REX ISOLDE and offline controls Mats Lindroos on

REX, ISOLDE and off-line controls Mats Lindroos on behalf of the ISOLDE team Mats Lindroos

Outline • • What at ISOLDE is CO responsible for? What has CO done and completed? What is started and in need of completion? What can be left for later? Mats Lindroos

CO at ISOLDE • CO responsibilities: – Controls for targets, Front-ends, separators and experimental area • Both low-level controls and application programs – Controls for REX-ISOLDE • Both low-level controls and application programs – Low level controls for off-line separator and pump stands Mats Lindroos

ISOLDE@CERN • 400 Users at 30 experiments each year • User operated facility Mats Lindroos

CO history at ISOLDE • The ISOLDE control system: – Produced by PS division in the 90’s – Windows based – Parts migrated to PS standards in 2001 -2003 • The ISOLDE controls September 2005: – ISOLDE is left with a mixed Windows/LINUX/Lynx control system • Major problem for maintenance, operationally required modifications and developments Mats Lindroos

The “new” ISOLDE controls Mats Lindroos

ISOLDE targets, Front-ends, separators and experimental area • Migrated to PS standards: – PLC slow control, VME FEs and JAVA applications – Well executed project with innovative solutions – Remaining problems: • Very poor response time measured from CCV to AQN • Poor stability of JAVA applications and support for JAVA programmers (in OP group) • Logging programs highly unreliable • Missing applications (instrumentation) Mats Lindroos

Off-line and pump stands • PLC low level controls provided by CO group – Application programs written by ATB-IF section student in CO provided EXCEL environment – Problems: • New class A lab will need remote control for some equipment Mats Lindroos

REX-ISOLDE • Migration of REX controls started in 2004 -2005 – Re-structuring of PO side including hardware modifications – Test with control of some elements – Control of RF amplifiers implemented • Problems: – No programmers available for application programs (OP) – REX-EBIS and REX-TRAP, no migration started – ISOLDE collaboration contributes with salary for one CO engineer • Agreement in memo from 2003: Complete migration and maintain existing system Mats Lindroos

New projects • RFQ cooler – Control hardware selected in collaboration with CO group – Specification note for integration in ISOLDE control published • HIE-ISOLDE – Proposal for the future of ISOLDE to be presented at the Nu. PAC meeting 10 -12 October 2005 Mats Lindroos

Summary: To be completed ISOLDE: • Improve response time of ISOLDE controls – • • • Objective: Match performance of previous control system Provide general support for JAVA programs (and programmers) to create a stable control environment Advice on PLC low-level controls for Class A lab Name link person for ISOLDE in CO group REX-ISOLDE: • Coordinate migration of Front-ends with available resources in the department for application programs • Complete migration of REX power supplies controls (with PO) • Complete migration of REX EBIS controls (with ABP) • Maintain and support remaining windows based controls – Transfer of responsibility from “student” to CO for REX-TRAP Lab. View control system New projects: • Integrate the standard parts (power supplies) of RFQ cooler controls in the ISOLDE control system (with ATB) General: • We need clear CO guidelines for REX extensions and other HIE-ISOLDE projects • We need a “synchronization” between changes in CO standards with available resources for application programming (OP and specialists). This includes modifications of FE software which requires re-compilation and/or modifications of software on consoles. Mats Lindroos

Summary: To be left for later • REX-TRAP migration • Non standard equipment for new projects – Dedicated control under specialist responsibility – The CO PLC hardware can be used with the Siemens built in software and Labview • FETCH/WRITE interface of a Siemens Ethernet Communication Processor (like CP 343 -1) • Already implemented for ISOLDE RFQ cooler • Could be used to control new off-line separator in the Class A lab (works over ethernet). Mats Lindroos

6 comments from the ISOLDE team concerning this review “My general feeling is that we have bottom priority and that there is no-one in the CO group who feels responsible for ISOLDE” “The PLC system was forced on the specialist groups. Now we are supposed to support this systems without sufficient CO support and documentation” “Key application requested by the Safety Commission are not “stabilized” resulting in incomplete logs and poor records of the number of protons taken on the targets. This is in direct contradiction to commitments made to the host states” “The application programs provided by the OP group are very good but we lack a central support in the CO group for this environments. The operators do their best to help but as they are on shift work they cannot make interventions, nor address general problems with the JAVA control environment” “The detail of requested specifications does not correspond to the support eventually received. After having spent months of specifying a program I got something back that wasn’t even de-bugged” “The control group seem to be detached from the real world in which their control system is supposed to work. They are only interested in making one small process run but not in ensuring that the full system really does what it should do” Mats Lindroos

Status ISOLDE September 2005 (1) • • Non-application-specific Problems – "Standard" controls (not Isolde specific) – – – – • Intermittent slow down or freeze, affecting entire control system Intermittent slow down or freeze, affecting Java controls only Working sets Equipment lists recently updated. OK. Knobs (old) OK Knobs (new) OK SEM-grid OK Eqp-survey Unsupported and no replacement available Timing survey (unused? ) Radiation monitors OK Isolde controls – – – – – Rilis Laser Imager OK Display Alarms Tolerances and error conditions not set correctly for all equipment Elogbook Intermittent slow-downs or freezes Vistar infos (GPS+HRS) Software in need of improvement/update Easy Eqp Access Not working QP Array OK Logger Intermittent stops - under investigation Synoptics (from CO) Out of date Synoptics (from OP) Being evaluated Mats Lindroos

Status ISOLDE September 2005 (2) • GPS+HRS controls – – – – • Controls not yet migrated to Linux – – – • Wiregrids Moving-needle scanners Fixed-needle scanner Other missing things – – • Target heater OK Tape station Some minor bugs remain (graph plotting) HT control OK Mass scan OK Gas mixing OK GHM/GLM deflectors Some cosmetic changes needed Proton integrator Some changes being made Calibration of the gas-system pressure gaugues (Volts -> mbar) Calibration of thermocouples (m. Volts -> degrees-C) Stepper motors: correct decoding of position encoders (ATB for hardware) Stepper motors: interlocks to avoid collisions in GPS switchyard Stepper motors: set the Z-offset from a console Slit control application (ATB for hardware) FAQ – How to interface experimentalists' data acquisition systems with the ISOLDE control system? Mats Lindroos

Conclusions It is time consuming to maintain legacy hardware and software… …but it is even more time consuming to only partly complete a migration to new standards. The CO group lose major resources on “fire brigade” actions at ISOLDE. Please, use them to complete the started jobs instead. Mats Lindroos