• H. Maesaka, S.I. Inoue, S. Matsubara, Y. Otake
  RIKEN/SPring-8, Hyogo

The x-ray FEL fa­cil­i­ty at SPring-8 pro­duces a very short-bunch beam by using bunch com­pres­sors (BC) con­sist­ing of mag­net­ic chi­canes. Since the bunch com­pres­sion ratio is strong­ly de­pends on the beam en­er­gy and the en­er­gy chirp, we need to mon­i­tor the en­er­gy from the beam po­si­tion at the dis­per­sive part of the BC with a 0.1% res­o­lu­tion. How­ev­er, a beam pro­file at the dis­per­sive part is hor­i­zon­tal­ly flat and wide, max­i­mal­ly 50 mm, due to the large en­er­gy chirp of the beam. There­fore, we de­signed a mul­ti-stripline beam po­si­tion mon­i­tor. This mon­i­tor has a flat rect­an­gu­lar duct with a 70 mm width and a 10 mm height. Six stripline elec­trodes at in­di­vid­u­al in­ter­vals of 10 mm are equipped on each of the top and the bot­tom sur­face. Due to the small height of the mon­i­tor, each elec­trode is sen­si­tive to the elec­tron po­si­tion with­in 10 mm in the hor­i­zon­tal. There­fore, the mon­i­tor pro­vides a rough charge pro­file and the beam po­si­tion which is cal­cu­lat­ed from the grav­i­ty cen­ter of the sig­nals. We pre­pared a pro­to­type of the mon­i­tor and test­ed it at the SCSS test ac­cel­er­a­tor. We con­firmed that the po­si­tion sen­si­tiv­i­ty was bet­ter than 0.1 mm, which cor­re­sponds to 0.1 % en­er­gy res­o­lu­tion.

• S. Matsubara, A. Higashiya, H. Maesaka, T. Ohshima, Y. Otake, T. Shintake, H. Tanaka, K. Togawa, M. Yabashi
  RIKEN/SPring-8, Hyogo
• H. Ego, S. Inoue, K. Tamasaku, T. Togashi, H. Tomizawa, K. Yanagida
  JASRI/SPring-8, Hyogo-ken

We re­port the de­sign, per­for­mance, and in­stal­la­tion of the beam di­ag­nos­tic sys­tem of XFEL/SPring-8. The elec­tron beam bunch­es of an XFEL ac­cel­er­a­tor are com­pressed from 1 ns to 30 fs by bunch com­pres­sors with­out emit­tance growth and peak-cur­rent fluc­tu­a­tion which di­rect­ly cause SASE fluc­tu­a­tion. To main­tain the sta­ble bunch com­pres­sion pro­cess, the ac­cel­er­a­tor re­quires rf caiv­ty beam po­si­tion mon­i­tors (BPM) with 100 nm res­o­lu­tion, OTR screen mon­i­tors (SCM) with a few mi­cro-me­ter res­o­lu­tion, fast beam cur­rent mon­i­tors (CT) and tem­po­ral struc­ture mea­sure­ment sys­tems with res­o­lu­tion under pi­cosec­ond. The per­for­mance of the de­vel­oped mon­i­tor in­stru­ments, such as the BPM, the SCM, and the CT, was test­ed at the SCSS test ac­cel­er­a­tor and sat­is­fied with the re­quire­ments. To mea­sure the tem­po­ral struc­ture of the elec­tron bunch, three type mea­sure­ment sys­tems, which are a streak cam­era, an EO sam­pling mea­sure­ment, and a trans­verse de­flect­ing cav­i­ty with a res­o­lu­tion of few-tens fem­tosec­ond, are being pre­pared. The streak cam­era and EO sam­pling shows the res­o­lu­tion of sub-pi­cosec­ond. The in­stal­la­tion of these beam di­ag­nos­tic sys­tems is going on smooth­ly.

