NATIONAL RADIO ASTRONOMY OBSERVATORY

MEMORANDUM

DATE: December 8, 1998
TO: Mark McKinnon
FROM: R.J. Maddalena
SUBJECT: Resources needed for GBT holographic tests on the 140-ft

A goal of the NRAO is to commission the GBT in as fast a way as is practical. One step in commissioning the GBT will be holographic measurements. However, a large fraction of the electronics and software we will be using for holographic measurements has yet to be tested. Many aspects of the software and electronics (for example, the dynamic range of I.F. and backend components and the response of these components to transient and large changes in power levels) cannot be satisfactorily tested in the GBT mockup.

To reduce the time for commissioning the GBT we want to test before commissioning as much of the hardware and software by making holographic maps using an existing NRAO telescope. As you know, the decision is to use the 140-ft for such a test. We already have holographic measurements of the 140-ft to which we can compare measurements made with the new hardware and software. And, the GBT M&C system, as proved by my tests two years ago, already knows how to command the 140-ft through a holographic measurement.

This report summarizes what resources we would need for performing a holographic measurement of the 140-ft using as much of the GBT hardware and software as is practical. Except where noted, my estimates are for the extra effort need for a 140-ft test and do not include the efforts that would be expended even if we didn't perform the test. To generate this report, I have canvassed the engineers and programmers who are responsible for the components we will be testing on the 140-ft. My gut estimate is that the 140-ft tests will cost NRAO one month of extra labor but should save us about one month in commissioning the GBT.

I. Hardware required:

• GBT/140-ft prime-focus holography receiver.

According to Steve White, with two weeks notice NRAO can reassemble the receiver we used in 1988-1992 for holographic measurements of the 140-ft and 45-ft telescopes. Due to the differences in telescope receiver packages, we cannot test the GBT holographic receiver on the 140-ft. However, it might be possible to use the L.O. and block converters that will be used for the GBT. Steve is aware of what we would like and, with only a few weeks notice, we can decide what receiver components we can use.

• GBT holography backend.

The GBT holography backend was completed more than seven years ago and tested both in the lab and on the 45-ft in a holographic experiment. Thus, we need a minimum of hardware testing in the mockup. However, we will need to test the software interfaces and data collating software for the backend. For the 140-ft experiment, we will most likely decide to keep the backend in the mockup and pipe, via fiber, two I. F.'s from the 140-ft to the lab addition.

• GBT L.O. (can be replaced with 140-ft L.O. system if necessary).

The decision as to what L.O. system we will use depends upon the decisions Steve makes regarding receiver components. Most likely we'll either use the existing 140-ft L.O's or will carry to the 140-ft a synthesizer typical of what we'll use on the GBT. In either case, the effort in providing a L.O. should be trivial.

• GBT I.F. system (can be replaced with 140-ft I.F. system if ).

We will attempt to make observations using an I.F. piped from the 140-ft to the mockup using the two fibers that already are in place. Except for the I.F. processing that will be required to get the signals onto the fiber, we'll be using the same hardware setup as we'll be using for the GBT. Some minor work will be needed to prepare equipment to send the 140-ft I.F. onto the fibers. This step will need to be done anyhow as we are planning to perform other system tests with the 140-ft, the spectral processor, and mockup equipment.

• 140-ft telescope.

Except for mounting the receiver the day the tests start, and some minor preparing of I.F. and L.O. equipment, the 140-ft is pretty much ready for the tests.

II. Software required:

• GBT M&C

We will need macros for starting, stopping, and pausing a holographic mapping set; this includes pointing, focus determination (maybe), and starting from a designated point within a map. I tested these routines two years ago but they might need tweaking.

• Aips++ and UniPops

We will need the ability to generate an ASCII table of telescope positions and data so that UniPops can read and process the data. This will allow us to compare the old 140-ft data and the new data using the same data reduction package. By the end of February, it is imperative that Aips++ provides us with a 'table filler' for the holography backend, something that has been on their to-do lists for two years. Note that this effort would be needed regardless of whether we do the 140-ft tests; only the deadline for preparing the software will be different.

• Aips++

We would like to have holographic data reduction routines within Aips++. The Aips++ software need not be available during the test (since we can feed the data into UniPops's suite of holographic data reduction software) but we would like the software to be completed soon afterwards. Again, Aips++ needs to work on this regardless of whether we do the 140-ft tests and only their deadline has been pushed forward.

III. Personnel:

• Receiver and backend engineer (Steve White) and technicians -- requires two week lead time for preparing the receiver, I.F., and L.O systems. Probably will be involved the first couple of days of the experiment
• Telescope control programmer (Joe Brandt) -- requires six weeks lead time working approximately two days a week. During that time, Joe will need one business day per week when the 140-ft can be tested using his software. The other day per week he'll be working on software for the tests. Probably will be heavily involved the first couple of days of the experiment.
• Other M&C programmers, in particular Arno Granados (in charge of the holographic backend code) -- Since the backend software has not been tested, we can expect that bugs will need to be fixed. Our mockup tests will flush out most of the interface bugs before we start the 140-ft tests. M&C will need a lead time of five weeks, just to be safe.
• Aips++ programmers (probably B. Garwood and J. McMullin) -- The GBT holographic filler will take about a week but the holographic data reduction software is a many week effort. The filler is required by the end of February while the data reduction software can wait 'till soon after the completion of the tests.
• Software and hardware testers (R. Maddalena, J. Braatz, and G. Sandell) -- requires two weeks to test the holography backend in the mockup, at most one week to prepare the data acquisition and analysis software, one week for preparatory work, one week for the tests, and, in post-processing, three to four weeks with UniPops and two to four weeks after the completion of the Aips++ holographic data reduction system. Thus, about twelve weeks work spread among three people. The time should be spread over 4- to 5-month period and with a lead time of three to four months before the start of the tests.

IV. Time line:

• The lead times requested by those involved indicate that, if we decide today to perform the tests, the tests cannot be scheduled until April at the earliest.
• The tests cannot be scheduled any sooner than about two months after Aips++ provides the holographic filler.
• Once scheduled, the preparatory time requested by S. White, J. Brandt, A. Granados, R. Maddalena, J. Braatz, and G. Sandell must be allowed for. For example, Brandt will need one day per week for six weeks on the 140-ft for tests.
• The tests themselves will require five days of 140-ft observing time.
• After a successful completion of the tests, approximately four weeks of data processing.
• After the completion of the Aips++ data reduction software, approximately two to four weeks of data processing.