Metsähovi Radio Observatory, Finland (vlbi@hut.fi) EVN Station Report, November 2004 Ari Mujunen (Ari.Mujunen@hut.fi) Jouko Ritakari (Jouko.Ritakari@hut.fi) Highlights MRO has continued using and developing the PCEVN data acquisition system, especially the eVLBI features. In the eVLBI data transfer test in Sep-2004, a 640 Mbit/s disk-to-disk speed record from the Otaniemi campus to JIVE was achieved with the PCEVN system. The experiment was extremely well received by the networking community and it was published in the EU Information Society Technologies newsletter as an example of European networking success story, for details see 'http://www.cordis.lu/ist/rn/ri-cnd/news_oct_04.htm'. At this moment MRO has delivered a total of twentynine PCEVN-based disk recorder systems, seventeen of them to Australia. Geodetic VLBI Project The S/X receiver built by TTI Norte, Santander, Spain, features standard geodetic RCP-only S and X bands (not wide-band X). The system temperature of this receiver is around 80 K when the LNAs are cooled to 15 K and the bulky feed system and polarizers are at room temperature. The geodetic S/X receiver is owned and operated by Finnish Geodetic Institute and is thus not generally available for EVN observations by default. PIs wishing to use the receiver need to contact Metsähovi to initiate arrangements with FGI to get the permission to use their receiver. Next Generation VLBI Project Since the termination of EVN-wide PCEVN project Metsähovi has continued the development of Metsähovi VSI Data System (MVDS). It is a simple and scalable multi-Gbps data acquisition system that adapts easily to the evolution of commodity PC technology. Using standard microcomputers and Linux has proven to be a very successful strategy. The performance of the system keeps improving when the computer industry develops faster computers. As predicted in the previous TOG station report, the nVidia nForce3 250 GB motherboards deliver sustained 512Mbps performance per PC in all directions (record, playback, disk-to-net, net-to-disk) with a 50--150Mbps safety margin. We expect the next notable improvement to be the arrival of the new nVidia nForce4 motherboards which have multiple native 1Gbps Ethernets and an increased number of native 3Gbps SATA-2 ports. Another interesting technology for VLBI is the Serial ATA port multiplier standard, which allows up to fifteen Serial ATA devices to be connected to one Serial ATA port. This greatly simplifies Serial ATA cabling and allows up to 60 hard disks (with today's 400GB disks that would equal 24TB or 40 thin tape capacity) to be connected to a standard microcomputer, if somebody wants to do that. Station Hardware/Software H-Masers and Maser Room Now that the temperature control system of our isolated maser "cave" seems to succeed in keeping the temperature constant within +/-0.5C, a new problem, water seepage thru concrete walls has appeared. At this moment we use two Russian Kvarz H-masers, with Kvarz-69 as the station frequency reference and Kvarz-70 as backup. They both have functioned very reliably throughout 2004. Receivers The 22 GHz receiver is functioning as before. The 43 GHz receiver is still being repaired for its broken RCP HEMT. A new pair of 43 GHz (LCP/RCP) HEMTs was purchased from the "YLINEN Electronics Ltd." in September 2003. The HEMTs have been installed but DC and bias wiring and testing is waiting for manpower and proposal pressure. The new dual-feed 80-115/150 GHz SIS receiver from Institute of Applied Physics of the Russian Academy of Sciences (IAP RAS, Nizhny Novgorod) delivered fringes for the first time in April 2004 in a Global mm-VLBI Array session. The 3mm part features dual polarization and 2mm single LCP polarization (same unit, frequency swap needs manual swapping of quasioptics, about one hour). All swaps between 1.3cm, 0.7cm, 13/3.6cm, and 3mm/2mm are manual and require one workday (Mon-Fri 7--13 UT) plus the cool-down period which is safest to count as overnight, meaning the receivers are available starting from UT 7 next workday. 13/3.6cm are available simultaneously (as in all geodetic S/X receivers). The VLBA data acquisition rack Several of the baseband converters had broken down during the last years, they had to be fixed because a total of fourteen baseband converters were needed for geo sessions. Unfortunately all of them had the same fault: The 10G070 prescaler in the oscillator divider was faulty. Three of the prescalers are totally dead, one divides erratically by 11 and two work only if the frequency is a multiple of 100kHz. For the repairs MRO obtained a EVN spare divider from Jodrell Bank. This divider was manufactured by Signatron and had a ecl-based prescaler. The Signatron divider had several interesting features. First interesting feature was that the divide-by-10/11 prescaler changes to divide-by-14 prescaler when frequency is over 990 MHz. Our oscillator could go to 1040 MHz, this caused the BBC to lock-up into the highest possible frequency. Another interesting thing was that the synthesizer did not lock in the frequency range 700-820 MHz. From the oscilloscope screen we could see that again the prescaler sometimes divided by 10 and sometimes by some other number. When we fed the divider box from a Rohde-Schwarz signal generator we could see that this behaviour was dependent on oscillator level. The normal 10dBm level specified in the VLBA BBC documents was not enough, the divider spent more and more time dividing right when the level was increased by a few decibels. VLBA tape recorder It has been known since 2001 that some high-numbered recorder tracks are producing marginal recordings. A Mk5A disk recorder is available, thanks to JIVE providing a loan unit. 512Mbps mode was tested in Nov-2004 in N04K2 but no performance testing at 1024Mbps has been done. Finnish Geodetic Institute is planning on purchasing a Mk5B unit sometime during 2005 (when and if it becomes available). A four-PC cluster of PCEVN units is available as well. The VLBA tape drive (with one headstack only, with upgraded Mk4 write electronics) is being decommissioned and should not be scheduled any more. VLBI Sessions in 2004 Metsähovi participated in Feb-2004 and Nov-2004 22 GHz EVN sessions. Geodetic test experiments for Finnish Geodetic Institute were performed in Mar-2004, May-2004, Sep-2004, and Nov-2004. Global mm-VLBI Array sessions were resumed with the new 3mm/2mm SIS receiver in Apr-2004 and Oct-2004. Manpower At the moment MRO has two effectively part-time persons working in VLBI: Ari Mujunen, whose main occupation is laboratory engineer jand Jouko Ritakari, who is working part-time at MRO. People from Finnish Geodetic Institute are helping in the geo sessions.