Science IPT telecon 15 June 2004 Wilson, de Brueck, Zwaan, Laing, Wright, Conway, Morita, Hills, Emerson, Holdaway, Mangum Wootten Carilli Hills: Initial report on antennas based on paper report is complete; members of the core team are taking the AEG report into consideration for a final report which accounts for this. Laing: Not a complete CDR for the backend, so there was some difference of opinion among panel members on whether it was passed. 1st LO and distribution were missing. Conclusion was a conditional pass, there will have to be some delta CDRs and a general tidy up at the end of the year. Report under construction. Digitizer is on track and looks good but is quite late and there are integration issues. Design OK. Fibers--do amplifiers need to be put in for DTS. Yrbium doped amplifier or replace the receiver or nothing at all. This should have been decided. EVLA card replacement found enthusiastic response. System specs also at issue as there was a disagreement with BE specs. Documentation was not in adequate state for CDR. IF processor was also a contentious issue--three designs presented. Overall most of the design work looked pretty good. Emerson: from the debriefing--end to end measurements of the system still missing. total power digitizer needs redesign, issues with speed of sampling. Preproduction review in Jan 2005. Sampling speed--originally a back of the envelope calculation. If this is done more accurately--2.5 to 3 bits per beam for example for Nyquist--500 Hz 950 GHz not even At 2 kHz one gets down to 1% loss of S/N ratio. Laing: Other issues--use of the TP detectors still somewhat vague. Laing will write on one sheet of paper what all of the issues are to make them clear. How do we relate the size of the bits with the actual power and its frequency distribution once it comes through the telescope. Laing: B7 review: IRAM had done an extremely good job on the prototype. A very nice mixer substantially exceeds the noise spec. Pretty good gain stability. Saturation not established but on the order of 1%. Many lab tests done, so this aspect was very impressive. Documentation and details for production and quality assurance need attention. 2SB 8 GHz design of prototype. Calibration device kickoff meeting. Three things will happen under Carter's leadership. Bordeaux and Madrid items under Guilloteau, Bacmann. Carter design multiload design (current one doesn't fit cryostat). One driver is that B3 and B7 rx thought vulnerable to saturation. But B7 isn't and we don't know for B3. Tests will involve getting a load from RAL which is temperature stable--spiky Al substrate with epoxy coating with temp controlled circuit measured against well controlled receiver in lab. Also same test with B3 evaluation rx at Socorro. Target date for completion of study end of October. No construction other than experimental setup. writeup will be circulated. ARCs: NAASC exists but no employees and no money other than the Head. EU elements consist of central node and smaller centers spread throughout Europe. STC of ESO and EuALMA Board and ESAC have approved the concept. Just now the ESO Council also has approved. A Draft call has been distributed for comments. REsponse with statements of interest 31 October. Configurations--Conway: Simplified version of baseline solution plan linked to agenda. The concepts have evolved considerably since then. Antenna specs have 30 micron repeatability on intersection of axes which works out well. Discussion also on whether delay stability of instrument interacts with baseline measurement accuracy. Now thought to be less of a problem. Go around 5-9 sources in as many mnutes, repeat many times in one direction and then in the other to avoid cable wrap. Constant delay cancels. Spec for drift is 22 fs per antenna in the cycle this is probably OK. A constant drift bigger than that might be OK--FS would take this out. In system requirements this is not spelled out adequately. Different interpretations could be taken if one did not refer to the LAMA memo. Drift not thought to be a problem now. Now can estimate accuracy, which is of course baseline dependent. If one does the calibration with the full array with all closure quantities one can define an antenna based delay error with a component of atmosphere and one of position. Relative error outside and center is the same number as if single baseline calibration. But all position errors which are accumulated do cancel--all antennas are on same coordinate system. Finish this document by the end of the month. Conclusions will be interesting--mode on how we do the cal, how often, which antennas are used from the rest of the array--only 1-2 needed. Specification on when solution is done--after move, again at night, apply retroactively. Also implies a min no of channels even for a continuum experiment. Need 0.5 cm antenna placement to start the process. Bandpass calibration may be a limiting factor but not time variable. This process may limit astrometry. Hills: on that type of task baseline errors only multiply the distance to reference object; with several surrounding the object of interest this is improved. Mangum: Cal group reactivated, to meet on 2nd Thursday at 1500 UT. July 8 will be first but falls on CIPT CDR. Holdaway: aips++ does indeed work fine with linear feeds and linear polzn so the alarm last month was unfounded. However, one cannot add Dterms--the polzn leakage terms, phase errors, reset random number generator seeds--so one doesn't get cancellation. MH to visit Socorro to visit the aips++ forces in July. Wright--all these things are in miriad uvgen if you want to go that route. Holdaway--have investigated the problem without Dterms and phase errors. So far believes that 1/f is not a problem for imaging. The 0.1% spec is ill conceived as for a bright point source there will be flux scattered into other locations in the field, so a weak spot nearby may be limited in what one can do for fractional polzn measurement. Laing: Read protomemo on gain flux which was great. Differential phase stability between channels most likely to cause a problem. We don't currently have a spec on this. ACTION: Propose a spec.