Beam Diagnostics for the BNL Energy Recovery Linac Test Facility Peter Cameron, Ilan Ben-Zvi, Michael Blaskiewicz, Michael Brennan, Roger Connolly, William Dawson, Chris Degen, Al DellaPenna, David Gassner, Martin Kesselman, Jorg Kewish, Vladimir Litvinenko, Joseph Mead, Brian Oerter, Tom Russo, Kurt Vetter, and Vitaly Yakimenko Brookhaven National Laboratory, Upton, NY 11973, USA Abstract. An Energy Recovery Linac (ERL) test facility is presently under construction at BNL. The goals of this test facility are first to demonstrate stable intense CW electron beam with parameters typical for the RHIC e-cooling project (and potentially for erhic), second to test novel elements of the ERL (high current CW photo-cathode, superconducting RF cavity with HOM dampers, and feedback systems), and finally to test lattice dependence of stability criteria. Planned diagnostics include position monitors, loss monitors, transverse profile monitors (both optical and wires), scrapers/halo monitors, a high resolution differential current monitor, phase monitors, an energy spread monitor, and a fast transverse monitor (for beam break-up studies and the energy feedback system). We discuss diagnostics challenges that are unique to this project, and present preliminary system specifications. In addition, we include a brief discussion of the timing system. INTRODUCTION The motivation for the development of the Energy Recovery Linac at BNL is to have high-quality, high-current electron beams (with acceptable RF power requirements) for high-energy bunched-beam electron cooling of ions [1], and possibly for an electron-ion collider [2]. By high quality, we mean beams with minimal emittance, and whose profiles are sharper than Gaussian. The ERL approach proposes to accomplish this not by relying upon equilibrium with radiation to establish the emittance (as in the case of storage rings) but rather by preserving the low emittance (typical of a linac) of a carefully designed source during acceleration for a single pass through the interaction region, then recovering the beam energy with a second pass through the acceleration cavities. This method permits significant reduction in beam emittance compared to a storage ring, and also permits high current without requiring unreasonable RF power. Figure 1 shows a possible layout of the test facility (others are under consideration). Expected parameters are shown in Table 1. Work performed under the auspices of the U.S. Department of Energy. 232
FIGURE 1. A Possible Layout of the Test Facility. ERL-SPECIFIC REQUIREMENTS The ERL imposes diagnostics requirements [3, 4] beyond those normally present in linacs and storage rings. One such requirement is to measure with high resolution the difference in currents between the accelerated and decelerated beams. In principle this could be accomplished by measuring the power requirement of the SRF cavity, but that presumes that the low- and high-energy beams are accurately anti-phased, and proper calibration of such a measurement is not completely straightforward. A simple and elegant method is to utilize two toroids placed immediately after the line to the beam dump, one in the dump line and one in the accelerated beam line, and to link those toroids with a figure eight winding. The output of one toroid is used to drive a nulling current through the figure eight, and the output of the second toroid is then the differential current measurement. This overcomes the dynamic range problem of measuring a small current difference in the presence of a large current signal. As implied in the above discussion of current monitoring, a second requirement is to have the bunches properly phased through the SRF cavity in both the acceleration and deceleration passes, to minimize momentum spread in the accelerated beam (crucial for electron cooling) and to maximize energy recovery from the decelerated beam. Phasing for the acceleration pass can be accomplished by measuring beam position and/or profile in dispersive regions. Phasing for the deceleration pass will be accomplished by adjusting path length with a small chicane, and can be monitored in a 233
