The Advanced Photon Source Upgrade (APS-U) undulator requirements were changed from the first and second field integrals to the entrance and exit angles of the particle beam. This gives the user the best radiation view angle via the storage ring closed orbit correction system. SENIS HALL PROBE BEFORE IMPROVEMENTS An issue regarding the general question of the accuracy of Hall probe measurements has been unclear until recently. Calibration of the probe and measurements of the undulators are done under quite different conditions. Calibration is done at rest with a stable field, and measurements are done on the fly in a high-gradient field. It is quite obvious that one cannot expect exact accuracy of the field obtained in this case. It is not difficult to find the difference: just measure the field of the device on the fly and at rest at a few points. An extensive test of the Senis 2-axis Hall probes was done at the Advanced Photon Source using a 33-mm period undulator A device and the calibration system [1]. In addition to these general considerations, we recently found that the LCLS-II prototype undulator measurement results collected at the APS and at SLAC are different, and an investigation of this issue was performed. It was found that peak field results are speed dependent. To identify the reason for this difference, measurements of the undulator A device using a Senis 131-15 two-axis Hall probe were performed. Figure 1 shows the difference between magnetic field scans performed at different speeds. It shows that the main difference occurs at the regions with strong field gradient. One of the possible sources of the errors we described earlier is Faraday’s law, which states that the electromagnetic force (EMF) through the wire loop is given by the rate of change of the magnetic flux. It allows us to assume that the reason for this effect is due to some inductance and capacitance components in the vicinity of the Hall probe sensor. The results of speed dependence for a 35-cm-period xLEAP-II wiggler are shown in Fig. 2. The speed dependence for such a device with a long period is very small. The difference between high speed and low speed can be seen only at the locations with sharp field change. Figure 1: Magnetic field vs Z for a 33-mm period undulator A device (left scale) and the difference of the field measured at scanning speeds of 150 mm/s and 50 mm/s (right scale). Figure 2: A 35-cm-period xLEAP-II device. Blue curve: By field; red curve: difference of the field measured at scans with speeds of 100 mm/s and 20 mm/s. The integral of the field difference for different speeds is very close to the locations with strong field change. This statement proves our conclusion that this effect is due to the existence of inductance and/or capacitance at or near the sensor. SENIS HALL PROBE AFTER IMPROVEMENTS The recent generation of Senis Hall probes features many very important improvements. One of them is the elimination of Hall sensors zero drifts (see Fig. 3). -400 -300 -200 -100 0 100 200 300 400 -1 -0,8 -0,6 -0,4 -0,2 0 0,2 0,4 0,6 0,8 1 -1240-1220-1200-1180-1160-1140-1120-1100-1080-1060 Field difrence [G ] Fi el d [T ] Z [mm] Field (T) Field difference [G] -400 -300 -200 -100 0 100 200 300 400 -2 -1,5 -1 -0,5 0 0,5 1 1,5 2 100 150 200 250 300 350 400 Field difrence [G ] Fi el d [T ] Z [mm] Field (T) Field difference [G] ____________________________________________ *Work supported by U.S. Department of Energy, Office of Science, under contract number DE AC02-06CH11357. 12th Int. Particle Acc. Conf. IPAC2021, Campinas, SP, Brazil JACoW Publishing ISBN: 978-3-95450-214-1 ISSN: 2673-5490 doi:10.18429/JACoW-IPAC2021-WEPAB127 MC2: Photon Sources and Electron Accelerators T15 Undulators and Wigglers WEPAB127 2907 C on te nt fr om th is w or k m ay be us ed un de rt he te rm s of th e C C B Y 3. 0 lic en ce (© 20 21 ). A ny di st ri bu tio n of th is w or k m us tm ai nt ai n at tr ib ut io n to th e au th or (s ), tit le of th e w or k, pu bl is he r, an d D O I

