Sci IPT
Nikolic, Hills, Wootten, Laing, Morita, Wilson, Kniesel, Hogerheijde, Wright, Brogan, Testi, Wilson, Brogan, Lonsdale, Hibbard, Hunter, Schwab, Myers, Remijan, Richer
PAI 2nd quadrant correlator competes with this meeting.
RH: First Fringes. Continue work. Dynamic fringes for the first LO are OK but second LO still not quite right.
Status: Installation of new software v6.1 has been intense. Two more weeks on SD on the three accepted dishes. Partly to complete amp cal. OPT offset pointing still needs some work. Thermal deformations: Melco showed significant deformation, more insulation has been installed on one antenna which has begun holography tests. On Vertex, not obvious that the deformations take the antenna outside of spec. More tests starting next week.
KIM: Melco completed mods on one antenna. Began measurements but problem with rx has slowed progress. Still expect good results by end of month.
RH: Good focus curves, x,y,z of subreflector as f(el) but night to night variations were traced to hexapod on DV01 which was replaced. Most recent data looks good. Generally, antennas and other parts of the equipment need better reliability for us to make progress according to schedule. These are early days and we expect reliability to improve. Taking them to the high site will require better reliability. DV02 has gone through first phases of its program, fully equipped, doing focus, beam maps and so on. EU antenna is active, SS has arrived. Assembly will begin soon.
RH: Antenna stations are being prepared for antennas. Foundations are completing well. Power and fiber temporary for A106, hoping for the final hookup for the next three. In the outer parts of the array some areas have had soft soil or permafrost and some stations may need to be modified by some hundreds of meters. See Sawada's notes.
MW: Compare the configuration with what happens when an antenna is dropped. It should not be worse than this.
Most northern location is close to road. This is a security issue. Some medium scale theft has occurred; this may not be wise placement for this antenna. The antenna may have to be moved nearer to the guardhouse, lessening the longest baseline to just over 15 km.
CB: Tutorials given in Canada. Few know exactly what it is.
RH: Most feel this acceptable.
Review of commissioning plans. This has not seriously moved forward. Need to decide on overseas members, one from each partner from ASAC, at review.
LT: ESAC met and discussed. Rafael Moreno may join. EA will meet next month. May be ESAC or ANASAC member rather than ASAC.
Commissioning Scientist, working on short list.
CB: Are the dates set? Can one attend by telecon?
RH: I would be loathe to move it. At the OSF.
RL: Are you likely to want written input and if so, when?
RH: We need to revise and circulate the plan over the next few weeks. By end of July this should be done, send to committee by two weeks before review.
Updating focus. This has been discussed, some data from AIPT on actual performance of the actuators. All have better resolution than specified, so there is no objection to continuous updating in principle. We ask CIPT to implement this. This can always be disabled. Mel thinks that it should have the ability to stop.
RL: Similar scheme on the WHT with similar parameters, this has worked very well.
SciSpecs. Nothing new
New issues: IF system attenuators in two places, one in IF switch (changing between front ends) also in digitizer to account for BP slope and adjust the 2 GHz sections. Switches have 10-90% time if nanosecs can take up to 5s to settle their last 5%, not even with 1% in 1.5 sec spec. This can cause a problem for fast switching. Phase is within spec after 1.5s. No immediate solution in hand, replacement would be significant rebuild. Intermediate solution is change the IF switch which does most, but leave the processor switches the same. When we look at cal source (phase cal) getting exactly the right level is not so critical. We could take a hard line but is this OK. This is a few tenths of % error dying away over seconds.
MW: Declare success if you can get below 1%.
RH: We have high specs on dynamic range, Walsh cycle can take 2 seconds, so perhaps amp changes during that will affect cancellation.
MW: How much does power change--may not want to change the first attenuator but this will have ramification on BP cal. If you only want phase, this may be OK. You can average over BP for phase.
RL: During Walsh switch could introduce closure error, could be a problem. Do we change it too often? Does it need a change most times?
RH: I think we should have had a variable attenuator in each front end. Just one in each IF system should be changed. We can tolerate the second one.
KIM: Only amplitude change?
RH: Measurement shows it is always within .25 deg within the settle time. Switching does change phase.
TH: The phase measurement was CW so there could be a gradient. I sent spreadsheet to Saini which has Tsky for B7-10 for weather conditions appropriate to those bands. He is going through a worst case scenario.
RH: Perhaps you could circulate that. I will send the report around from BE.
Nonlinearity of the Front Ends. This was discussed in the calibration telecon last week. We've known this for a long time. Naive calibration using an ambient load results in an incorrect answer by a few per cent. There was a lot of work on small signal injection in the early days. We ended up with ambient and hot load however owing to problems with those schemes. I did not realize the compression, 5 to as much as 8%, was as large as this. Spec is 5%, interpreted as differential LN2 to ambient. Actual amount can be larger.
RL: Design review numbers were smaller.
TH: I can get a discussion with the B6 people before the next meeting.
RH: Pavel sent a subsequent email with additional reports. We can try to characterize the nonlinearity or we can again look at low level cal signals, i.e. blocking small part of the beam. Anything has complications.
MW: I discussed diffraction some. This needs lab measurement. I think this is attractive. Need a large LN2 to calibrate in the lab, better than on the sky. On the sky, diffraction needs calibration. Anything needs calibrating, a clean nonsaturating load seems desireable.
RH: Role of cal source overall is unclear to me. In practice we rely on sources of known brightness.
MW: Yes, this is a secondary calibration. Bandpass cal poses similar problem--often one uses a noise source, which introduces its own problems also.
RH: We assume all BP cal is on astronomical sources.
MW: At the highest frequencies maybe not.
RH: Let's think about this. We have a project to build an improved cal load. That is producing a very elaborate scheme but expensive. Review of cal system in October.
RL: EU seems satisfied for a 1 year no cost extension to WVR project.
AR: Good progress on including spectral line database in OT.
SM: New CASA release end of week.
CB: Gradually learning how to do tutorials better. A challenge occurs when novices are the tutees.
RL: I agree, that was our feeling here.
CB: This isn't a CASA issue, but a student training issue.
RH: It worries that the consumer may not really understand their image.
MW: Ultimately inline processing must be improved. Concept of offline data processing.
CW: The
McMaster workshop had a focus on grad students, so the group was less experienced. There were lectures on principles followed by a day of handson. Feedback I received was that with some prior exposure people gained more from the workshop. Our goal is to have a robust pipeline eventually, but not in early science.
CB: The pipeline will eventually do a very good job on calibration but there is no way the pipeline can guess the scientific intent, and tweak the image to realize that intent. People will need to know something.
RH: We need to do some mitigation to bring reality in line with the promises which have been made.
JR: State of simulation?
RH: I was not on the telecon.
CB: We joined in the end. The issues are that many subsystems are involved and a clear prioritized list of work needed needs to be made. There are online and offline issues. Each affected subsystem will be polled for prioritized list, then planning will occur.
RH: Originally there was a simulation tool available in SSR. Two other possible needs for simulation have come up--one is during commissioning, when we try to understand the instrument's affects, and another for design of the pipeline. If we have a new list of requirements then we need actions, which is difficult. BN has floated a proposal for a virtual meeting on simulation
BN: Since the last workshop there has been a year passed, we thought it would be good to assess where we stood via a virtual meeting.
RH: I endorse virtual meetings. Whether more generic workshops are needed or management decisions are needed can be discussed. I suspect the latter is needed. CDR7 is next week.
--
AlWootten - 2009-06-17