variety of ways. In a method similar to that used for the accelerated beam, beam position and/or profile can be measured in the dump beamline. A second possibility is to directly monitor the SRF cavity drive power, a minimum in the required power indicating good phasing. A third method is direct measurement of phase of either or both beams. Our intent is to I/Q demodulate the outputs of all BPMs. This will provide accurate phase measurement with single beams, but would impose excessive demands on time resolution if both beams were monitored by a single pickup. BPM phase measurements will require a means of accurately correcting time delays relative to the SRF. There is the possibility to construct a longitudinal pickup based on the same principle as the SRF cavity, where a null measurement is accomplished by the antiphasing of the two beams. However, it is not yet clear that there is a need for such a pickup, that it would provide any advantage over simply monitoring the SRF cavity drive power. TABLE 1. Machine parameters Parameter [units] high chg low emit injection energy [MeV] 5 5 beam energy [MeV] 15-20 15-20 rms bunch length [ps] ~20 ~20 RF frequency [MHz] 704 704 revolution freq [MHz] 9.4 9.4 bunching freq [MHz] 9.4/ 704 28.2 charges/bunch ~1e1 4e9 1 beam current [ma] ~150/ 500 450 rms energy spread 10e-3 10e-4 εx, εy [mm-mrad] 30 5 beampipe dia [cm] 5 5 current recovery [%] 99.95 99.95 A third ERL-specific requirement is to have the accelerated and decelerated beams on a common center and in addition to have that common center centered in the SRF cavity, to minimize excitation of higher-order transverse modes and thereby raise the threshold for the beam breakup (BBU) instability. This implies position monitors immediately before and after the SRF cavity. In low duty cycle operation the time separation (~100 ns) of the low and high energy pass through the pickup would impose a not-unreasonable lower limit on position monitor electronics bandwidth, so that independent position measurement and correction of both beams is feasible. Centering of the aligned beams on the SRF cavity might then be accomplished either by reliance on survey data, or as beam current goes up by measuring higher-order mode power in the SRF cavity. However, this measurement could not be accomplished when most crucial, during normal full-current operations. Alignment of 234
the two beams relative to each other during normal operations might be accomplished via the TM120 and 210 modes of a rectangular resonant cavity pickup. Such a pickup might be operated in the vicinity of the third harmonic of the SRF cavity frequency to minimize impedance seen by the beam. Positioning such pickups both immediately before and after the SRF cavity would permit proper transverse alignment through the cavity, and the TM111 mode might also provide a high quality longitudinal phase measurement. A method [3] which is similar in principle but not yet demonstrated in practice, would center the two beams relative to each other by looking at a conventional BPM output at the RF fundamental and second harmonic and tuning to minimize the fundamental and maximize the second harmonic. Bunch length measurement for the BNL ERL is perhaps not so critical as in ERLs designed to produce ultra-short bunches. Ultra-short bunches are not a requirement for either e-cooling or erhic. Bunch length can be measured most directly with a streak camera. Less direct measurement can be accomplished by comparing transverse profiles in dispersive and non-dispersive regions, perhaps in combination with variation of the RF phase. Finally, it is crucial to monitor and localize the development of halo in the ERL. Useful information on halo monitoring may be found in the proceedings of a recent workshop [5]. The measurement requirements in the ERL will require careful implementation of the best methods described in that workshop. ARCHITECTURE AND SPECIFICATIONS One responsibility of the Diagnostics Group for the Brookhaven ERL will be to provide an economic, reliable and uniform interface to the RHIC timing and control systems. We intend to utilize well-developed expertise to accomplish this with a straightforward extension of the approach taken in the RHIC BPM system [6], and later refined in SNS Diagnostics [7, 8]. In both cases the interface to timing is encoder logic embedded in a gate array in each data acquisition module. In RHIC the data acquisition is controlled by a DSP in each custom module, where calibration and initial processing occurs before data is transmitted via firewire to VME and the control system. Additional processing and display are accomplished at the application level. In the SNS, data acquisition is controlled by LabVIEW running in Windows, and communication to the control system is via Ethernet. The timing decoder gate array board also provides the interface to the PCI bus, serves as a motherboard for the analog electronics and digitizers, and permits the possibility of fast pre-processing before delivery of data to LabVIEW. The intent is to employ a similar architecture wherever possible for the ERL diagnostics systems, a plan that may also be compatible with future upgrade plans for the RHIC BPM system. A second responsibility of the Diagnostics Group will be the instruments themselves. Design, implementation, and utilization of the instruments is always a collaboration between Diagnostics and Accelerator Physics. In the case of several of the systems (position monitors, loss monitors, wire scanners, scrapers, current monitors, etc.) most of the needed expertise already resides within the Diagnostics Group, and involvement of Accelerator Physics will probably not extend beyond 235
