A rapidly developed control system for WITCH A

A rapidly developed control system for WITCH A. Lindroth, V. Kozlov

Contents • • • Requirements Design choices Architecture Status Future improvements: next shutdown

Requirements • • • Ready simultaneously with experiment!! Easy to develop further Control everything in the experimental set-up Scans & other things needed for trap experiments Cheap

Control System ”definition” Physical system Experimental hardware Experiment Electronics Computer control Control System

Order of decisions Buy this or that piece of experimental hardware Buy/design a Controller unit The physical side allways must come first flexibility needed System ready fast less flexibility possible, but one should not ”paint oneself into a corner”: commit only to few decisions System ready fast not complete functionality at first What type of connection between controller and PC? Include the relevant code in the Control System

Design choices: Software Commitments • Use Lab. View: Save time • Use DSC, not replaceble: Save time, NI is big this dependence is not likely to cause problems • Make some minimal architecture choices, that can be kept even when the system keeps developing, thus not constraining the possibilities – Communiction architecture: Queued TCP/IP – Possible to hook on new programs to communication architecture

Software architecture Ca. DAQ Gate Diagnostics GUI Analysis Traps Cycle GUI main Communication Scans GUI Ca. DAQ Heart Config-GUIs HV GUI Vacuum GUI Communication DSC Physical-Parameter namemap GPIB CAN RS-232 CYCLE SCANS ISOBUS ANALOG OUT SMC FILE INTERLOCKS GPIB Config CAN Config RS 232 Config LOOP? Config? DIO Config BI Process Computer Queue TCP/IP DSC Write Datasocket Variables which are used in the VIs, and loaded from files at start-up. Or simply files.

Design Choices: Computer Hardware • One powerful I/O PC (but making sure the software allows addition of further I/O PCs) – Passive backplane: many PCI and ISA slots – GPIB, 18 ch RS 232, CAN, DIO, AIO, ISOBUS, PPG 100 all on one PC – Many National Instruments cards GPIB Card USB--> MIO 24 DIO 96 PPG 100 RS-232 box Card Cadaq Heart CAN Card 1 RU Screw panel Long cables (~15 m)To devices Screw Panel 2 RU LEMO panel

Design Choices: Controllers/Electronics • Many controllers are built by our electronics department: cheaper, we get exactly what we want, they can repair/upgrade it, some things are not available commersially, we decide how to pack different functionality into separate modules, . . . DELAY BOX • 10 ns jitter • 100 ns precision • 0 ns to ~20 s (+50 ns allways) • 20 ch • 2 outputs/ch: delay/gate • rs-232

Design Choices: Network • Our own, independent network, with a gateway to the ISOLDE network. • The gateway will allow control of WITCH via the Web. We are now learning how to get the appropriate permissions from the CERN Network Security guys. • An ISOLDE network PC for ISOLDE control. To ISOLDE network ”Cadaqheart” ”Cadaqgate” TCP/IP ”Cadaqheart” In light building Repeater? In ISOLDE hall

Recent developments: Vacuum system Valve Control box RS 232 Gauge Controller Relay Vacuum Interlock box Valve DIO card on Ca. DAQheart

Recent developments: Trap Electronics Temporary solution! COMPONENTS • PPG 100 • Delay box Delay Box • 52 ch static voltage supply/switcher +/-20 V • 24 ch rf splitter+static offset box Switchable Static Voltage supply rf func. gen Rf splitter, inverter, offset adder

Recent Developments: Software Status Ca. DAQ Gate Diagnostics GUI Traps Cycle GUI Communication Vacuum GUI Ca. DAQ Heart Communication DSC Physical-Parameter namemap GPIB RS-232 CYCLE SMC GPIB Config RS 232 Config LOOP? Config? DIO Config

Outlook Near-future work Long-term work (starting in 6 -12 months) • Fill the software functionality gaps • Readback of electrode voltages: fieldbus • Secure quality of code • Major upgrade of software functionality – where needs are found • Software Documentation • Implement capacity enhancements based on test experiments this summer • Upgrade to Lab. View 6. 1 • Software control of Voltages • Get new, computer controllable trap electronics • Get, install detector electronics • Change so all PCs are of in-rack type • Focus on DAQ: separate computer with (highly) optimized readout speed