PREx/HAPPEx Collaboration Meeting Minutes - December 7 and 8, 2007

Attending: D. Armstrong, G. Cates, K. de Jager, P. Decowski, G. Franklin, B. Hahn, T. Jinasundera, K. Kumar, D. McNulty, L. Mercado, B. Michaels, S. Nanda, D. Parno, K. Paschke, P. Souder, A. Rakhman, R. Snyder, R. Wilson By phone: none

(Minutes by Dustin McNulty)

Summary of Meeting Discussions (Friday):

• Target change after CSR experiment has been extended from 3 to 5 days and may require additional time depending on the level of Pb target radioactivity. During this time, the hall will be open for us to work.

• At best, the PREx test will start on Sat. Jan 26, 2008.

• The lab may extend the current Accelerator run by 11 days (up to Feb. 11, 2008, per hall B's formal request). Potential chance here for extra beamtest days.

• In the Feb/March down, we have permission to inspect the injector setup.

• After our beamtest, we must remove anything we install in the L-HRS focal plane (due to NaI detectors).

• PREx Goal for upgraded Compton Polarimeter: Measure longitudinal Compton asymmetry at the level of ~3% uncertainty and provide electron beam polarimetry with 1 to 2% uncertainty.

• To achieve goals: Upgrade to green laser/cavity, replace PbWO4 array with single GSO crystal, and use flash ADCs to sum the signals for each helicity window.

• New GSO crystal (6cm diam X 15cm length, about 12Xo) to arrive at CMU within the month. Currently they are using a 1 cm cube sample from the manufacturer (Hitachi Chemical) for response function studies using a Sodium-22 source.

• They have a working DAQ with FADC and are investigating the pedestal noise levels seen using the signal integration technique.

• Developed simple GEANT simulation for studying the GSO response to mono-energetic photons.

• Will continue to add more realistic elements to Monte Carlo in order to investigate and finalize detector design.

• Additional comments and discussions: PREx will run at 850MeV, so Compton photon radius (6 meters downstream of interaction point) at the calorimeter location is ~3cm. "Resolution doesn't matter here, its linearity that's important" (Paul). "Maybe we need to worry about photons hitting the pmt photo-cathode?" (Kent). We would like to test the true response function at HIgS.

• Lumi detectors are being modified to employ a selectable Neutral Density (ND) filter between the light-guide and the PMT in order to achieve the desired lightlevel for optimal detector performance.

• Filter box design is complete and being constructed by William & Mary machine shop. First box to be completed before end of month, the rest before beamtest.

• Light-guide material has been ordered/received and new light-guides will be fabricated by W & M machine shop.

• Commercial ND filters (2 x 2 in) can be custom cut into 1 in squares and mounted in an Aluminum frame (to be inserted into the filter box). W & M machine shop will also construct these. Need to order remaining filters. Need to investigate method of stacking filters to achieve more flexibility in attenuating lightlevels.

• Tests are ongoing in the counting house test cage to understand the linearity of the Lumi PMTs and to determine the optimal lightlevel, PMT HV, and ADC gain settings needed to achieve minimal differential non-linearity (Beta*No).

• Will characterize all 8 Lumi PMTs (gain curves and linearity).

• Currently testing 16bit ADC, will soon test 18bit ADC.

• Tests indicate that we want lightlevels around 20nA and PMT anode current less than or equal to 50 microAmps.

• Currently pedestal noise is at the 100ppm level (10 count pedestal widths with 50,000 count signals). Working to reduce this noise by an order of magnitude.

• Making steady progress toward achieving 18,000 finesse cavity resonance.

• Currently working toward locking Prometheus green (100mW) laser in high-finesse Fabry Perot Cavity. Reprofiled Prometheus laser.

• Achieved mode matching of LGR cavity mirrors.

• Using OptoCAD for complete ray tracing of the optics setup.

• Using Pound Drever Hall (PDH) locking scheme with fast (PZT on laser mirrors) and slow (temp change on PPkPT crystal) feedback loops to tune phase of the cavity resonance.

• Short duration lock has been achieved. Cavity feedback loop tuning in progress to increase locking duration.

• Additional comments and discussions: "There is less than 1% pulse to pulse cavity power fluctuation at 30Hz" (Sirish). "We may need to monitor these fluctuations" (KK). "Measuring total power in cavity does not necessarily give you the luminosity because power fluctuations can cause spatial distortions in the beam shape" (Sirish). "The green cavity has a sharper resonance and is much harder to lock than the IR cavity (there are 12 parameters to optimize for the green cavity)" (Sirish). "If don't see TEM00 mode during the experiment, then we will need to do more optical tuning" (Sirish).

• Developing high power green laser via Second Harmonic Generation with PPLN crystal. This method will provide the necessary green laser power to achieve the ~1.5kW intra-cavity power required for the upgraded system.

• PPLN frequency doubling requires temperature stability at 0.05 degree Celsius level. Have achieved this stability using a Thermal Electric Cooler (TEC). Currently achieving 0.02mW green power stability.

