V2C Telecon Notes, April 25, 2013 --------------------------------- In attendance: Arthur Niell, Chris Beaudoin, Brian Corey, Chopo Ma, John Gipson, Dirk Behrend, Dan MacMillan, Zinovy Malkin, Johannes Boehm, Mamoru Sekido, Bill Petrachenko. 1. V2PEG presentation at the EGU -------------------------------- Considerable time was spent discussing the plots of scale and x/y-pole difference (VLBI - IGSfinal) for CONT11 in the context of R1 and R4 performance. Both in terms of formal error and scatter the CONT11 results are considerably better. It was pointed out that part of the improvement is due to the fact that a lot of care is taken to ensure that stations are operating at their best in CONT11; but another important factor is that in the CONT sessions, the same network is used every day so that network bias is not a factor. In R1 and R4, a core set of stations is used every time with the rest changing according to station availability. It was suggested that perhaps it would be interesting to plot the R1 and R4 data with different colours to see if there is systematic bias between them or perhaps to select a subset of R1 and R4 sessions with very similar station composition to see if that would improve the R1/R4 data and support the network bias hypothesis. Another suggestion was that perhaps once a month a very high quality (many stations) R1/R4 session could be performed. Since there is fixed station availability this would need to be done somewhat at the expense of the remaining R1/R4 sessions. It was also pointed out that the more precise CONT11 data showed what appeared to be significant systematic trends. There was some speculation as to possible origins of the x/y pole difference (VBLI-IGS) trends but no physical cause could be suggested for the the apparent ~2 wk sinusoidal trend in the scale data. It was agreed that a longer set of data at this level of precision or better would help resolve if the trends were real and if so to help discover their origin - something to look forward to with VGOS. An interesting aspect of the x/y-pole difference (VLBI-IGS) data is that the rms scatter is smaller than the formal error. This probably indicates that the VLBI and IGS data sets share a correlated error source, e.g. the atmosphere. Based on a presentation given by Jim Ray at EGU2012 the IGS final accuracy is <30ps which is approximately the same as the scatter of the differences with VLBI (not including the apparent linear trends). However, the IGS formal error quoted with the series is considerably smaller and according to Dan decreased to this value suddenly not too long ago. Also from the same presentation it was noted that the VGOS network expected for about 2017 is weak in the South America/South Pacific/ Antartica region with nearly half the globe without sites. However, it was agreed that nevertheless the expected network is still quite good. Straw Man VGOS Operational Plan ------------------------------- A question was asked about correlation requirements which are not mentioned in the plan. It was pointed out that for early operations (2014/2015) current correlators are probably adequate and that with software correlators the development cycle is much shorter than with hardware correlators. It was also mentioned that Walter Alef gave a talk at EVGA about the next generation Bonn correlator in which VGOS requirements were considered. It was suggested that better defined targets (e.g. specific stations) be included for the years 2014 - 2016 as was done for 2017. It was mentioned that this is also a concern of the V2PEG and information is being requested from the stations to add detail to the plan. A question was asked about network requirements for the period of the plan. Although as much data as is practical will be transmitted in near real-time, it is recognized that data shipping will still be required into the foreseeable future. 2. Broadband system development and tests ----------------------------------------- Arthur and Chris needed to leave at the end of the hour and weren't able to make a report. Although Gino couldn't attend he sent a report for the DBBC3 which is inserted below. ***** From Gino ***** The dewar for the broadband receiver DBBR is under construction. The feed under test showed good performance in the first measurements and probably a slight revision will be realised to move higher the range in the lower end. Indeed it seems to be actually not worth to be sensitive around 2.3 GHz for the strong RFI present in Noto. LNAs are still an issue and until now different amplifiers have been used in different ranges. It would be very useful to have summarized somewhere the different options available. This receiver is designed for a 12 m Intertronics antenna illumination but for now the receiver will be adapted to the Noto antenna with a tertiary mirror, still being usable in this actual VLBI2010 telescopes. Output is the full broad band as analog and fully digital within a set of network connections, after sampling and band selection. The broadband sampler devices will be available in late summer and the first complete unit will be adopted in this receiver. The firmware for both Core3H and Core3L is under development and a prototype board is used for the testing within a real Virtex7 device. The goal is to have a complete front-end DBBCR-DBBC3 system to be coupled with a DBBC2010 available by the end of this