Status of Kappa Goniometer on PROXIMA 1 B

Status of Kappa Goniometer on PROXIMA 1 B. Guimaraes, P. Legrand, P. Gourhant and A. Thompson Report for Kappa workshop, 28 th - 29 th November 2011.

Kappa Gonio of PX 1 • Large α (50˚), hence large range of crystal orientations. • Ω or Φ scans, hence large range of geometries possible. • But anti-collision is harder. We use « simple » approach (limit movements) ! • Closer to 20 μ S-O-C than to 1 μ (although < 7 μ part of angular range). In most cases, this has not been a handicap.

News from SOLEIL – PX 1 • Goniometer used routinely for 4 years now – collection around ω or Φ with κ non zero (≈ 10%), « normal » collection (≈ 90%). • 5 user groups request, at proposal, « oriented » data collection (large cell, anisotropic diffraction, difficult phasing from « big » object. They are « autonomous » …. • « Most » users have used κ non zero (help from local contact - exceptions « industrials » ) • Major upgrade to beamline in July – now have PILATUS detector and shutterless data collection. • Hence gonio motor control / data collect soft changed from « proprietary » (Rigaku) to « in house » . • Additional anti collision hardware added.

PART 2 – MORE DETAILS (AS REQUIRED) ON PX 1.

Mechanical Integration of Pilatus 6 M on PROXIMA 1 – S. Le. Couster. Mechanical design : Integration of the detector into the existing experimental environment: Modification of existing detector support with minimal changes Requirements for translating / offsetting the detector and their impact of design Counter weighting scheme to avoid vibration / oscillation during detector movements Protection of the active detector surface (Guillotine): Positional and translational constraints, limited weight supported by detector, detector sensitive to vibration Double function of protection and support for X ray imaging caméra (beam alignment) I 45° degree X-ray images: Based on in-house design (detector group) with higher magnification, stabile support and mounting on guillotine Beamstop integrated in detecteur Space limited by guillotine – proximity to detector surface, motorised alignment Detector connection: Movable cable tray for coolant and connection to detector needs to permit > 1. 25 m of translation 3 D models to avoid collisions: - Goniometre/Guillotine Lighting/Guillotine Robot/Detecteur

Electronic around the Experimental station New motion control for Goniometer • Specific trigger system which manages all possibilities of the PILATUS mode • New anti-collision system to secure Pilatus from Goniometer movements • New motion control Control. Box Driver. Box Servo. Box Experimental station PLC Pilatus Goniometer Anti-Collision System SPIBoard Specific Trigger System

Trigger PILATUS Using a SPIETBOX to trig the Pilatus allow us to select easily trigger modes : • - External Trigger depending on Time - External Trigger depending on goniometer position - External Enable adjustable in time - External Enable depending on goniometer position All parameters needed to trig the Pilatus are transmitted from tango device to the SPIETBOX through the SPICONTROLLER board. • Two other SPIETBOX are used to duplicate encoders signals and so isolated the motion control from the Pilatus trigger control. •

New motion control: Servo. Box New standard power unit for servo motor at SOLEIL (direct current and brushless motor). Servo. Box contains : • Standardized connectors • Power supply: 24 VDC (logic) et 48 VDC (1 KW to motor) • Fans and back plane board • 4 slots for 4 power board (see below) Directly compatible with Control. Box Preliminary study and specification written by Soleil Power board SDB 10 based on an Elmo Whistle OEM component Numeric amplifier for DC or brushless motor • Rated output current: 10 A (Voltage 48 V) • Feed back by resolver or incremental encoder • Signals managed: limit switch, thermal sensor, brake • • Entire design by Soleil (electronics and mechanics)

Anti-collision The Anti-collision system is made using position sensors of the three concerned motors. • Each sensor is connected to a PLC. The anti-collision equations is embedded in the PLC which determines if motors must be stop or not. • When specific conditions occurred (see table below), the PLC send to the Control. Box the order to stop the motors and the conditions that activated the security. • Until the operator do an action to unlock motion, all motors are disabled. After an acknowledge, the Control. Box allow movement only on the direction that will correct the default. • When we are far from the collision area, movements are allowed in both directions. • Still in progress … •

PROXIMA 1 data collect with Pilatus detector MXCube PX 1 Collect Manager PX 1 Logging Manager Process & shutter control Experimenta l data Filer ramdisk SPIETBox encoder positions, shutter & detector states, triggers to detector logging Goniometer axis control Detector control Meta data Filer Beam intensity & X, Y positions logging XBPM control SAC - 24 novembre 2011 - L. Roussier & P. Gourhant