• Have achieved 17%/Watt conversion efficiency for PPLN doubling.

• Currently using lightguide IR laser + PPLN to produce ~50mW of green laser power. Plan is to eventually use fiber amplifier to amplify IR power and achieve ~5000mW of green laser power.

• Preparing setup for coupling IR laser into the high power fiber amplifier.

• New electron detector status: Delivery of repaired detector and electronics on Jan 15, 2008. Installation of new support structure/motion system starts on Feb 8, 2008.

• High power fiber laser raises big safety issues (at 10W, you could cut your finger off).

• There are three sets of cavity mirrors that could be used: REO (Gain = 15000), LGR (Gain = 6000), CVI (Gain = 2000). Better mirrors (higher Gain) are harder to lock. Currently in hall A, IR system has Gain = 5000.

• There are 4 possible ways to produce green laser light for the new cavity: lightwave + PPLN (50mW), Prometheus (100mW), Prometheus + PPLN (500mW), and lightwave + fiber Amplifier + PPLN (5000mW). One of these lasers combined with one of the mirror sets will provide the 1.5kW of intra-cavity power.

• Schedule: Immediately following Jan beamtest, install new electron detector and commission it during the next experiment (threshold pi0 production). Install new cavity between April and June 2008. Install new photon detector in the Fall.

• Test concept for integrating Compton Photon detector.

• Test ability to run 50 microAmp on Lead target.

• Test linearity and noise sensitivity of Lumi and detectors.

• Source work needed both before and after Jan beamtest.

• Additional discussions: Pass 2,3, and 4 give purely longitudinal beam polarization. Pass 1, and 5 have a decent amount of transverse beam polarization (30 - 40% ?). "Can we run with same pass as hall C (with hall C at high current) for cross-stalk tests?" (Kent).

• Goals for the beamtest: Measure noise when integrating (15Hz pulse to pulse difference). Evaluate noise due to big Bremsstrahlung pulses. Evaluate other sources of noise. Measure an asymmetry (measure bkgds carefully since we don't have an electron detector).

• Preparations for beamtest: Need to mount test detector/12 stage pmt. Need accumulator software. Need Compton DAQ to handle FADC signal integration technique.

• Compton hardware during test: Get IR laser cavity working. Tune beam through Compton chicane with old photon detector array. Test 12 stage PMT using single PbWO4 crystal borrowed from PrimEx collaboration. During 12 stage PMT detector checkout, monitor bkgd with old array.

• No formal presentation was given, but here are some discussions: "We may want to install an LED flasher (fiber optic) into the 12 stage PMT test detector (to facilitate easier checkout)" (Bob). "We will turn on IR cavity before Xmas down; if cavity is not tuned very well, we will use 700mW laser (feeding PPLN crystal in lab now) to boost the cavity power without solving cavity tune problems" (Sirish).

• Test compatibility with PREx and concurrent high current experiment in hall C: Check cathode interaction (observed non-linearity of intensity vs. laser power), and problems associated with quantum efficiency hole.

• Establish near parity-quality source configuration to support Lumi tests.

• Look into pockels cell HV switch (for high rate flipping, moves much less current).

• Schedule: Dec 22 - Jan 3, 2008--parity quality source setup and establishment of baseline. Jan 3 - Jan 30, 2008--during any accelerator down, attempt RHWP scans, pita voltage scans, etc. After January beamtest--perform "two-laser" tests (to help understand hall C cross talk).

• Center beam on target (with 4mm raster); use a holey target (with 1mm hole, for example) or one of the tilted Lead targets.

• Measure rates in the scintillators with the vertical drift chambers (VDCs) off.

• Establish rates up to 500nA (do not exceed 10MHz).

• Protect HRS detectors (if necessary detune HRS magnets).

• Ramp beam up slowly to 100 microAmps (in 10 microAmp steps, pausing 20min at each step).

• Monitor rates in scintillator and Lumis, target temperatures, vacuum pressures, and radiation monitors.

• New cavity receiver boards have been installed in hall A.

• Need to checkout cavity signals using HAPPEx DAQ in December during CSR experiment.

• During January beamtest, we need to: Perform X and Y beam scans at high and low current with raster on and off, calibrate steering coils, measure current between 50 and 1000nA, and deduce resolutions and absolute accuracy.

From Piotr (detector tests):

• Optimal configuration for thin detector: 0.5cm thick quartz block with 4cm X 4cm upper surface and 4.5cm X 4cm lower surface, and 6cm long cone mirror, 2in diameter PMT with 2mm thick glass. Simulation results show ~40 Cerenkov photons (on average) reach the PMT for each 850MeV incident electron, providing ~17% resolution.

