Bunch Length Monitor Development Facility Advisory Committee Oct 12-13 2006 Josef Frisch

Bunch Length Measurement Requirements Single shot bunch length measurement after BC1, and BC2 BC2 required later – will use the same technology Can calibrate single-shot system measurement using multi-bunch measurements Transverse cavity available for absolute bunch length measurements TCAV far downstream, want to minimize use for calibrating BC1 measurement

Bunch Length Measurement Requirements Parameter 1nC BC1 BC2 0.2nC Nominal RMS bunch Length (microns) 200 20 60 8 Coherent radiation frequency (THz) 0.25 2.5 0.8 6.0

Electro-optical Sampling Non-invasive Directly measures bunch longitudinal profile Single Shot Measurement Resolution down to ~100fsec ~200 fsec demonstrated Allows direct laser vs. beam measurements Requires high peak current LCLS BC1 current ~10X lower than at SPPS where some experiments were done. Best at short pulses (BC2) which have higher current Expensive and Complex – femtosecond laser, etc. Probably will install after BC2 David Fritz

Transverse Deflection Cavity SPPS Can measure longitudinal profile vs. energy (if spectrometer included). Single Shot Resolution <100fsec (15fsec demonstrated at TTF2 / FLASH DESY) Good Dynamic range (can adjust sweep strength) Invasive. Pulse stealing May require optics change for best resolution Expensive and complex (but uses standard SLAC supported technology) Installed before bend DL-1 and after second compressor BC-2 Primary bunch length calibration for LCLS DESY (bad pulse)

Millimeter Wave Coherent Synchrotron Radiation Single Shot (assuming single shot spectrometer, or multiple detectors) Non-Invasive Measures from arbitrarily short to ~mm bunches (with appropriate filters). Simple high rate readout – can use signal from single detector with input filter Measures power spectrum (no phase information) – cannot reconstruct bunch shape Variations on spectral response must be calibrated using external bunch length measurement – not practical to provide a calibrated signal To be installed after BC1 and BC2.

Millimeter Wave Gap Radiation Single Shot (assuming single shot spectrometer, or multiple detectors) Non-Invasive Simple high rate readout – can use signal from single detector Very simple, low cost Low noise readout <1% RMS demonstrated Diode detectors work to ~300GHz -> ~200 micron bunch length Possibly can be extended to ~ 1THz, ~70 micron bunch length Provides only relative measure of bunch length To be installed after BC1.

Optical Synchrotron Radiation Noise Measurement P. Catravas et al, Physical Review Letters 82 (1999) 5261 1.5ps. 200-500pC, 44MeV beam using a spectrometer with a resolution of 0.05nm/pixel 1.5ps 4.5ps RMS distribution measurement does not require calibration Non-invasive Not single shot Will test after BC1 Can upgrade to (near?) single shot measurement using optical spectrometer

Transverse Cavity - Injector 0.5 Meter long S-band transverse cavity 1.4 MV deflecting voltage (max) Powered from 20-5d Deflected spot (E vs. T) viewed on OTR screen in front of injector dump. Operation initially manual, developing software for automated measurement Screen configuration not designed for “pulse stealing”. Could add slightly off-axis screen, and operate TCAV slightly off of 0 phase if continuous measurement is necessary. Unfortunately, not easy to transform measured bunch lengths to BL11, BL12 bunch length monitors after bunch compressor.

Injector TCAV TCAV tested successfully at full gradient in klystron lab 3MW, 120Hz, 3us pulse 2MW, 120Hz, 3us pulse was requirement

Transverse Cavity Sector 25/29 Cavity now operating in sector 29. Will be used there for commissioning, moved to sector 25 later. 2.44M long structure, 20MV deflecting voltage Easily resolves 20 micron bunch (in simulation). 8 micron bunch (low charge case) more challenging Can change optics for measurement to have larger B function in deflection cavity (?) 5 micron bunch measured at DESY / FLASH Would like to improve optics and camera

Transverse Cavity results from ESA Experiment Simulation (including longitudinal wakes) Experiment Note: double image probably Due to camera problem

Millimeter Wave Gap Monitor Gap, with 2 pairs of millimeter wave detectors 100GHz detectors Very good intensity sensitivity Initial tuning, ~1mm bunch length sensitivity Same detectors used for End Station A run 300GHz detectors 200 micron bunch lengths for normal operation New type of detectors Detectors produce short (~100 picosecond) output pulse, set by dispersion in waveguide Amplifiers Best sensitivity with ~1KOhm input impedance amplifier Best linearity with ~50 Ohm input impedance amplifier Both types tested, will try both in actual operation. Data acquisition similar to LLRF system

Millimeter-wave gap monitor tests in End Station A Output of 100GHz detectors as phase (bunch length) is adjusted M. Woods SLAC Comparison of 2, 100GHz detectors for a range of operating conditions RMS difference 1.4% for 10,000 pulses