The Advanced Photon Source Upgrade (APS-U) project is developing a multi-bend achromat (MBA) lattice at 6.0-GeV beam energy to replace the existing APS storage ring lattice operating at 7.0 GeV. One of the key components of the project is to design, fabricate, and install optimized insertion devices (IDs) for 35 beamlines. A plan was developed to standardize on four new undulator period lengths for 44 new undulators and to reuse 23 existing undulators with four more different period lengths. Early in the Upgrade project we anticipated there would be large challenges in meeting the tight fabrication and tuning schedules so that all undulators would be ready for installation in the upgraded storage ring prior to beam commissioning. With recent developments and techniques used in the magnetic measurement laboratory, we have successfully tuned many of the new and reused undulators to demanding magnetic field requirements. We will report on the tools and techniques used and on results to date.

We developed a new method for tuning the undulator phase errors by shimming the undulator gap profile mechanically. First, the phase errors of a device are calculated based on the initial field measurement. Then the desired field strength modulation along the device length is derived from the phase errors. Finally, the gap profile is mechanically shimmed to produce the desire field strength modulation. The method has been successfully applied to the tuning of many new and reused APS Upgrade (APS-U) hybrid permanent magnet undulators. The method is especially effective for tuning the legacy undulators with large phase errors. For instance, an old 33-mm-period undulator with a 23∘ initial RMS phase error largely due to radiation damage has been tuned to better than 3∘.

For the newly designed and fabricated APS Upgrade (APS-U) hybrid permanent magnet undulators (HPMUs), the development of magnetic shims has been critical to successfully tuning the undulators to meet the tight APS-U physics requirements [1]. Different types of side and surface shims have been developed and applied for this purpose. The side shims are primarily used for trajectory tuning, and the surface shims are for phase and multipole tuning as well as trajectory tuning. Current design, applications, and measurement of the shims for the newly designed and fabricated APS28 (28 mm period) undulators are presented in this paper. INTRODUCTION The APS chose many years ago to use hybrid undulator magnet designs to generate x-rays, which is also the case for the new undulators but with improved mechanical structures. However, the new undulators as well as the old undulators that will be reused, require magnetic tuning to meet the strict APS-U physics requirements [1]. The development of new shims was critical to the successful and efficient magnetic tuning of all the planned new APS28 undulators. The key factors to consider are the gap-dependent magnetic shim signatures, ease and repeatability of installation, and stability under strong magnetic forces. Figure 1 shows the first APS 28-mm-period undulator installed on the 3-meter-long measurement bench in the magnetic measurement laboratory. Figure 1: The first APS28 undulator on the measurement bench with the measurement coordinate system overlayed. The undulator trajectory and phase analysis are based on on-axis magnetic field measurements with Hall probes, and the multipole analysis is based on long-coil measurements. There are 168 and 148 poles on each jaw for the 2.4-mlong and 2.1-m-long APS28 undulators, respectively. SIDE SHIMS In a hybrid undulator a pole is sandwiched between two magnets except at the ends, which terminate with poles. The side shims are placed on one or both sides of a pole to alter/tune the magnetic field in the vicinity of the pole. They are normally used in pairs, with one shim installed on the top jaw and one installed on the bottom jaw at the same pole location. Side shims are preferred over surface shims because they do not reside near the beam axis in case a shim would come loose. However, side shims have limitations: due to their predominant normal dipole characteristics, they can be used primarily for trajectory tuning. Configurations For the newly designed undulator, the elemental side shim is made from a 0.5-mm-thick sheet metal of low carbon 1010 steel. A piece of side shim of a certain thickness consists of a number of stacked 0.5-mm-thick elemental shims sandwiched between two aluminium holders, as shown in Fig. 2. Figure 2: Photograph of a 1.5-mm-thick side shim (three 0.5-mm elemental shims stacked) placed against the tealcolored keeper. To the left of the shim is the pole and behind it is the first magnet. The shim was placed at the end pole for illustration only. Magnetic Signatures For magnetic signature measurements several pairs of side shims (more pairs for thin shims) are placed at poles of the same polarity in the core of undulator to obtain better averaged signatures. Figure 3 shows the measured normalized gap-dependent signatures for the APS28#1S undulator. All signatures were normalized to their values at a 10-mm gap. ___________________________________________ * Work supported U.S. Department of Energy, Office of Science, under contract # DE-AC02-06CH11357. † ypiao@anl.gov 12th Int. Particle Acc. Conf. IPAC2021, Campinas, SP, Brazil JACoW Publishing ISBN: 978-3-95450-214-1 ISSN: 2673-5490 doi:10.18429/JACoW-IPAC2021-THPAB077 MC2: Photon Sources and Electron Accelerators T15 Undulators and Wigglers THPAB077 3941 C on te nt fr om th is w or k m ay be us ed un de rt he te rm s of th e C C B Y 3. 0 lic en ce (© 20 21 ). A ny di st ri bu tio n of th is w or k m us tm ai nt ai n at tr ib ut io n to th e au th or (s ), tit le of th e w or k, pu bl is he r, an d D O I

