PEP - II LFB - Status and Predictions

Transcription

1 PEP - II LFB - Status and Predictions Measurements and observations from o/30 and 10/31 high-current operation at 2.2A Predictions and impications for upgraded high-current configurations J. Fox, S. Prabhakar, D. Teyteman, A. Young, et a December 7,200O 484

2 Tak Outine Operationa history, measurements of growth rates, moda patterns, and phase transients Considerations for 1A HER, 2A LER operations noise foor moda growth and damping rates interaction of RF and LFB systems PEP-II upgrade proposa-considerations of added RF stations, higher currents Instabiity contro Therma management Summary - suggestions for future work 485

3 Hardware summary, operationa history LER, HER systems in routine use - Both HER, LER each configured with 2 kickers, 1500 watts kicker power (HER running on 1 kicker, 1 kw) Both running Woofer ink to contro modes -1Ocm40 Growth rates in HER ower than anticipated, benefit of tune spread from gap transients Operationa issues (as viewed from LFB perspective) Tuning of RF, woofer equaization very important for contro of ow modes, saturation-free contro of true HOM modes Fi-pattern dependence of growth rates, woofer configurations, beam induced kicker power PEP-II damped HOM s give bands of unstabe modes, unike ALS,PLS, BESSY-II and SPEAR- 486

4 ~..._... _. rpd.diunphg.mte.....! ~i~>o..... _: z c 10- r::......_ $0,:.. c F._ ~ ::::::2 ::::::::::::::j:::::::::::::::: ::::: 1...:...,...:...:.....,......, I... I::::::,:: :.::::::::::: 1:::::;:::::: :::::: : : i::::::~::.:.:::::::::::::::::.:::::::::::::.: t....~...~... r I : \I 487

5 a)oscgwdopesinimedawwn b)evowondmodos c) Exp. Fit to Modes (pm-brkpt) PEP-JI LERoct31 OW1346: b= mA. &amp= 6, SMGain= 2. Nbun= 872, Gair= -1, GairQ= 0, P= M4, Ptnase2= 7Q. 650, Cai& k 0 s 0 Pa 2u 300 Qoo ModeNo. 488

6 High-current MD Resuts October 30/3 1 High current LER measurements by2 pattern, 1658 bunches contro at ma - gain increased 2X from production configuration for tai contro Gap Transients RF noise foor - direct oop integrator off, ots of 360 and (mosty) 720 Hz. Fastest growing modes are a in cavity bandwidth. HOM growth rates consistent with expectations kicker couped beam power not unreasonabe in by2 at 2138 ma - ess than by3 at 1450 ma Issues to sort out Noise foor, gain margin. Rough anaysis of 2138 ma data suggests RF noise wi saturate feedback with a rough gain increase of 2. Study coud be done deiberatey adding noise to understand dynamics, stabiity of saturated contro 489

7 G&= -1, Gain2= Phase1 = 70. Phase2= 70. 6&ot= 370. Caib= ~~ b)evomionofmodaa..r. :.. :....:., m y :.....I f.._._. ;.-.../...:....:..... : anqs I._:- : : _... y(.i;.;. jj,... : I d)--@re-krpt) ; ; ii6 0.1 j )_... B m ;......_....i ;..... #odeno. 0 0 Time (ms) 0 0 ooo 1500 MO&No. ub) Exp. Fit to h&2&s (pod*) x 103.:..:- _,:... : :. i..: 30 I: :I i-i -10..;. 11:...j....:. 1. I -12r... :.; _j j....., i 0 YodeC 1500 PEP-II LER:oct : o= 1997mA. Dsamp= 6, -ifgain= 3, Nbun= 1744,

8 1 PEP-II LEFkoct3OOW2304: o= 2136mA, Dsamp= 6, ShifGak 3, Nbun= 1744, Gain= -1, Gain2= -1, Phase1 = 70, Phase2= 70, Btipk 691, Caib= I 1 I I I L, k O*& -....:....t... i i..t.. i i...._ %

9 N=-EnvebpesinTiM~..-_.1 f :. :.... : b)evoumofmodes anch No. 0 The (ms) c) Exp. Fe to Moth fprwbdcpt) _.j j _ I tm4jl 2 i So ;.....; i... JE c $ I... i _... d h I c _; _j_ j i i : a _..._. ~.._.._; a q)&.....;..... _i ; oo ModeNo. PEP-I LER:oct31100/1524: IO= 1-A. Dsamp= 6, *ifgain= 3, Nbn= 1744, Gain1 = -1. Gain2= 0. Phase= 70. Phase2= 70. Brkot= 370. C&b= I, I 492

10 Predicting High-current performance from ow-current measurements Growth Rates LER, HER HOM growth rates consistent with cavity impedances, gap transient Landau damping effects feedback system gain must counteract fastest growth rate - Achieved gains comfortabe for IA X 2.2A operations - but gain is infuenced by many factors Operating Gain and margins Processing channe must be unsaturated to preserve inear contro- noise on the beam imits gain Feedback contro fiter heps define noise bandwidth, phase margin - Woofer performance, RF system configuration and RF noise foor is the LFB gain imiting mechanism. Noise Foor OK for 1A X 2.2A operations when RF, Woofer optimay configured (based on A LER measurements) - LER modes -2, -3, -4 show increased motion with increasing current (direct oop?) - o/30 operations required turning off the integrator of the direct oop - we see 360 and (mosty) 720 Hz ines on the beam as a resut 493