BL12 Gap Monitor

300 GHz diode detector

Diode Amplifiers / Electronics Diodes act as ~1K Ohm output impedance devices, ~ 1 V/mw sensitivity Maximum linear output ~50mV Very high output bandwidth ~1GHz Have tested 1KOhm input amplifier (100MHz BW), and fast 50Ohm amplifiers Choice will depend on observed signal level. Ideally would use fast gated integrator, ~200ps gate width Will start with ~30MHz bandwidth limit amplifiers Bandwidth limit to match to 100Ms/s digitizers In principal loose ~x5 in sensitivity From End station A tests, think we are OK Can add commercial (SRS 255) fast gated integrator if needed

Data Acquisition Change w to provide Pulse integral rather Than I/Q Use LLRF type digitizers Operate at 102Ms/S, 16 bit 500 samples can be processed in real time 16K samples available for diagnostics LLRF software algorithm requires only slight modification for diode signal Change w to provide Pulse integral rather Than I/Q

Gap Monitor with Pyroelectric Detectors Tested at End Station A Just put a pyro detector with RF horn next to the ceramic gap. Same detector as used for BL12 (discussed next) Sensitivity too low for single shot measurements, but should extend measurement range to <100 microns Limit not yet tested Lower performance than the mm-wave CSR monitor, but simple to commission M. Woods Pyro detector looking at gap in end station A during RF phase change

BL12 Gap Monitor Overall Status Expect system to be ready at turn-on Detector distance from gap must be adjusted Damage threshold ~100X max normal operating range, but signal level is difficult to calculate. Only see signals when bunch is fairly short (few mm) Can find shortest bunch length without calibration, but for length measurement, Need calibration – use TCAV at end of linac. 1Km of beam line away: this is probably our most serious problem Usual commissioning problems – if no signal: Long bunch? Bad alignment? Dead diode? Bad cables? Bad timing? Ceramic doesn’t transmit mm-waves?

BL11 Millimeter-wave CSR bunch length monitor Mirror with hole after bend to collect synchrotron radiation stripe Reflective optics (off-axis parabolas) to collect and transport light Beam splitting filter for high pass / low pass to 2 mm-wave detectors Different filters available Compare power on detectors for (uncalibrated) bunch length measurement Similar in concept to gap monitor, but bend and collecting optics give larger (>X10) signal, at cost of increased complexity Need higher signal for short bunch measurements where diode detector do not work

Minor modification of existing design used for screens. Vacuum chamber Minor modification of existing design used for screens. Flat mirror with hole for beam, directs mm (and optical) radiation vertically through Z-axis quartz window. No particular technical issues Chamber is in fabrication, but will not be ready for installation before turn-on Will install spool piece until ready.

Optical Path Off-axis parabolas for transport / focusing Millimeter-wave radiation awkward Too small for waveguide Too big for free space optics Alignment not critical – mm, not micron wavelength. Will be aligned with visible source Filters not yet designed In principal easy – pattern of squares, or wire grid on PC board to give low pass or high pass Commercial solutions too sophisticated, very expensive (need multiple units for flexibility) Difficult to test – millimeter wave test equipment expensive. Do not need exact performance, only stability (calibrate system using TCAV) Can be quickly installed Most significant outstanding technical issue Horns used to concentrate signal on detectors Chamber purged with dry air / nitrogen Choice is safety paperwork vs. performance

Detectors Pyro-electric detectors used as good compromise on performance / cost Wanted to avoid cryogenic detectors Large area detectors needed to collect long wavelength signals. Large area -> large capacitance ~700pf. Complicates amplifier design Approx. requirement, comparison Of 9mm and 2mm detector energy collection

Detector Amplifiers Need good charge sensitivity at high capacitance (~700pf) Similar to physics detector amplifiers Have simple op-amp based amplifier with ~3000 electrons RMS noise Expect ~1000 electrons noise with external FET front end Expect ~ 1uJ X 1uC/J = 1pC charge on detector max ~ 107 electrons. Easiest to make low charge noise amplifier with slow output ~10-100 microseconds

Will use same digitizer as LLRF Data Acquisition Will use same digitizer as LLRF May need to run digitizer at ~10Ms/s to record long waveforms (up to 50usec) Have synchronous divider for clock Low frequency bandwidth of digitizer too low (~1MHz) Use detector signal to modulate RF tone Ugly, but simple, and avoids need for new hardware or software.

BL11 Millimeter Wave CSR monitor, overall status Vacuum chamber slightly delayed, will probably not affect commissioning Electronics / data acquisition not finished, but no significant problems foreseen Filters need to be designed! Similar issues to gap monitor for calibration – need downstream TCAV. Largest operational issue

Optical Synchrotron Radiation Correlation Measurement Plans to add optical detector with narrow band filter to BL11 bunch length monitor Optionally can add single shot optical spectrometer New plan – no real schedule yet, but might be possible to install quickly Would provide calibrated measurement of RMS bunch length (but not shape).

BC1 vs. BC2 systems BC2 will have TCAV available for calibration, or for continuous measurement in pulse stealing mode Assuming BC1 mm-wave monitors are successful, will use nearly same design for BC2 Gap monitor probably not useful for short bunches, but may still be a useful tuning tool CSR monitor easier for short bunches since shorter wavelengths are easier to focus. Hope to have operational electro-optical bunch length monitor at BC2 If possible might then install similar system after BC1 for calibrated measurement Optical spectrum fluctuation measurement may be useful in both locations.