Several undulator lines for the Advanced Photon Source Upgrade (APS-U) will consist of two inline undulators. In order to keep the undulators operating with optimal phasing over the full range of gaps, a phase shifter will be included between the undulators. A design has been developed for a phase shifter that will serve for a variety of undulator period lengths and gap ranges. The permanentmagnet phase shifter will use SmCo magnets to reduce the risk of radiation-induced demagnetization. The available space between the undulators is tight, so magnetic shields are placed between the undulators, the phase shifter, and the corrector magnet that is also located in the inter-undulator space. While these shields guard against magnetic cross-talk between the devices as the undulator and phase shifter gaps change, they do have an effect on the end fields of the devices. These end-field effects are examined and relevant tolerances are set and presented. INTRODUCTION Preparations for an upgrade to the APS are well underway. Changes to the insertion devices are included in the upgrade, both because of the change from an e-beam energy of 7 GeV to 6 GeV, and to allow users to take advantage of the improved electron beam characteristics. For eleven beamlines, the plan is to have two inline undulators that can be operated in tandem. Getting the full photon amplitude from the pair of undulators requires that they operate in phase with each other over the full range of usable undulator gaps so that destructive interference between photons from the two undulators is minimized. INLINE UNDULATOR SECTOR The inline undulators for APS-U will be the accustomed full length of 2.4 meters. They will come in a variety of period lengths, depending on the user’s choice, and some will be revolver undulators that offer a choice between two period lengths for users seeking greater flexibility. Between the undulators will be a phase shifter and a corrector. The anticipated layout is shown in Fig. 1. The distance from the end of one undulator to the beginning of the next is 197 mm, which puts the magnetic components close enough together that magnetic shields are required to interrupt magnetic crosstalk. Shields are at both ends of the undulator so the undulators are more readily swappable without retuning. The distance between the end of the undulator strongback and the nearest face of the shield was set to 20 mm based on experience tuning many undulators at APS. Note that the distance from the face of the last undulator pole to the nearest face of the shield will be slightly larger than 20 mm and will vary with period length while the strongback length can be set to 2400 mm, the magnetic structure length must be an integer number of half periods. Figure 1: Sketch of the inline undulator sector. The undulators are not to scale. The overall machined length of the vacuum chamber flat section is 5050 mm; keeping the distance from the far ends of the shields to 5045 mm allows for the radius at the end of that flat section. The gray borders around the undulator and corrector are non-magnetic support structure. Undulator length 2400 mm (not to scale) Shields are 4 mm Corrector 20 20 20 20 49.5 Pase sifter

