This may be difficult if the source is large compared to the beam, requiring many pointings, or if the desired angular resolution is high. This is because the observation may pass completely through the basic uv cell before it is possible to complete all the different pointings. It is always possible to carry out the observations by using tracks on successive days, but here we discuss the limitations to obtaining the complete sampling in one single track.

This memo describes a B array which, when combined with the present C array, improves the resolution, by about a factor of three, to 5". Unlike earlier B arrays, this one has only a small range of shortest to longest spacings and gives reasonable sampling of that range for a seven hour track.

This memo provides an easy recipe for calibrating and imaging data from the BIMA array. Images can be made automatically, although at present the data must be edited to obtain the best dynamic range.

The purpose of this memo is to consider two options for implementation of long baselines on the BIMA Array: coaxial cable and fiber optics.

We obtained optical pointing data for antenna No. 5 on 11-12 May 1992 using the unix operating system. An rms residual of 6 arcsec was obtained with a sin(2Az) residual pattern in elevation. We investigate the effects of large tilts and collimation errors on the pointing equations. An antenna tilt, or an elevation axis mis-alignment of 1', or a collimation error of 5' produce errors of ~1" rms in the pointing.

This cookbook is designed to help the new or slightly experienced user make maps from BIMA data using the MIRIAD data reduction package. Different methods are given for determining the success of the calibration procedure, and some detailed guidelines are presented on how to make dirty and clean maps. A familiarity with MIRIAD is assumed, such as how to run MIRIAD programs and how the observations from a single track on an object are organized into the different MIRIAD datasets. Advanced topics such as mint are not covered here.

This memo explores the optimum calibration of millimeter array data. The tradeoff between integration time on source and calibration, and the choice of calibrators is discussed. For strong sources, or at short baselines, calibration errors become more important. On-line calibrations of instrumental subsystems made with high SNR reduces the number of parameters needed to define the gain and passband functions. The gain calibration can be improved by fitting antenna gains to both sidebands simultaneously. This facilitates the selfcalibration and analysis of relative positions of images in both sidebands. For strong sources, or line and continuum observations, special care needs to be taken with the passband calibration.

A fiber-optic link for the distribution of LO and IF signals was successfully installed and tested on antenna #3 at Hat Creek. The link performed to specification; there was no measurable degradation of system performance over the standard coax cable system. Appendix A presents and costs a plan for upgrading the entire array to fiber-optic LO and IF distribution.

This cookbook is designed to help the new or slightly experienced user calibrate BIMA data using the MIRIAD data reduction package. Some detailed guidelines are given for editing the data and estimating the quality of the data. Some familiarity with MIRIAD is assumed, such as how to run MIRIAD programs and how the observations from a single track on an object are organized into the different MIRIAD datasets. Advanced topics such as selfcal and mint are not covered here.

It is well known that the drives on antenna 3 are presently unable to maintain pointing in the presence of even a small amount of wind. However the effect this has on data is probably not widely appreciated. This memo uses solar observations to point out that on a typical summer afternoon antenna 3 may be pointing off the source (i.e., the source is outside the primary beam) for up to 50% of the time without the data being flagged bad. These pointing errors must produce high noise levels on the baselines including antenna 3.

Tests were run on three of the Maryland SPARC workstations (a SPARC2, an IPC, and a SPARC1), running three standard MIRIAD tasks, INVERT, CLEAN, and RESTORE. These tests were run both using each machine's local disk and using disks on the other machines, in order to test the relative performance of the machines, as well as the importance of working on a local disk. Also, running time versus map size was measured for the three tasks, and a test was run with another user on the machine. The results of these tests is presented.

Antenna shadowing occurs if the projected baseline is less than the antenna diameter. This may effect the choice of calibrators in compact array configurations. Here is a script to generate a list of shadowed data. Lists of shadowed data for the c-array are appended.

It is shown that pointing offsets determined by program CROSS exhibit systematic errors as large as 70%. These errors are due to improperly fitting a Gaussian to the offset measurements in subroutine RCGAUS. The results of fitting model Gaussians with and without additive noise show that a non-linear fitting method based on the Levenberg-Marquardt is superior to the the cross-correlation method of RCGAUS.

This memo analyzes the effectiveness of using one or two outrigger antennas placed along the lab road combined with 6 or 5 antennas in the conventional "T" configuration. Each configuration is evaluated for three source models at declinations of -20, +15, and +50 degrees. UV tracks are generated for each model, and inverted with uniform weighting to yield a beam. Plots of the UV coverage and beam shape are presented for each potential configuration. Maximum negative sidelobe level and beam size are used to evaluate the quality of the configuration. The optimal location for a single outrigger antenna appears to be at a distance 1570N (feet) and 340E. The optimal location for a second outrigger appears to be at a distance of 2800N. If this location is not practical, a suitable alternate location is at 2440N 220W. These configurations yield sub-arcsecond resolutions with beam sidelobes levels approaching the 10% level for a full track.

We present a model analysis of interferometer data corrupted by receiver noise and gain instabilities, and atmospheric phase fluctuations which are correlated with total power fluctuations. If the atmospheric fluctuations are larger than the fluctuations due to gain instability, ground radiation etc., then we can estimate the phase errors and correct the data. We have developed software tools to estimate the correlation between atmospheric phase and total power fluctuations, and correct the data. A receiver gain stability of about 10E-4 is required, and the variations in total power from ground radiation must be smaller than about 0.2 K on the time scale of the atmospheric fluctuations.

Pointing data for 3 configurations Nov-Dec 1989 is analysed. The effect of antenna structure and solar heating on the pointing is also discussed. The parameters derived from optical pointing are reproducible after an antenna move. There is a large dispersion in the parameters derived from radio pointing, even with large data sets. The pointing parameters should not be changed based on small data sets. The radio pointing can be based on the parameters derived from optical pointing without significantly degrading the rms. Using the refraction computed from temperature and relative humidity does not significantly change the rms.