year. The FILA40G design has been completed and the first two units will be realised in the next months. SAS2/3 Interfaces for recording on external SATA diskpacks are provided. ***** From Gino ***** 3. RFI ------ Brian reported that a report on the V2PEG RFI survey is being prepared. He also mentioned that at the ERATEC-RFI workshop in Bonn advice was given that publishing information on RFI is not permitted by law. As a result of this the RAEGE RFI reports were removed from the web. It is not fully understood exactly what the legal situation is and what the implications are for the the V2PEG RFI survey report but Brian is looking into it. 4. Source Structure ------------------- Johannes reported that Jamie McCallum and Stanislav Shabala visited Vienna from University of Tasmania and added a source structure element to the VieVS Monte Carlo simulator. The idea is that, based on the source structure index of the source, an offset could be added to the source position. A core-jet model for the source was assumed with the orientation selected randomly and the offset along the jet proportional to the structure index. This could be used to investigate the impact of source position offsets on station position, EOP, or scale bias. It was also mentioned that Chris Jacobs had enquired about the expected size of VGOS source position errors. Dan recalled that he had intended to look into this and would put it back on his list of things to do. 6. Schedules ------------ Johannes mentioned that Jing Sun is back in China and that he has reestablished communications with her. She will continue to work on scheduling issues. Dan mentioned that work is continuing at NVI on mixed mode (i.e. broadband and S/X antennas in the same session) scheduling. The issue is that the smaller 12 m antennas are not sensitive enough in S/X mode and lead to lengthened integrations. To avoid the resulting degradation, a mode of scheduling is being investigated in which an S/X only subnet and broadband only subnet are scheduled separately for a fixed period and then all antennas are scheduled together for another period. It was cautioned that an important aspect of mixed mode sessions is to tie the legacy sites to the broadband sites which would be degraded through the use of the separate subnets. No solution was offered but it was suggested that the ratio of the different modes (separate and joined subnets) could be varied depending on the purpose of the observation. 8. Atmosphere ------------- Johannes reported that elevation angle cut-offs were being investigated with the VieVS simulator. For sessions involving many stations it was found that cut-offs of 10 or 15 degrees worked better than 5 degrees. The conclusion of course depends on whether or not the simulator atmosphere turbulence model is realist at low elevations. Japan ----- Mamoru Sekido reported that NICT is planning to test its fixed frequency broadband (3-14 GHz) system in June or July. The test will use the 34 m antenna at Kashima and the 1.6 m transportable antenna. Single polarization will be used and the 1.6 m receiver is ambient temperature. The choice of frequencies and receiver architechture is dictate by budget constraints and issues related to the optics of the Kashima antenna When asked about progress of the GSI 12 m antenna at Ishioka he reported that completion is expected in spring/summer of 2014. Australia --------- Jim Lovell could not attend but he sent the following report: ***** Jim Lovell ***** We recently secured operations funding until the end of 2014. Its more than we have had previously which means we can use all three antennas for every session (up from typically 2 per session) and we can double the number of observing days per year from 70 to 140. 2. We're continuing to work with Callisto Space on a Stirling-cycle cooled 2.2 - 14 GHz prototype receiver. Callisto are currently cryogenically testing a LNA from Low Noise Factory and we've just put in an order for design and construction of a 2.2 - 14 GHz QRFH (from Caltech: note the higher freq coverage compared to the Goddard feed). Callisto will test this in a dewar with the LNA at 70K. We currently don't have the funds to build a full prototype receiver for Hobart12 but this is a big step forward. We're expecting results on the LNA-only tests any day now but there's a ~6 month lead time on the feed. Steve Rawson (CEO of Callisto) is visiting Hobart late in May. 3. RFI at Hobart is about to get a lot worse at S-band due to wireless broadband internet transmitters to be installed in the area. Frequencies *above* ~2300 MHz are likely to be affected (the upper 3 S-band channels in most of our IVS observations). We're worried that our LNA may be saturated at least for some pointing directions but probably won't know for sure until the transmitters are switched on. Best case is probably losing everything in S-band above 2300 (the upper two S-band channels). For VLBI2010 though, it may be better for us to have a feed with a low frequency cutoff above this new interference, say 2.5 GHz. ***** Jim Lovell ***** There was some discussion of how quickly feed performance degrades below the lower cut-off frequency. Brian thought that the cut-off was quite sharp perhaps down by 10 dB within a few hundred MHz. Chris offered do some analysis to come up with a more definitive answer. ------------------ Next v2c telecon will be Thursday May 23.