At the Advanced Photon Source (APS) we have assembled, measured, and analyzed 14 new undulators of the exact same mechanical design, some of them with sorted magnets and some unsorted. The performance differences appear to be insignificant in meeting the tight APS Upgrade (APS-U) undulator requirements. SORTING It is worth mentioning that sorting of the magnets assumes that we know the most important sources of the errors. Very different types of errors affect the performance of an undulator, mainly: errors in sizes of the magnets, location of the magnets between the poles, differences in the pole gaps, and magnet recesses [1-4]. Justification of the fact that sorting is not perfect comes from the fact that with the same sorting techniques (simulated annealing, total moment, many points of the magnet field data, etc.) results are different. An important fact to mention is the quality of permanent magnets, which has been improved considerably. One of the ways to see how the sorting affects the performance is to compare the results with and without sorting. Magnet sorting at the APS was made on the total magnetic moment (M) using either the nominal magnet block volume or the real magnet block volume. Sorting using the nominal volume is equivalent to sorting of the total moment, and sorting using the real volume is equivalent to sorting on the total moment density. Previously, the horizontal component of the total moment Mx was included in the procedure, and sometimes the result led to unwanted changes in multipole components (see Fig. 1). Figure 1: Comparison of normal quadrupole components with Mx sorted and unsorted for the APS27#3 device (a 27-mm-period undulator). One of the 28-mm-period undulators, APS28#13S, was used for specific testing. It was first assembled using sorted magnets and then disassembled and reassembled with unsorted magnets (the magnet keepers and the so-called super-strongbacks were unchanged). Sorting did not consider either of the transverse components Mx or My, or the difference in field from the north and south faces of the magnets. The comparisons are shown in Figs. 2 and 3 for the sorted and unsorted cases, respectively. Any differences noted, whether before or and after tuning, including tuning efforts and the number of shims used, are not essential. Figure 2: Computed as-assembled trajectory at 6 GeV (top) and RMS phase errors (bottom) before tuning for the sorted case for the APS28#13S undulator at 8.2-mm gap. The asassembled RMS phase error was 4.88°. Figure 3: Computed as-assembled trajectory at 6 GeV (top) and RMS phase errors (bottom) before tuning for the unsorted case for APS28#13S undulator at 8.2-mm gap. The as-assembled RMS phase error was 5.51°. The final RMS phase error achieved was 2.64° for the sorted case and 2.67° for the unsorted case at 8.2-mm gap. -300 -200 -100 0 100 200 300 5 10 15 20 25 30 No rm al Q ua dr up ol e [G ] Gap [mm] Unsorted Sorted ____________________________________________ * Work supported by U.S. Department of Energy, Office of Science, under contract number DE AC02-06CH11357. 12th Int. Particle Acc. Conf. IPAC2021, Campinas, SP, Brazil JACoW Publishing ISBN: 978-3-95450-214-1 ISSN: 2673-5490 doi:10.18429/JACoW-IPAC2021-WEPAB128 WEPAB128 C on te nt fr om th is w or k m ay be us ed un de rt he te rm s of th e C C B Y 3. 0 lic en ce (© 20 21 ). A ny di st ri bu tio n of th is w or k m us tm ai nt ai n at tr ib ut io n to th e au th or (s ), tit le of th e w or k, pu bl is he r, an d D O I 2910 MC2: Photon Sources and Electron Accelerators T15 Undulators and Wigglers As seen in Fig. 4 there is large difference in the average top-bottom jaw total moment distribution for the two cases. The RMS value for the sorted case is about 30 times smaller than for the unsorted case. Nevertheless, clearly the end results of the tuning are very close for both (see Fig. 5). Figure 4: Average top-bottom jaw total moment distribution of device APS28#13S for sorted and unsorted cases. The RMS value is 2.54×10-5 Tesla for the sorted case and 7.73×10-4 Tesla for the unsorted case. Figure 5: Comparison of RMS phase errors vs. gap for the APS28#13S undulator with sorted and unsorted magnets after final tuning. The RMS phase error requirement for the APS-U is less than 3°. Table 1 lists all new undulators tuned to date, with magnet sorting status indicated, including two recently assembled shorter-period devices, and the achieved RMS phase errors after tuning. Table 1: Comparison of All Tuned New Undulators Device ID RMS Phase Error [Deg] Sorting Status APS28#1S 2.6 Unsorted APS28#2 2.6 Sorted APS28#3S 2.7 Sorted APS28#4S 2.5 Sorted APS28#5S 2.9 Sorted APS28#6S 2.8 Sorted APS28#7 2.7 Sorted APS28#8 2.4 Sorted APS28#9S 2.4 Sorted APS28#10S 3.1 Sorted APS28#11S 2.4 Sorted APS28#12S 2.1 Sorted APS28#13S 2.6 Sorted APS28#13S 2.7 Unsorted APS25#1S 2.1 Unsorted APS21#1S 1.6 Unsorted CONCLUSION Any differences between using sorted and unsorted rare earth permanent magnets from Kyma and Shin-Etsu (magnet vendors) appear to be insignificant in meeting the tight APS-U undulator requirements. Therefore, we have decided not to sort the magnets for the remainder of the new undulators to save time and effort.