Process control implementation TANGO devices PX 1 Collect. Manager & PX 1 Logging. Manager written in Python by L. Roussier & P. Gourhant Enabled easy testing « off line » (hardware + motor test bench ) • quick and easy adaptation for beamline requirements & MXCube interfacing • Shutterless (continuous) collect Hardware synchronisation between axis position and shutter opening /closing 4 Pilatus trigger modes implemented (external trigger or gates) Mode currently used (via SPIETBox) : open the shutter, give an external trigger for the first image • count acquired images from the detector, close the shutter when all are acquired • Experimental data directly saved on ramdisk by the detector software 10/28/2020 SAC - 24 novembre 2011 - L. Roussier & P. Gourhant

Meta data logging along the collect Objectives : - Collect system characterization and improvements in terms of synchronization, reproducibility, axis velocity. - Preventive degradation of equipment. - Future correlation with diffraction images Implementation : - Flyscan mechanism - Frequency : 1 k. Hz - In parallel : Periodic beamline sensors archiving First results of analysis (Python scripts) - Parameters optimisation of the collect ( Delay before first image for constant speed) - Omega behaviour better than Phi (speed regularity, …) 10/28/2020 SAC - 24 novembre 2011 - L. Roussier & P. Gourhant

Meta data first analysis Collect axis : Omega Have improved PID, acceleration time for omega – phi axis is much worse by comparison : we have not yet got the same level of control / following error as the Rigaku hardware achieved. Data quality is good, but not quite as good as using omega – improvements needed. 10/28/2020 SAC - 24 novembre 2011 - L. Roussier & P. Gourhant

• • ADSC 315 r Max 3 images/sec. Fastest on PX 1 is 12 images/min. Binary images of 18 Mb. Up to 180 Gb per day (uncompressed). PILATUS 6 M Maximum speed of 12 images/s "cbf" format (binary 32 bit compressed images. CIF) 6 -7 Mb/image. Collect of 3 Gb in 1 or 2 min in continuous mode (85 Mb/sec), every 5 to 10 min – gives max of 800 Gb per day. We typically collect 400 – 500 Gb per day (compressed).

Staff directly involved on PX 1 – PX 2. • PROXIMA 1 – R. Barbosa, P. Gourhant, B. Guimaraes, P. Legrand, F. Blache and S. Pierre-Joseph • PROXIMA 2 – W. Shepard, D. Duran, G. Fox, T. Moreno, A. Buteau Collaborations – Global Phasing, Insititut Pasteur (F. Rey), CNRS – Marseille (S. Spinelli), CNRS Gif sur Yvette (S. Fieulane)

PART 3 – BIOSTRUCT-X

Bio. STRUCT-X – What we proposed Sept 2011 – Sept 2015. • 2 Workshops for Scientific / Technical Development. • 2 User training courses (crystallography? ) • Accessibility at each SR site • Validation

Validation • High α vs low α. Do we have enough knowledge to advise people? • Do our measured performances mean the same things? What are the best characterisation methods? • Precision of alignment methods and impact on experiments? • What protocols are we sure of, and how do we choose between them (automatic) / when do we ask for user input.

Actions / Discussion. • Who does not want to participate (no funding for non EU of course)? • Review of available beamline facilities – underway : I will circulate a version around Christmas for you to verify and complete. • Plan for first full meeting. Subject, when (before Oct 2012) and where. • Think about validation (not comparative) criteria (technical and crystallographic) / measurement methods.

What we need to do better at SOLEIL • Anti-collision. In some orientations can use κ of 90˚, but we don’t to make collisions impossible. • Radius of S-O-C. Can get the gonio to work fine (≈7 μm) but only in hemisphere « towards source » . The moment ω « goes over to the dark side » , centre jumps 20 μm! Solution is to repair it, but we also need to better characterise (automatically? ). • Between first two, only us 25% of capability of our gonio. • Properly understand the best way to use the goniometer, to advise users correctly. • Train the 90% of users that are « not autonomous » !

Difficult structure solution - 1. • Contacted SOLEIL with inhouse data he couldn’t index. • Came with unfrozen crystals. • Collected native data 1. 9 Ǻ resolution, with a maximum cell edge of 500 Ǻ. • Gd derivative obtained on site – structure solved by SAD (in refinement).

Difficult structure solution - 2 • Problem was to get phases (quickly!) as MR solution incomplete. • Tricky data collection because of diffraction “spot shape” (Ho derivative). • Os derivatives – site had high B factor and gave phases to low resolution. κ geometry critical. • Collaboration with I. P and Global Phasing.