JEM-X Meeting: Monday 26th November 2007. DNSC Present: NL, SB, CBJ, JC, NJW, CAO Instrument Status: Quite noisy at the moment. Seems to be low-energy electronic noise around 2 keV. Could be due to sun on the mask, space weather, or raised background looking at north eclliptic pole. HV was set down last week in revolution 623. Observations are quite close to the Sun at the moment, so with a dithering pattern can mean a significant pointing towards the Sun. OSA Status: j_ima_iros 7.05 is finished. Need to study Crab response as a function of off axis angle to see how the various free parameters vary. NL to show JC where the program is so that he can take a look at this data with the new program. Pulse-height limit (on board) is not yet implemented in j_ima_iros, though it can sometimes swallow the Xe line if gain is very high. This is one of the fast rejection limits used to filter out particle background. Next step in j_ima_iros development is to tackle electronic efficiency. CBJ has already made these tables and must find them for NL. Spatial gain variation is also implemented, and NJW's j_shadowgram now calculates average detector gain for each SCW if it can't find CAO's output -GAIN-SCP produced by j_cor_gain. NL's program also takes into account the onboard `curtain´ correction. SB points out that the electronic efficiency has a very sharp edge around PHA=120 due to nature of rejection criteria, and this plays an important role being so rapidly varying over the lower PHA values, and cannot be fitted by the averaged electronic efficiency produced by CBJ, which also is not spatially varying but an average over the detector. This spatial dependancy could also couple into the collimator tilt, or be misinterpreted as such. All these considerations are of course most important for very weak sources. NL has also introduced a new bit flag that can be used by mosaicking, to signal the presence of a found source in the image. SB, Spatial gain map and position corrections: SB discovered that archived corrected data is garbage - which is why we recommend that all users should start in COR level and create the their own new corrected data with the latest software and the latest calibration tables. Us SDAST folk have complained about this for years. Archive needs updating with the latest software and tables, and especially the new timing correlation. SB looked at event lists for 100 orbits and made spectra for each pixel (about 10000 counts per pixel). Also looked at Xe line for each revolution and the energy resolution, which agrees with CAO's results. Xe line varies by about 0.5 keV around about 29.6 keV, and this variation has no real significance. Resolution has degraded from about 9% to 12-14%. Molybdenum line is clear but lies on a steeper background. Clear grounds to believe that a new SPAG table would improve the energy resolution of the the instrument. Each pixel is changing its gain as a slightly different rate, giving a spread in the energy data. There are some pixels however where it's hard to see and fit the Xe line, but an extrapolation based on neighbouring pixels gives and improvement in the Xe position. For each RAWX (anode) during the course of these orbits, it can be seen that the relative position of the Xe line changes quite a bit, though in a sort of pattern, not a pure noise and not a monotonic development. We must be sure that the new SPAG table is clean of statistical noise, so that systematic changes are the most significant and come through. Ideally we would have a SPAG for each IMOD epoch, but this is limited by statistical considerations. SB's also looked at the detector counts (20 -25 keV band) and can clearly see dead anodes, some of which work for part of their lower length, and regular jumps in countrate along the Y direction, though exact pattern of the variations differs from anode to anode. This has to do with the size of the individual pixels as they are calculated by the onboard software, the RAWY position determination software. This may need to be taken into account in the PIF calculation. Bright (weak) horizontal strips are abutted by dark (stronger) strips, suggesting that missing counts in one pixel are picked up by its neighbours. Our current position correction table is based on Ferrara data, but this was plagued by a problem in the X/Y scanning table, whereby there was a one mm difference in X and Y motion over the length of the detector. Current corrected positions show irregular jumps in Y position moving along the X axis. Also the distance between Y values (anode positions) are not constant. SB not sure that the sharp changes in coordinates are actual physical differences. Are we just encapsulating measurement noise in these corrections? Using these values as the centres of each pixel will change the effective area of each pixel quite a bit. SB suggests smoothing out the wiggles in each dirrection. Some of the clear patterns across anodes are probably due to the way the Ferrara measurements were performed with the collimator in place and a finite-sized pencil beam that would be blocked or reflected by the collimator. Other places, the position measurements are somewhat chaotic. It seems that the table should be processed to smooth out these wiggles and chaotic regions, and generally correct the problems that are known to exist in the Ferrara calibration data. Could an in-flight position correction table be made with a single bright source. At lower energies we should be able to see the collimator in the detector counts shadowgram. Are some of the st ripes we see actually the collimator? SDAST Meeting: SPAG should be added to the agenda. Supper will be Wednesday evening, though CBJ and CAO can't come to the meal. JC will come with some good examples of how a joint JEM-X/SPI cookbook could be an advantage. Copenhagen 2008: NL has made an application letter for funds to be sent round to various agencies, but the budget maybe needs to be updated: national agencies should support students and other needy people who might want to attend. Pens, paper, bags etc. need to be found, with `coporate identity´ stamped on them. DTU probably knows a supplier. What about corporate sponsorship from the danish firms that helped build the instruments? A firm could sponsor the welcome drinks or similar. Research and Innovation Agency will need a full budget with the application. AOB: None.