The Advanced Photon Source (APS) is undergoing a major upgrade to its storage ring. The APS Upgrade (APS-U) project plans to build over 40 new hybrid permanent magnet undulators (HPMUs) and rebuild over 20 existing HPMUs. To meet the APS-U undulator requirements, the quality of the undulator magnetic field needs to be fine-tuned to the specifications. The traditional methods that depend on the tuning specialist’s experience are not desirable for tuning large quantities of undulators. We developed algorithms that automate the tuning of permanent magnet undulators. For tuning of the undulator trajectory and phase, the algorithms optimize the tuning parameters with differential evolutionbased global optimization. The algorithms have been successfully applied to 24 APSU HPMUs. The results and experiences of the tuning are reported in detail.

As part of the LCLS-II R&D program, a novel 3.4meter long undulator prototype with a horizontal main magnetic field and dynamic force compensation — called the horizontal gap vertical polarization undulator (HGVPU) — has recently been developed at the Advanced Photon Source (APS). Initial steps of the project included designing, building, and testing a 0.8meter long prototype [1]. Extensive mechanical testing of the HGVPU has been carried out. The magnetic tuning was accomplished by applying a set of magnetic shims. As a result, the performance of the HGVPU meets all the stringent requirements for the LCLS-II insertion device [2], which includes limits on the field integrals and phase errors for all operational gaps, as well as the reproducibility and accuracy of the gap settings. The HGVPU has been included in the baseline of the LCLS-II project for the hard x-ray undulator line.

As part of the R&D program of the LCLS-II project, a novel 3.4-meter-long undulator prototype with horizontal magnetic field and dynamic force compensation has recently been developed at the Advanced Photon Source (APS). Previous steps in this development were the shorter 0.8-meter-long and 2.8-meter-long prototypes. Extensive mechanical and magnetic testing were carried out for each prototype, and each prototype was magnetically tuned using magnetic shims. The resulting performance of the 3.4-meter-long undulator prototype meets all requirements for the LCLS-II insertion device, including limits on the field integrals, phase errors, higher-order magnetic moments, and electron-beam trajectory for all operational gaps, as well as the reproducibility and accuracy of the gap settings.

The Horizontal-Gap Vertically-Polarizing Undulator (HGVPU) is a compact, innovative, variable-gap insertion device developed by Argonne National Laboratory for the LCLS-II HXR beamline at SLAC. A full sized 3.4-meterlong prototype has been built and fully tested meeting all LCLS-II undulator specifications. An array of conical springs compensates the attractive magnetic forces of the undulator jaws. These springs are designed to exhibit nonlinear spring characteristics that can be closely tuned to match the force curve exerted by the magnetic field, thereby minimizing the overall deflection of the strongbacks. The HGVPU also utilizes the existing LCLS-I support and motion system along with other existing equipment and infrastructure, thus lowering overall cost and installation downtime. Figure 1: Rendering of final 3.4m HGVPU prototype.