• T. Ohshima, N. Hosoda, H. Maesaka, S. Matsubara, Y. Otake
  RIKEN/SPring-8, Hyogo

The pulse width of an X-ray laser at XFEL/SPring-8 is sev­er­al tens fem­to-sec­onds, which re­quires ref­er­ence rf sig­nals to have the same time-sta­bil­i­ty. The ref­er­ence sig­nals with a low phase-noise os­cil­la­tor are sent to in­stru­ments in 19" racks de­vel­oped along an ac­cel­er­a­tor by an op­ti­cal fiber sys­tem. The tem­per­a­ture drift of the fiber makes phase shifts of the ref­er­ence sig­nals. There­fore, the fiber is put in a ther­mal-in­su­lat­ed duct. By feed­ing tem­per­a­ture-con­trolled water (26.1 ± 0.1 deg. C) in a pipe at­tached to the duct, the fiber tem­per­a­ture was kept to be 26.2 ± 0.08 deg. C at the am­bi­ent tem­per­a­ture change of 29.1 ± 1.7 deg. C. From this tem­per­a­ture con­trol­la­bil­i­ty, the phase shifts of the sig­nals through a 400 m fiber of a ther­mal co­ef­fi­cient of 5 ps/km/K are 160 fs. Fur­ther re­duc­tion of the shifts is re­quired and will be achieved by a fiber-length feed­back con­trol in a fu­ture plan. Vi­bra­tion of the fiber also de­grades the qual­i­ty of the sig­nals. The fiber is em­bed­ded on a vi­bra­tion buffer ma­te­ri­al. A test to eval­u­ate the ef­fect of the vi­bra­tion to the trans­mit­ted sig­nal phase was car­ried out. The test re­sult will be also shown in this paper.

• T. Hasegawa, T. Inagaki, Y. Otake, T. Sakurai
  RIKEN/SPring-8, Hyogo
• S. Takahashi
  JASRI/SPring-8, Hyogo-ken

This paper de­scribes the pre­sent sta­tus of a pre­cise tem­per­a­ture-reg­u­la­tion sys­tem for the C-band ac­cel­er­a­tor at XFEL (X-ray Free Elec­tron Laser)/SPring-8. It is es­sen­tial to main­tain a con­stant tem­per­a­ture of an rf cav­i­ty for sta­ble las­ing. We there­fore in­stalled a heater-as­sem­bly unit into a cool­ing water cir­cuit of each rf cav­i­ty. By con­trol­ling the heater power, the tem­per­a­ture of the cav­i­ty can be sta­bi­lized. We con­struct­ed a pro­to­type of this sys­tem at the SCSS (SPring-8 Com­pact SASE Source) test ac­cel­er­a­tor to check its fea­si­bil­i­ty for the XFEL. The pro­to­type sig­nif­i­cant­ly con­tributes to a sta­ble sup­ply of SASE to users. For the XFEL, we sim­pli­fied this sys­tem in con­sid­er­a­tion of cost and con­trol­la­bil­i­ty. For ex­am­ple, to make one reg­u­la­tion sys­tem si­mul­ta­ne­ous­ly con­trol­ling two C-band ac­cel­er­at­ing struc­tures was tried. Keep­ing a tem­per­a­ture vari­a­tion as tight as ±0.02 K at any op­er­a­tional mode could be achieved by this sys­tem. The pre­lim­i­nary test re­sults of the sys­tem are also re­port­ed in this paper.

• H. Hanaki, T. Asaka, H. Ego, H. Kimura, T. Kobayashi, S. Suzuki, M. Yamaga
  JASRI/SPring-8, Hyogo-ken
• T. Fukui, T. Inagaki, N. Kumagai, Y. Otake, T. Shintake, K. Togawa
  RIKEN/SPring-8, Hyogo

The in­jec­tor of the 8 GeV linac gen­er­ates an elec­tron beam of 1 nC, ac­cel­er­ates it up to 30 MeV, and com­press­es its bunch length down to 20 ps. Even slight RF in­sta­bil­i­ty in its mul­ti-stage bunch­ing sec­tion fluc­tu­ates the bunch width and the peak cur­rent of an elec­tron beam and it ac­cord­ing­ly re­sults in un­sta­ble laser os­cil­la­tion in the un­du­la­tor sec­tion. The ac­cept­able in­sta­bil­i­ties of the RF fields in the cav­i­ties, which per­mit 10% rms vari­a­tion of the peak beam cur­rent, are only about 0.01% rms in am­pli­tude and 120 fs rms in phase ac­cord­ing to beam sim­u­la­tion. The long-term RF vari­a­tions can be com­pen­sat­ed by feed­back con­trol of the RF am­pli­tude and phase, the short-term or pulse-to-pulse vari­a­tions, how­ev­er, have to be re­duced as much as pos­si­ble by im­prov­ing RF equip­ment such as am­pli­fiers. Thus we have care­ful­ly de­signed and man­u­fac­tured the RF cav­i­ties, am­pli­fiers and con­trol sys­tems, giv­ing the high­est pri­or­i­ty to the sta­bi­liza­tion of the short-term vari­a­tions. Com­po­nents of the in­jec­tor will be com­plet­ed by the end of the April 2010, and the in­jec­tor will be per­fect­ed in the sum­mer 2010. We will pre­sent the per­for­mance of the com­plet­ed de­vices in the con­fer­ence.

