LCLS Undulator Tuning And Fiducialization Zack Wolf Yurii

LCLS Undulator Tuning And Fiducialization Zack Wolf, Yurii Levashov, Achim Weidemann, Seva Kaplounenko, Scott Jansson, Ralph Colon, Dave Jensen August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 1 Zachary Wolf wolf@slac. stanford. edu

Introduction LCLS: Linac Coherent Light Source (Free Electron Laser) The LCLS will consist of 33 undulator segments. The last 1/3 of the linac is used to produce the electron beam. Some Parameters: Ebeam = 13. 64 Gev λr = 1. 5 Å Planar permanent magnet undulators Nd Fe B permanent magnets λu = 30 mm Gap = 6. 8 mm Tapered gap, 4. 5 mrad Bfirst = 1. 249 T, tapered, nominal field, first undulator Kfirst = 3. 5, tapered, each undulator has its own K 226 poles per segment Each segment is 3. 4 m long August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 2 Zachary Wolf wolf@slac. stanford. edu

LCLS Location J. Galayda August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 3 Zachary Wolf wolf@slac. stanford. edu

LCLS Installation and Commissioning Time-Line Paul Emma June 29, 2007 linac/BC 2 Install LTU/Und/Dump FEE/NEH Install First Spont. X-Ray tunnel/FEH Light Install CD-4 a now PPS J A S ON D J F MAM J J A S ON D 2007 2008 2009 Controls checkout & re-commission injector …Injector Commissioning Re-commission Inj/BC 2 to SL 2 Linac/BC 2 LTU/Und/FEE FEL Commissioning August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 4 Zachary Wolf wolf@slac. stanford. edu

~18 m Magnetic Measurement Facility ~23 m Floor plan divided into three functional areas Magnetic Measurements (± 0. 1º C) Fiducialization and Assembly (± 1º C) Storage (± 2. 5º C) Test stand lay-out is driven by requirement to match the Earth Magnetic Field conditions in lab to Undulator Hall, i. e. azimuth and gap orientation need to be identical LCLS-TN-04 -1 August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 5 Zachary Wolf wolf@slac. stanford. edu

Undulator Movements In MMF Tuning Time: about 4 weeks 1 Week, Fine Tuning 1 Week, Rough Tuning 1 Week, Come to 20 C <1 Week, Finish Fiducialization Peak Throughput: 1 undulator per week August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 6 Zachary Wolf wolf@slac. stanford. edu

MMF Temperature Achim Weidemann In general, the system meets stability requirements Time history at the Kugler bench August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 7 Zachary Wolf wolf@slac. stanford. edu

Rough Tuning Bench Components August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 8 Zachary Wolf wolf@slac. stanford. edu

Rough Tuning Bench Reference Pole Undulator Cam Movers Zero Gauss Chamber Cable Handling Bench August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 9 Zachary Wolf wolf@slac. stanford. edu

Capacitive Sensor Alignment Performance Undulator mechanically aligned to bench in 15 minutes Yurii Levashov August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 10 Zachary Wolf wolf@slac. stanford. edu

Scans Begin And End In A Zero Gauss Chamber Beginning zero field measurement Ending zero field measurement Amuneal Chambers Sentron Hall probe Zero Offset Correction Assume linear offset dependence on time August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 11 Zachary Wolf wolf@slac. stanford. edu

Undulator Tuning LCLS-TN-06 -17 • Gives requirements • Describes equipment • Enumerates each step of tuning • Describes output data Rough Tuning Steps: straighten trajectories set gap (set K and tune phase) do phase shimming do phase matching add magnetic shield Fine Tuning Steps: check trajectories, phase adjust field integrals shim field uniformity glue shims in place perform checks make final data set fiducialize August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 12 Zachary Wolf wolf@slac. stanford. edu

X Trajectory Shims Isaac Vasserman’s Ideas ANL/APS/TB-48 § Apply shims to top and bottom poles § Shims weaken By, don’t cause Bx § Can’t strengthen pole, instead place shims on next pole to reduce deflection in other direction § Developed software to automate shim placement LCLS-TN-04 -7 August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 13 Zachary Wolf wolf@slac. stanford. edu

Y Trajectory Shims Bx Shim New Design Smaller external fields Uniform Bx field in gap Old Shim Design Inconel Steel Spot weld Yurii Levashov August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 14 Zachary Wolf wolf@slac. stanford. edu

Straighten X And Y Trajectories Measured X Trajectory Calculate Shims Modeled X Traj With Calc Shims Measured X Traj After Calc Shims Applied The Shimming Is Automated August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 15 Zachary Wolf wolf@slac. stanford. edu

Gap Shims Tapered shims give the pole cant angle and are used to set the gap. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 16 Zachary Wolf wolf@slac. stanford. edu

Phase Shims LCLS-TN-04 -7 ANL/APS/TB-48 § Measure By with a Hall probe § Calculate phase error § Calculate shims to correct error § Apply shims, repeat § We don’t have shims to strengthen a magnet, only weaken § Must locally reduce the gap to increase phase shift § Developed software to automate shim placement August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 17 Zachary Wolf wolf@slac. stanford. edu

Fine Tuning Bench Components August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 18 Zachary Wolf wolf@slac. stanford. edu