A preliminary design of the multi-meter long collinear dielectric wakefield accelerator that achieves a highly efficient transfer of the drive bunch energy to the wakefields and to the witness bunch is considered. It is made from ~0.5m long accelerator modules containing a vacuum chamber with dielectric-lined walls, a quadrupole wiggler, an rf coupler, and BPM assembly. The single bunch breakup instability is a major limiting factor for accelerator efficiency, and the BNS damping is applied to obtain the stable multi-meter long propagation of a drive bunch. Numerical simulations using a 6D particle tracking computer code are performed and tolerances to various errors are defined.

27-mm period undulator with a gap of 11 mm. The calculated field roll-off with the new optimized model is within the requirements of the MBA, in the range of ± 5 mm.

length of ~0.5 m. These two steps led to the final design of ~1.1–m-long magnets, which are currently being fabricated. The quench behavior of each 0.5-m-long undulator magnet, as well as undulator assemblies from these magnets, was studied. The first SCU magnet assembly did not meet the design specifications due to the breakdown of the insulation. With an improved design and fabrication process based on lessons learned, the second SCU assembly achieved the design undulator field of 1.2 T. The design was further optimized, and the third set of magnets was fabricated and successfully tested.

A short prototype (847-mm-long) of an Insertion Device (ID) with the dynamic compensation of ID magnetic forces has been designed, built, and tested at the Advanced Photon Source (APS) of the Argonne National Laboratory. The ID magnetic forces were compensated by the set of conical springs placed along the ID strongback. Well-controlled exponential characteristics of conical springs permitted a very close fit to the ID magnetic forces. Several effects related to the imperfections of actual springs, their mounting and tuning, and how these factors affect the prototype performance has been studied. Finally, series of tests to determine the accuracy and reproducibility of the ID magnetic gap settings have been carried out. Based on the magnetic measurements of the ID Beff, it has been demonstrated that the magnetic gaps within an operating range were controlled accurately and reproducibly within ±1 μm. Successful tests of this ID prototype led to the design of a 3-m long device based on the same concept. The 3-m long prototype is currently under construction. It represents R&D efforts by the APS toward APS Upgrade Project goals as well as the future generation of IDs for the Linac Coherent Light Source (LCLS).

As the western hemisphere's premier x-ray synchrotron radiation source, the Advanced Photon Source (APS) continues to advance the state of the art in insertion device technology in order to maintain record high brightness, especially in the hard x-ray wavelength region. Due to the unique bunch pattern used for normal APS operations and its ultimate capabilities, the APS has chosen superconducting technology for its future hard x-ray undulator sources. In the last several years, the APS in collaboration with the Budker Institute of Nuclear Physics has being developing the technology for planar, small-period superconducting undulators (SCUs). These developments include the design and construction of several prototypes and the construction of the necessary mechanical, vacuum, and cryogenic infrastructure at the APS site. Several prototypes of the SCU magnetic structure have been built and tested. The first SCU is assembled and will be installed in the APS storage ring at the end of 2012. Expected SCU performance in terms of x-ray brightness should noticeably exceed that of existing APS undulators. Immediately after commissioning, the SCU will be used at APS Sector 6 as the radiation source for high-energy x-ray studies.

An enhanced field bending magnet (EFBM), a pie-shaped dipole magnet that produces a peak field of about 1.2 T at the center of its gap of 4.9 cm, was recently designed for Sector 2 at the Advanced Photon Source (APS). The By field is maximized and its multipole components over X and Y are minimized by introducing two different chamfers on each edge of the iron poles. The calculated roll-off fields around the gap center in the region of ±1.5 cm in X and ± 1.8 cm in Y are 1.4 × 10-4 and 8.3 × 10-5, respectively, at the center of the EFBM length in Z. This is a different idea as compared to shaping the top surface of iron poles by changing the gap size over the horizontal direction to get a good roll-off field, as was done for the standard APS dipole. The peak field at the center of the gap of the EFBM as a function of the current shows there is no magnetic saturation on the iron poles up to a field of 1.2 T.