specification of system requirements. However, for those systems specific to electron machines (synchrotron light monitors, transition radiation monitors, streak cameras, etc.) and more specifically to the ERL (phase monitor, energy spread monitor, high resolution differential current monitor, etc.), the experience within the Diagnostics Group is somewhat limited, and there will be greater reliance on Accelerator Physics in design, implementation, and operation. As suggested by Table 1, the ERL will operate in either of two modes. In the first (high bunch charge) mode, the 9.4/28.2 MHz bunch frequency is appropriate for electron cooling of 120/360 bunches in RHIC (although other bunching frequencies are under consideration). In the second, every bucket of the 703.75 MHz RF will be filled for low emittance studies. The additional dynamic range required by these two modes adds only minimal complication or expense. However, some forethought is required as a result of the sparse spectrum with the 703.75 MHz bunching frequency. This makes it more difficult to work away from the RF frequency, and imposes more stringent demands on RF shielding. Our intent is to provide flexibility where possible (for instance, by using programmable synthesizers to generate local oscillator frequencies) to permit avoiding the RF fundamental when operating with the 9.4/28.2 MHz bunching frequency. Processing at 14.1 MHz looks particularly attractive, as this will permit the same LO frequency for either e-cooling or high current mode, and the resulting 56.4 MHz clock frequency for BPM I/Q demodulation is comfortable for the intended digitizer. A preliminary list of proposed Diagnostics, together with their preliminary Accelerator Physics specifications, is shown in Table 2. In addition to the two modes (high charge/low emittance) mentioned above, these diagnostics must also meet the needs of both low duty-cycle commissioning and high current operations. Refinement of specifications to include this consideration is in progress. CONCLUSION We have discussed ERL-specific diagnostics requirements, presented a proposed data acquisition platform and timing/controls interface, and made a first attempt to define diagnostics devices, quantities, locations, and specifications. More detailed system-by-system designs will be presented in a series of forthcoming tech notes. 236
TABLE 2. Diagnostic devices and AP Specifications Device Qty Range Accuracy Resolution Comments Position/Phase BPM (button) 12 1/2 pipe 500µ 1µ (av)/100µ Dual plane rad BBU/Energy Feedback 1 Sample scope Beam Transfer Function 1 Include BTF kicker Energy Spread 2 10-4 10-5 Dispersive BPMs Phase 8 +/- 180 deg +/- 2 deg 0.2 deg BPMs w/ I/Q Loss BLM (PMT) 10 1-1000 rem/h 30% 0.5 rem/h 20µsec and 1sec BLM (cable ion 10 1-1000 30% 0.5 rem/h 20µsec and 1sec chamber) rem/h Current Current 12 5% 1% BPM sum signal Differential 1 10-4 10-5 2 toroids w/ null Profile Flags 4? 0.2σ 0.1σ Phosphor + TR 3 inj line, 1 dump Wire Scanner - profile 4? Full aperture 0.2σ SEM mode, 1 inj & dump, 2 ring Wire Scanner - halo 3? 10-6 BLM mode Scraper 2? 0.2σ SEM + BLM Synch Light 8? 0.2σ Every bend mag Compton 2? 0.2σ Inj line Streak Camera 1? 0.2σ Dual sweep REFERENCES 1. Ilan Ben-Zvi et al., R&D Towards Cooling of the RHIC Collider, PAC 2003, Portland. 2. M. Farkhondeh and V. Ptitsyn, erhic Zeroth Order Design Report, BNL CAD AP note 142, March 2004. (available at http://www.agsrhichome.bnl.gov/ap/ap_notes/cad_ap_index.html) 3. G. A. Krafft and J-C Denard, Diagnostics for Recirculating and Energy Recovered Linacs, BIW 2002, Brookhaven. 4. P. Piot et al., Performance of the Electron Beam Diagnostics at Jefferson Lab s High Power Free Electron Laser, PAC 1999, New York. 5. P. Cameron and K. Wittenberg, Halo Diagnostics Summary, Halo 2003, Montauk, and references therein. 6. T. J. Shea and R.L. Witkover, RHIC Instrumentation, BIW 1998, Stanford. 7. W. Dawson et al., BPM System for the SNS Ring and Transfer Lines, BIW 2002, Brookhaven. 8. K. Vetter et al., RF Beam Position Monitor for the SNS Ring, these proceedings. 237