From: F. Bertoldi (MPIfR), 24 June 2002 To: W. Brunswig (IRAM) CC: IRAM Granada, Bolometer Group MPIfR, Thum, Grewing Re: Bolometer systems at the 30m: issues, problems, plans. Dear Walter, To follow up our conversation from a few weeks ago, I would like to give you a written summary of some issues relating to the bolometer systems at the 30m. We would much welcome feedback. greetings, frank 1 We intend to return MAMBO-117 (also called MAMBO2) to PV in late October [FB 21.8.: early November] with improved electronics and other modifications. MAMBO-117 will then become the standard, and MAMBO-37 the backup bolometer system. 2 MAMBO-37 will stand by at liquid nitrogen temperature throughout the summer. The summer bolometer observation pool will likely be mid October to mid November [FB 21.8.: 27.9.-12.11]. Soon after the winter pool will begin. 3 Next year we shall try to redesign the optics system, to also provide for a second bolometer that could be operational at the same time as MAMBO-117. The idea is that we would have another, smaller system to operate at 0.87 mm, which could be switched to quickly during pooled observations whenever weather permits. This modification would mean to also redesigning the M5-M6 structure, since the best location of the second cryostat would be approximately where the ABBA rack is now. We would need to compact the ABBA rack to move it further back somewhere below M4. 4 HUMBA (the 2mm array) will be returned to Bonn along with MAMBO-117 in early July (pickup by Reichertz & Racanelli) [FB 21.8.: pickup postponed to October/November]. It will be returned to PV after modifications for further tests, and hopefully normal operation, likely next spring. HUMBA uses the ABBA backend. 5 The current windows-based MAMBO-backend ABBA has recently shown to be vulnerable to virus attacs of unknown origin. A proper installation with adequate security measures will hopefully make the system more reliable. We hope to provide a linux-based BE (which we are now designing for LABOCA, the APEX bolometer array) sometime next year to replace the windows ABBA. We shall keep you informed on developments related to the LABOCA-BE and in the course discuss the redesign of the MAMBO-BE with you. 6 We must provide an ABBA backup system in case of hardware failure. One option is to mirror the hard disk. Additionally, a second PC may be configured just like the ABBA PC --although without the ADC card. (It would be too expensive to buy a backup ADC card. A spare is in Bonn that could be shipped in a day.) 7 The bolometer system at PV has reached a considerable level of complexity, and will likely become more so in the future. Modifications to the system are done frequently by a number of people both from Bonn and IRAM. Although some modifications and problems are now being logged in the current.comments file of the flexible pool web page, is is prudent that a complete documentation of all changes to the system is accessible to all persons involved, especially to those observing. We suggest that you install such a system --e.g. a protected web page, or a CVS depository-- so that all modifications to the hardware, backend, or the merger/dataflow can be logged there. The operators must be current on the state of the system, how to trace errors, and they shall provide all the necessary passwords. 8 We had installed and tested observing modes with continuous data sampling throughout a scan. However, such observing modes are currently not possible because DAPs are not available between subscans. We stress the importance that a solution be found to this problem during this summer, so that the continuous sampling of data can be tested in October. Continuous, fast data sampling promises to improve the efficiency especially of on-off observations. (We should also record the signal from the second wobbler encoder). 9 The fits merger will have to be modified when continuous DAPs are provided. After discussions with the Grenoble software people we suggest that the current fits format be maintained for the next winter season. Although Albrecht favours a simpler, unsynchronized mbfits format (which would eliminate many sources of errors), we do not have the personnel resources to modify the reduction software in time to accomodate the new format. The new data reduction packages in planning shall accomodate the new format, but they are unlikely to be ready for next winter. 10 We would like to branch off signals from an AC and a total power bolometer channel to monitor the sky opacity and noise along the line of sight. You have suggested that you might be able to provide ADC-equipped VME systems in the receiver cabin to transmit such signals (as well as air temperature etc) to the LAN, where they could be picked up for processing and display, e.g. using the chart-recorder software Albrecht experimented with last year. I would like to provide the observer with a continuous display of the sky opacity measured by the tau-meter, the line of sight opacity computed from the total power bolometer signal, and the skynoise level computed from the AC or/and total power bolometer signal. This information would be extremely valuable to help guide the observations. 11 Axel Weiss will continue to expand the capabilities of the php/MySQL pool observation management system, including public and restricted access to the database. To provide users with the data we need a restricted ftp area from where much of the bolometer data from a given season can be accessed. 12 We need to explore ways to record bolometer data without using the wobbler, in order to take absorber measurments at any time to check the state of the system. Currently we have no simulated wobbler signal for the backend. The fast scanning or skydip modes of ABBA may do the job. ---------------------------------------------------------------------- Frank Bertoldi Max-Planck-Institut fuer Radioastronomie phone: 49 - (0)228 - 525 377 Auf dem Huegel 69, D-53121 Bonn fax: 525 229 Bertoldi@MPIfR-Bonn.MPG.de