11 Predicting High-current performance from ow-current measurements, continued Gap transients must be consistent with front-end detector range, gain variation vs. synchronous phase gap transients for typica production fis HER - approx 8 degrees RF at 600mA LER - approx 18 degrees RF at 2138mA (by2 5% gap) LER - approx 12 degrees RF at ma (by3 trains) Looks OK for 1A X 2.2A operations, for reasonabe gap structures and bucket current uniformity. Opportunities to experiment with back-fied LER gaps, carefu current vs. bucket profies. by3 pattern, ma gap transient neary fis front end ADC range with 2mA/bucket current popuations - we woud want to reduce front end signas a bit, increase DSP gain to compensate to run production in this sort of pattern and current 494

12 Averagesofbunchsignsk. -1QO I 1 I 1 I I I I oo0 a3o ' I 1 1 I I I I I um ooo bunchfwnber 495

13 , Averages of bunch sigmas syrchmous phase (reative to mfwenm OdhiW) I I 1 I Ii s -. LERin.by3.vm.... j t i.(/ t f//;'..;.; /, k 0 8 ji//// 4 I j 'I 0 -.._... i i.. [pf;.....i... f....!j #j&j A: j 1,,~~)(,,,fw' ;./,jr u,w!.i! _:.....~...~..._... et ; ;j I 1 I 1 I Ii bunchnumber I I I i Ii, 496

14 Predicting High-current performance from ow-current measurements, continued Overa timing, phasing stabiity must be consistent with kicker operating frequency and gain variation with synchronous phase Operating experience - with controed ring phasing, use of front-end phase tracking servos, not a probem 497

15 ha EizY... -&&... * 0.0s. r : Considerations of added RF stations, higher currents Growth rates - scae with impedance (scae with tota number of RF cavities) Feedback.Gain must scae with impedance for equivaent damping rate RF noise foor scaes as sqrt(number of stations) Tota gain must be consistent with noise foor adding power amps increases gain by sqrt(power) Therma management of beam ine (kicker) structures LER upgrade - add 3rd,4th,%h RF stations Growth rates increase by 2.5 noise foor increases by 1.58 HER upgrade - add 6th, 7th RF stations Growth rates increase by 1.4 noise foor increases by

16 Feedback System High-power Configurations and Overview High-Power Components 500 W 1-2 Ghz Ampifiers (up to 4 per system) 5 KW Absorptive fiters 5KW water cooed oads (3/system in tunne) 5KW oad monitor coupers Kicker Vacuum Structures (2 per ring) 1 5/8 Heiax cabe pant, 5/8 Heiax monitor pant In our configuration the maximum power imits are reached first in the cabe/connector assembies on the kicker oad ports and the water-cooed oads. Tunne oad and couper connectors being changed to 7/ 16 DIN from SC based on temperature measurements and operating experience Diode-type power detectors added to oad couper monitor signas. Signas in PEP database, usefu in evauating various fi patterns 499

17 Beam-induced kicker power Using Corett s mode the numbers for beam induced power (uniform fi) in PEP-II are: LER 3A Kicker tota Load port feed port every 2nd 4.5kW 1.75kW 0.75kW every 3rd 7.5kW 2.6kW 1.125kW every 4th 1OkW 3.5kW 1.5 kw Max amp 1KW 5oow 5oow 1400 ma by4 production fi - measured oad power (one side, in 1-4 GHz band) 1.1 kw. Tota oad power might be higher, depending on ro-off of high frequency components (see spectra) ma by3 production pattern - Tota induced kicker power (cacuated) 3kW, pus ampifier power 1.4A by4 trains of ten - tota induced power 2.8kW 2.5A by2 trains of 14 - tota induced power 6.1 kw 2SA by3 trains of 14 - tota induced power 6.1 kw 500 f

18 current (ma) 4 Ln 0 P 8

19 Transfer Impedance of two eement PEP-II Kicker I I I I I I I I F g fi data from oct3000/ Fre;juency Induced cker power I I II Cumuati$?#%%ker power Tota induced power is e+O3 watts Frequency xog 502

20 Transfer Impedance of two eement PEP-II Kicker I I I I I I I I r\ I/w x og Fre uency Induced R. cker power x og 1200 I I I I I I I I I 1000 u 800!i O Cu~uat&?~!%!ker :ower 7 a II I II I. II I It 8000 I I I I I I I I I x og I A by2 trains of 14 case 0 I I I I I Frequency x og 503

21 Possibe upgrade paths Higher growth rates require increased gain - is this gain consistent with the noise foor? Broadband and Woofer gain/phase margins? Options - add power amps and/or kickers to increase system gain - deveop Woofer equaizer in new RF? reduce RF noise contribution via RF system modifications? run narrower contro fiters, consistent with moda tune spreads and oop stabiity? Frascati-stye ongitudina kicker possiby usefu for high beam-induced power - repace drift-tube kickers above??? A? Potentia issue for added RF cavities - existing LFB processing, downsamping imit aowed maximum synchrotron frequency - Tunes for HER with 6,7 RF systems? Deveopment of new transverse digita processing system, aowing instabiity diagnostics, programmabe contro fiters? 504

22 150 Woofer gain, Vkomt _..._.: ,,. I I I I.I I

23 Summary LER LFB, Woofer ook good for 2.2A operation HER LFB, Woofer ook good for 1 A operation Predictions based on measured/extrapoated growth rates, noise foor scaing with current, MD resuts at 2.1 A Oct. 30 LER 2A operation wi imit possibe fiing patterns to those with acceptabe gap transient magnitudes. Practica use of 2/3 of theoretica 30 degrees (at 476 MHz) Operation above the HER 1 A, LER 2 A regime needs more study to understand feedback gain imits, woofer margins and therma issues from beam-induced power. We need to understand the origin of the impedances (and concomitant growth rates) for modes within the RF cavity bandwidth. We shoud (re)deveop the simuation capabiity to understand the performance of the direct/comb RF oops, and the gain/phase margins of the woofer. We may need to impement a more sophisticated woofer path in the RF processing as part of the 4A upgrade path 506 f