• Optimal configuration for stacked detector: 10 trapezoidal blocks identical to above thin detector block interleaved with 1.7mm thick plates of Tungsten. Using 2in PMT with 2mm thick glass, simulation results show ~120 Cerenkov photons per shower (on average) reach the PMT from the stack. Note that simulation also shows 1 to 2 stray electrons hit the photo-cathode per shower -- need to understand how many photo-electrons these strays produce (a few is OK, but ~100 is a problem).

• Goals for Jan beamtest: From measured detector output and known PMT gain, determine number of Cerenkov photons and compare with simulation results. Acquire detector spectra and resolution in counting and integrating modes. Study PMT signal backgrounds and noise.

• Activities for Jan beamtest: Observe PMT signals in counting house (need 2 cables), change quartz block in thin detector from 5mm to 10mm, and perform blinded PMT tests. For counting mode, HRS at 20 degrees, expect 100Hz rate. For integrating mode, HRS at 12.5 degrees, expect 50kHz rate. Need 1 pass beam for these studies.

From Krishna (Lumi detector tests):

• Optimize light input and PMT current output: Set ND filters to give 10nA lightlevel, look at some PMT responses in counting room, set appropriate PMT HV settings to limit anode current to 50 microAmp.

• Ensure PMT responses are dominated by high energy scatters from the target: Blinded tube tests (shielding needed?), examine dithering coefficients vs. PMT number.

• Study noise properties: Minimize pedestal noise, examine widths vs. flip rate, different oversampling, different beam currents (10 and 50 microAmps), and different targets.

• Lumis will be useful for measuring transverse asymmetry. Will want 1 pass beam for this.

There was no formal presentation for this item. Plans for "Width" studies are given in item 15 above.

• This discussion was deferred to Saturday morning.

• Whiteboard discussion

• Warm septum for PREx. $ coming from JLab, but no design time granted yet.

• Get design and have constructed by outside machine shop. (Take ~1 year to finish)

• Target design final for PREx and e06007. It will hold 4 isotopically pure Pb-208 targets (3 are tilted for e06007). Pb targets ($5000 each) are sandwiched between diamond foils and passively cooled by the loop 3 cryo target (liquid H2 at 19K). The target ladder has room for 4 additional "standard" targets.

• Need to decide what "standard" targets we want: Thin Tantalum, thick Carbon-12 (how thick?), 3 X 3mm holey target, and BeO (for spot viewing).

• Currently we have two 18-bit boards; 4 new boards (with updated design) are due/expected Jan15, 2008--however, Bob gives 50/50 chance that we will receive them in time for beamtest.

• We may need to borrow some of hall C's 18-bit ADCs to cover all our signals (8 Lumi, 2 dets, ? striplines, 6 cavity). We can test hall C boards in EEL 122 (Bob has working CODA test crate with a few empty slots).

• Monte Carlo of Compton photon spectra and detector has been developed. Effects of synchrotron shield and accidentals (in the form of pileup) have been included.

• Discrepancy exists between Monte Carlo and real data; fitting procedure has an energy dependent systematic associated with it. (With no accidentals it works, but with accidentals result doesn't work).

• Hardware studies of detectors true response function (especially low energy tail) are important. Can we test at HIgS?

• Still a work in progress.

• Not clear if we will have this ready for Jan beamtest.

• Even if we do not flip helicity at greater than 30Hz, we could ~simulate rapid flipping by oversampling within the usual 33ms window.

• When we oversample, we will not have the injector (V to F) information.

• "We need to get used to understanding BCMs, pedestals, and analysis of rapid flip data" (KK).

• For PREx detectors: Find S_0 (scintillator paddle), check it out (set trigger and timing,...). Talk to Jack Segal about frame and mounting, and cabling (HV, signals, and patch panel to counting room).

• For BCM checkout: Need to take data during CSR experiment, compare new cavity signal's correspondence with existing striplines. Measure pedestal noise.

• For Compton FADC: Setup in hall--FADC, HAPPEX timer, in/out scaler. Note that injector helicity information is currently "in-time" by default (we may not be able to change this to "delayed"). Need to do LED tests (need negative HV on PMT for integrating). There are 6 new FADCs. "Make sure we have control signals/gates (dither on/off, laser on/off, etc.) in real-time data stream" (Kent),

• For Compton DAQ: Still much work to be done on Compton DAQ. We cannot currently make modifications to DAQ because it uses a custom version of CODA.

• Compton laser/cavity checkout: Sirish and Dustin will attempt to lock IR cavity before December down.

• For targets: Test target assembly techniques using cheaper (chemically pure) Pb targets. Need ESAD and RSAD (Bob will make these). Be careful that we do not exceed hall's radiation budget during our long-run target tests.

• For Lumi: Need to remove light-guides from beamline, remove quartz detectors, and install them in new light-guides (follow radiation safety procedures). Install first filter box assembly during December down. Install all 8 Lumis after PMTs have been characterized and additional filter boxes are ready. Minimize pedestal noise in the hall.

• Online data analysis: Need to consider how analysis will work with oversampling (and/or fast flipping) and "in-time" helicity information.