5/25/2016J-PARC1 High Level Physics Applications – The Big Picture Day 1: Relation of High Level Applications to Other Systems

June 16-27, 2008USPAS2 Accelerator View: Satellite It can be big!

Accelerator View: Engineer Engineer’s view: lot’s of real stuff in it that requires interfaces and meets design requirements

Accelerator View: Theoretical Physicist  An abstract object that performs manipulations on ideal beams under ideal conditions

Accelerator: Operations View  A control room with interfaces to whatever is required for seamless accelerator operation.

Accelerator: Manager’s View  A device that meets the promises made to the funding sponsor

High Level Control Room Physics Applications  Need to consider the many points of view of the disparate systems that an accelerator is composed of: Magnets RF Diagnostics Timing Control system  All of these are input/output to the high level applications  There can be many devices Effect on the beam is coupled Not feasible to control each device individually Need a physical model to help guide the setting of these

Relation of High Level Applications to the Control System and the Hardware  The high level applications communicate with hardware through a control system  It’s important to carefully spell out the requirements / interfaces for the control system and the underlying hardware. Control System

EPICS View of the Data Flow  Here XAL would be a client  Most “MEDM” screens are preconfigured interfaces to hardware MEDM Client MEDM Server IOC MeterPower SupplyCamera IOC Channel Access

High Level (physics) Applications  High level applications typically integrate information from multiple systems over an extent of the accelerator  These applications often require information about the beam (from beam diagnostics) and perform manipulations on the beam (magnets / RF)  They communicate with these systems through a control system  Often a physics based model is used to provide information how to control the beam  Examples: Orbit display and correction, orbit bumps, setting RF phase and amplitude, …

Specify the Interfaces, and all should work  In principle things are simple:  Specify what you want to receive from measurement devices, e.g. a measure of the horizontal beam position at some place and time and return its value in mm, with name “xyz”  Specify what you’d like to control E.g. a magnet field level, in Tesla, for magnet xyz.  Wire it all up and hit the “On” button

What Could Go Wrong???  Since high level applications use information measured from many sources, it is important to understand the strengths and limitations of these sources  It is also important to understand the limitations of equipment you may try and control  There are inherent difficulties in trying to communicate quickly with many information sources – control systems are complicated and not consistently reliable.  Often the majority of a high level application is exception handling – checking for the things you are aware of that can go wrong.  Let’s examine some of the primary players in accelerator measurement and control

Magnet Controllers  Typically engineers provide a current or voltage setpoint and readback Physicist need magnetic field  Power supply can be tripped (setpoint is OK, but readback may be bad)  Iron core magnets take time to settle – need to wait after setting them before measuring the effect on t the beam Hysteresis affects the field level for a given current setting

Diagnostics - Eyes and Ears to the Beam  BPM – Beam Position Monitor measure beam position  BCM – Beam Current Monitor (current transformer)  BLM - Beam Loss Monitor Ionization chamber (IC) Photomultiplier tube (PT)  Profile measurement Wires Fluorescence screens Residual gas stripping Lasers

Beam Position Monitors  Moving charge induces current in surrounding electrodes Zo beam electrode

BPM  BPM signals: Amplitude = sum of top + bottom +left + right Vertical position = (top-bottom)/amplitude Horizontal position = (left – right)/amplitude Zo R I beam Zo Zo Reflection R I wall

Beam Current Monitor (BCM)  is proportional to I I out is 1/N times I in v R Varying Current Magnetic Field Toroidal Core N turns  A changing current induces a voltage on a current loop surrounding it:

Typical Analog Front End & Digitizer Block Diagram  There is typically a fair amount of information processing in the analog / digital conversion  Often there are gain settings (saturation effects) -40 dB AD602A or AD dB to +30dB FILTER -34dB to +26dB ADC 40MHz or 60MHz Clock (continuous) DAC 8 Bits 14 Bits 8 Bits GAIN DATA 78 Ohms SWITCHED ATTENUATOR 0dB,-20dB,-40dB 6mV to 23.4 V AD Bit +/- 1.1V 1.05MHz rf REG. AD8131 Diff. Out dB DIGITAL INTERFACE DIGITIZER FRONT END Dual-amps switched Duplicated for switched amps 7MHz Gaussian LP Filter

Example Digital Interface  Plenty of room for problems in the digitizer to control system interface as well  Timing triggers are always an issue 1.05MHz (   f) RING Ref. 1.05MHz RING Ref PLL and TIMING 67MHz Clock 14 Bits CIRCULAR DATA MEMORY FIFO 8 Bits GAIN & SETTINGS MEMORY Gate Control 1mS GATE PC INTERFACE EVENT TRIGGER LOCAL COMPUTER DIGITIZER DATA Data Timing Out Data Timing In TIMING Timing DATA

BCMs  BCMs are quite simple looking

Controllers for diagnostics are located in racks in service buildings  Sometimes someone brushing against a cable can mess up things  They have their own expert system interfaces (be careful that the proper values are restored after re-boots!)

RF Systems  Use Oscillating Electric Fields to Accelerate a Charged Particle High Voltage Convertor Modulators High Power RF Low Level RF Electric Force in Cavity Time accelerated decelerated sec

RF System Components Klystrons Linac LLRF Transmitter Ring Cavity Klystrons Reference Line System Ring LLRF Klystrons

Typical LLRF Control Rack Typical LLRF control rack installation in the superconducting linac. The VXI crate contains: Input/Output Controller: PowerPC running VxWorks Utility Module: Decodes events from Real Time Data Link (Global Timing System) Timing Module: Generates RF Gate timing signal Two FCM/HPM pairs – generates the patterns for the high power RF, includes feedback, feed-forward, interfaces to the control system, etc.

B lock Diagram of the SNS Linac LLRF Control System

Field Control Module (FCM)  Regulates frequency control, field amplitude and phase regulation  Corrects for correlated and uncorrelated errors  Adaptive feed-forward control used for correction of repetitive field errors caused by beam loading and Lorentz force detuning  RF Sequencer is used for normal operations – startup, warming up cavities, tuning cavities, etc.  Primary interface to the Control system, e.g. what phase and ampltude does the user want the cavity to run at.

High-power Protection Module (HPM)  Responsible for protection of the RF systems Monitors up to 7 RF inputs  Detects cavity and klystron arc faults utilizing the AFT FOARC chassis  Vacuum interface is connected to LLRF via HPM  Soft interlocks from Cryo, Water, HPRF and Coupler Cooling  Chatter faults are expert settable to fine tune individual channels  History plots are available to monitor any two input channels

Timing Systems (Master)  Master Timing system generates event signals and propagates these throughout the accelerator to be used as triggers for systems Pulsed magnets Diagnostics RF …

Timing Systems (Clients)  A difficulty is that different systems receive separate triggers and handle these triggers differently

Timing Signals from Multiple Sources  Providing common units and ways of providing information amongst disparate groups is important

High Level Applications – Systems Interfaces  The interface between high level applications and the different machine systems is through the control system  The high level application developers should request the information they would like to receive from the system experts Measureable/settable entity, units, context, Good communication between experts is important – needs to be initiated as early as possible Access to raw data is valuable for verification

Putting it All Together – Control Room Software Applications  We want to know what the beam is doing where and when  We want to control the magnets and RF with “physically meaningful units” relative to the beam  We want to adjust the controllable parameters in a predictable manner  This is what High Level Physics Applications tend to do  Lots of things can go wrong: “trust but verify” all information that you use. The beam never lies!!!