GBT Test/Maintenance Requirements for Trimester 06C

TBD

• This covers Ka-band, Q-band, CCB, Penn Array, Zpectrometer, PTCS, and all other requests.
• For 06C we need:

NOTE: in the above table "1 shift" means one 8 hour observing shift, on average (some longer and some shorter will be wanted).

A - clear skies (low opacity), low wind (less than 2.5 m/s), thermally stable (00:00 - 08:00)
B - clear skies (low opacity), low wind (less than 2.5 m/s).
C - clear skies (low opacity)
D - any observing weather.
E - can be done in parallel with maintenance, but needs antenna at access and M&C available. When scheduled this work may exceed the time required for other maintenance activities.

Summary: We must preserve at least 1/3 of each week to high frequency science with an effort to have 2/3 of the time for science whenever possible. Penn Array will use 1/3 of the good weather days in Sept. Between Oct 1 and Nov 1, Penn Array, Q-band, and Ka-band commissioning will use 2 shifts of A weather per week. That's about 2/3 of the good weather. In November, Q- and Ka-band, along with the Z-Spec and PTCS tests should take up 1/2 of the good weather time. All commissioning activities seem to be scheduled for completion by end of Nov. That's appropriate since after the end of Nov. we have the holiday season and would expect little progress until the start of the new year. RonMaddalena

Total 06C requirements: 0 hours

• Characterization of the C-band receiver baselines before modifications (two nighttime shifts). Date?

No longer needed -- see below RonMaddalena

• ?

At the end of Cycle 6 of 2006 and Cycle 1 of 2007 only (in the following order):
(i.e. end of September 2006 and the end of January 2007)

Approximately 6-10 days prior to release date:

• Two blocks of 4 hours each for integration tests; full system and antenna. Preferably during a weekday.

Approximately 2-6 days prior to release date:

• One block of 6 hours for regression tests (full system and antenna) followed by around 8 hours of commissioning time.
  • Notes: The commissioning time could be used for calibration tests, pointing, etc. The theory is that if the four hours of regression tests went horribly wrong, we'd revert to the current M&C version, the commissioning tests would then use the old version, and we could re-schedule. If the four hours of regression testing went well, the next eight hours would be a "sympathetic" observer using the new version, and a chance to catch/fix trivial problems. We'd revert to the current M&C version at the end of the commissioning tests.

Within +/- 2 business days of end of development cycle:

• One block of 4 hours for integration tests (to resolve/test any issues/solutions found during regression testing).
• We will schedule our release for a scheduled maintenance day. No additional time required. Ideally, it would be good to have a "sympathetic" observer scheduled to use the system directly after maintenance that day.

Averages 18 hours of time per cycle.

-- AmyShelton - 19 Jul 2006

• PF1 - Two weeks in the lab Jan or Feb 07.
• C band
  • Assuming the Penn Array is on the GBT starting about Sept 1, we would like to do the C-band upgrade in two phases: Phase 1 consists of the refrigerator upgrade and would be accomplished Sept 1 - Oct 15. RCO on the GBT after Oct 15 (or when the Penn Array comes off). RCO should be straightforward (2-3 hrs) since none of the electronics components change, only some cabling. The receiver could then be available for astronomy during November and December, including the global VLBI run in November.

• • Phase 2 consists of the replacement of the room-temperature RF electronics with an integrated package designed by Mike Stennes. We could begin that on Jan 2, and would like 5 weeks in the lab for installation and thorough testing, including with the Lab Spectrometer - so RCO would fall in 07A. Note that there might be some minor receiver manager (a few digital control bits) and Cleo changes needed during this phase - the details are not yet defined. I anticipate having a design review during the Fall at which the details of the integrated package will be presented, and final decisions will be made.

• • We can be flexible about the timing of the two phases in order to minimize impact on the telescope schedule (do most of the work whenever the Array is up for instance). For example, if the Penn Array does not go up ~Sept 1, Phase 1 could be delayed. And, Phase 2 could also be pushed later to mesh with the Array schedule. However, for scheduling purposes, we should assume Thanksgiving week and Christmas - New Years week are "nothing accomplished" due to holidays and vacations.

• Oct 1 - Jan 31
  • One maintenance day (08:00-16:30) per week dynamically scheduled (Tue/Thu)
    • priority of paired day possibilities: Tue/Thu, Wed/Fri, Mon/wed (confirm)
  • Every Friday morning 08:00-12:00 fixed or dynamically scheduled

• Twelve hours are requested for collecting gain calibration data once per month
  • Every other month, a contiguous 12-hour session LST 0500 to 1500, in good weather.
  • In alternate months, the 12 hours can be in 4-hour segments scheduled as convenient.

The good weather sessions will be for determining gain curves, the other sessions for SCALs

Due to heavy commissioning schedule in Oct - Dec., we can reduce our needs to a single 12 hr run in that time span. In January, we can go back to one run per month. RonMaddalena

-- Frank Ghigo and Ron Maddalena

• Avoid scheduling project changes exactly at 0700 or 1900 ET to minimize operator overload
  • Attempt to have project changes around 0730 and 1930 ET instead of at 0700 and 1900 exactly
  • If not possible, attempt a half hour earlier around 0630 and 1830 ET