Fine Tuning Bench Undulator With Magnetic Shield Cable Handling Cam Movers Kugler Bench August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization Steel Girder And Stands 19 Zachary Wolf wolf@slac. stanford. edu

Kugler Bench Performance Measurements by Georg Gassner X Y straightness spec: < 14 μm Pitch and yaw spec: < 4. 2 μrad Specs met in long block region August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 20 Zachary Wolf wolf@slac. stanford. edu

Coils For Fine Tuning Bench Bx, By One Undulator Period Long Coils Short Coil Zero Gauss Chamber I 1 I 2 Dave Jensen Field Integral Coil August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 21 Zachary Wolf wolf@slac. stanford. edu

Hall Probe Position Problem: The vertical trajectories on the Dover bench and Kugler bench did not agree. The horizontal field integrals from the Hall probe had a large y dependence. Coil measurements showed no such dependence. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization We stumbled onto a paper Isaac Vasserman wrote in 1997! The Sentron probes are sensitive to planar Hall effects after all! 22 Zachary Wolf wolf@slac. stanford. edu

Solution To The Hall Probe Problem Isaac’s hypothesis: planar Hall effect (30 microns Dover bench, 10 microns Kugler bench) Solution: Measure the positions of the Hall elements. Tune the horizontal trajectory with the By probe on the midplane. Tune the vertical trajectory with the Bx probe on the midplane. Use only the long coil for final field integral measurements. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 23 Zachary Wolf wolf@slac. stanford. edu

Effect Of Hall Probe Error On Trajectories μm X Trajectory μm 2 12 -2 Y Trajectory -2 y=+200 μm 6 y=+100 μm -1 2 y=0 μm -2 1 y=-100 μm -6 2 y=-200 μm -14 August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 24 Zachary Wolf wolf@slac. stanford. edu

μm Constant Bx 0 Corrects Bx To Long Coil Field Integral Y Trajectory 12 -2 μm Corrected Y Trajectory 2 y=+200 μm Bx 0=+0. 52 G -2 6 y=+100 μm Bx 0=+0. 29 G -1 2 y=0 μm Bx 0=0. 00 G -2 1 y=-100 μm Bx 0=-0. 29 G -6 2 y=-200 μm Bx 0=-0. 63 G -14 August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 25 Zachary Wolf wolf@slac. stanford. edu

Hall Probe Calibration Water NMR Chiller For probe temperature Magnet HP 3458 Metrolab NMR Seva Kaplounenko August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 26 Zachary Wolf wolf@slac. stanford. edu

Hall Probe Calibration Problem Probe temperature? Poor fits? Probe offset? Calibration magnet non-uniformity? Impurities in copper probe chiller? Calibration Change Before and after first undulator August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 27 Zachary Wolf wolf@slac. stanford. edu

Calibration Magnet 0. 25 G Very uniform! 2. 2 G 0. 2 G August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 28 Zachary Wolf wolf@slac. stanford. edu

Hall Probe Calibration Study Problem G-10 Temperature, ok Fits, ok Calibration magnet uniform Copper non-magnetic Offsets change, why? Stress on probe August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization G-10, permanent mount Copper block 29 Zachary Wolf wolf@slac. stanford. edu

Hall Probe Calibrations In Fixed Mount Only 03 -16 -07 and 04 -03 -07 are used so far. The Hall probe is calibrated in the same mount it is used in. Starting measurements in reference magnet before set K August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 30 Zachary Wolf wolf@slac. stanford. edu

Hall Probe Noise ~4 G Noise ~0. 2 G The Senis probe did not work. We are only using Sentron probes. We own one Sentron probe and borrowed one from Isaac. We are having two custom Sentron probes made for us. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 31 Zachary Wolf wolf@slac. stanford. edu

Kugler Bench Noise ~0. 2 m. V Noise ~0. 03 m. V Rough tuning bench uses linear motor driver. Fine tuning bench uses chopped motor driver. Problem solved by adding heavy filtering to the motor lines. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 32 Zachary Wolf wolf@slac. stanford. edu

Fiducialization, Determine K vs X K K 0 X Set the probe position to give the nominal K value. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 33 Zachary Wolf wolf@slac. stanford. edu

Fiducialization August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 34 Zachary Wolf wolf@slac. stanford. edu

Tuning Bench Fiducialization Checks • Reference fiducialization magnets • Optical alignment check • Optical fiducialization check August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 35 Zachary Wolf wolf@slac. stanford. edu

CMM Fiducialization Checks August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 36 Zachary Wolf wolf@slac. stanford. edu

Conclusion • MMF construction is complete, temperature stability and set point accuracy requirements have been met. The fine tuning bench construction met all specifications. • The electrical noise problems are solved. • The planar Hall effect problem is solved. • The Hall probe calibration problem is solved. We check the Hall probe frequently in a reference magnet. • The software is complete. We compared our software results to other programs and found no bugs. • Mechanical alignment takes about 15 minutes. • The field integral measurement system has been tested and is working well. • We have a written test plan and a written fiducialization plan. • We demonstrated a throughput of one undulator per week. Each undulator spends about 4 weeks in the MMF. • One reference undulator gets re-measured every fourth undulator. • We are 1/3 finished with production undulator tuning. August 21 -24, 2007 LCLS Undulator Tuning And Fiducialization 37 Zachary Wolf wolf@slac. stanford. edu