• N. Hosoda, H. Maesaka, S. Matsubara, T. Ohshima, Y. Otake, K. Tamasaku
  RIKEN/SPring-8, Hyogo
• M. Musha
  University of electro-communications, Tokyo

The in­ten­si­ty of SASE gen­er­at­ed by un­du­la­tors is sen­si­tive to the peak in­ten­si­ty fluc­tu­a­tion of an elec­tron bunch. The bunch is formed by ve­loc­i­ty bunch­ing in an in­jec­tor and mag­net­ic bunch­ing in bunch com­pres­sors (BC). The peak in­ten­si­ty is sen­si­tive to rf phase and am­pli­tude of off-crest ac­cel­er­a­tion at in­jec­tor cav­i­ties and 5712 MHz cav­i­ties be­fore the BCs. Thus, de­mand­ed sta­bil­i­ties of the rf phase and am­pli­tude for sta­ble SASE gen­er­a­tion are very tight. These are 0.6 de­gree (p-p) and 0.06 % (p-p) at the 5712 MHz cav­i­ties, re­spec­tive­ly. We are con­struct­ing a low-lev­el rf (LLRF) sys­tem com­pris­ing a mas­ter os­cil­la­tor, an op­ti­cal rf sig­nal trans­mis­sion sys­tem, and a dig­i­tal rf con­trol sys­tem using IQ mod­u­la­tor/de­mod­u­la­tor to drive klystrons. To re­al­ize the de­mands, much at­ten­tion was paid to tem­per­a­ture sta­bi­liza­tion for the sys­tem. A wa­ter-cooled 19-inch rack and a wa­ter-cooled cable ducts are em­ployed for al­most all part of the sys­tem. Tem­per­a­ture sta­bil­i­ty of the rack was 0.4 K (p-p) even though out­side was 4 K (p-p). The phase and am­pli­tude sta­bil­i­ties of the LLRF mod­ules were mea­sured to be 0.30 de­gree (p-p) and 0.56 % (p-p). These sta­bil­i­ties are suf­fi­cient for our de­mands.

• T. Asaka, H. Ego, H. Hanaki, T. Kobayashi, S. Suzuki
  JASRI/SPring-8, Hyogo-ken
• T. Inagaki, Y. Otake, K. Togawa
  RIKEN/SPring-8, Hyogo

XFEL/SPring-8 is under con­struc­tion, which is aim­ing at gen­er­at­ing co­her­ent, high bril­liance, ul­tra-short fem­to-sec­ond X-ray pulse at wave­length of 1Å or short­er. The in­jec­tor con­sists of a 500kV thermion­ic gun (CeB6), a beam de­flect­ing sys­tem, mul­ti-stage RF struc­tures and ten mag­net­ic lens­es. The mul­ti-stage RF struc­tures (238MHz, 476MHz, 1428MHz) are used for bunch­ing and ac­cel­er­at­ing the beam grad­u­al­ly to main­tain the ini­tial beam emit­tance. In ad­di­tion, in order to re­al­ize lin­eariz­ing the en­er­gy chirp of the beam bunch at three mag­net­ic bunch com­pres­sion sys­tems after the in­jec­tor sys­tem, we pre­pared extra RF struc­tures of 1428MHz and 5712MHz. It is im­por­tant to sta­bi­lize the gap volt­age of those RF struc­tures be­cause the in­ten­si­ty of X-ray pulse is more sen­si­tive for a slight vari­a­tion of the RF sys­tem in the in­jec­tor. We de­vel­oped some sta­ble am­pli­fiers for those RF struc­tures, and con­firmed the am­pli­tude and phase sta­bil­i­ty of an RF sig­nal out­putted from the am­pli­fiers. The mea­sure­ment re­sults achieved near­ly the re­quire­ment of de­sign pa­ram­e­ters. In this paper, we de­scribe the de­vel­op­ment sta­tus and the achieved per­for­mances of RF equip­ment of the in­jec­tor sec­tion.