Performance and Reliability of the EVN; April 2009 ====================================== Session 3/2008 -------------- The session consisted of five frequency blocks (7mm, 13/3.6cm, 6cm, 18cm, 1.3cm). In each block, NME including ftp tests and calibration experiments were carried out before user experiments. The ftp data were processed in a semi-automated way. In the first 7mm NME experiments N08Q1, we successfully detected fringes from all the participated stations (Eb, On20, and Nt). In the other ftp tests, Urumqi had no fringes in N08X1 due to the wrong LO frequency. Its data were saved in the user experiments EK028A. Shanghai had no fringes in N08K1 and unfortunately the reason (poor pointing) was found after the session. There were 16 user experiments (1 at 7mm, 2 at 13/3.6cm, 7 at 6cm, 5 at 18cm, and 1 at 1.3cm) in the session. There were four 1024 Mbps expriments (GW019B, EF021A, EF021B and GP047) scheduled and observed. Russian QUASAR stations participated three EVN expriments (EK028A, EK028B, and EK028C). Yebes 40m radio telescope started to perform the observatins at 13/3.6 and 1.3 cm. Station and correlator feedback for individual stations: Ef - Lost some hours due to problems with the antena control computer in EK026 and EF021B. Wb - Used in phased array mode. MK5B formatter gave error 'Mark5B sync error gt_3' in GP047, EF021A, and EK028B. Resynced the FS time a couple times in GW019B and EK028A because the FS computer clock was synced by ntp server but the 'rate model' was still on. Subband 7 polarizations wrong (bad X-Y phasing) in RT006D and RF006A (24-25 March 2009) Nt - BBC08 was in overflow during a long part of the experiment GW019B. The devices ia and ic were in overflow from 10 to 12 UT in EZ018B, probably because of the presence of satellite interference. On - Quit the observations of N08C3, EF021A, EK028B, and GC031B at 6 cm due to too much wind and mechanical problems. Lost some scans during the experiments of EL036B and EK026 because of strong wind. Lost noisediode and phasecal after around 21:30 UTC due to problems with the Power Supply in EK028C. Tr - No Tsys data till 7:15 UTC in EF021A. Hh - No observation in the session due to the serious polar-mount bearing problem. Mc - MK5 recording problems led to some scans being missed in EF021B. Lost 3.5 hour data due to a bad disk bank in EK026. Lost fringes in RT006D and RF006A (24-25 March 2009) because of a clock jump. The reason was not identified during the observations. Ur - Lost fringes due to wrong LO frequency in N08X1 and GW019B. Low amplitude in IF4/RCP (BBC08) EL036B. No fringes in BBC03 LSB in the whole session. Sh - Lost fringes at 1.3 cm due to poor pointing casued by loose screws. There was something wrong with the LNA, and the problem was solved in EK026. During EF021B, the operator had to type the gpstime command because formatter failed to do sych test from UT 6:56 to 8:34, and restarted MK5 at 8:20 UT because the pointer of the MK5 was not increasing. Ys - No fringes at S and X bands due to a wrong setup in EK028A. Mh - No fringes in IF 5 and 6 (LCP) EB037E due to the dead BBC06. Jb - No Tsys data available at 1.3 cm. LT had initially low sampler statistics in LCP subbands in the preparation phase for the e-EVN run RP015A (10 February 2009). Fixed by increasing the VC power levels. The LT had to stop one hour after the project started, because of an azimuth rail crack. Cm - No known problem. Some of the problems detected during e-EVN runs can be easily fixed. On-the-file tools exist for data quality checking, which should be carefully monitored by both JIVE and the telescope operators. Jun Yang for the Science Operations and Support Group (JIVE)