Instrumentation for the UCNB experiment
Transcription
Instrumentation for the UCNB experiment
Instrumentation for the UCNB experiment Sky Sjue December 13, 2011 Requirements – protons Maximum energy O(800 eV) so we need: 0.0016 a=0 a=-0.1 0.0014 bias voltage I minimal noise – electronics I minimal noise – cryogenic cooling 0.0012 Probability per eV I 0.001 0.0008 0.0006 0.0004 0.0002 0 0 100 200 a= 300 400 Ep [eV] 1 − λ2 1 + 3λ2 500 600 700 800 Requirements – electrons Maximum energy O(800 keV) so we need: 0.0025 I thick detectors I position resolution I timing resolution Probability per keV 0.002 0.0015 0.001 0.0005 0 0 100 200 300 400 500 Ee [keV] 600 700 800 Requirements – coincidence Coincidence is necessary to suppress backgrounds and possible systematic effects like electron backscattering. I segmentation, multipixel instrumentation I timing resolution I data acquisition dσ dΩ ∝ R Ze 2 2µv 2 sin2 (θ/2) 2 Large, thick segmented Si detectors Particle view: I I I D=15 cm p-type implant contact to minimize dead layer metallized Al mesh (visible) for faster response Segmented Si detector Hexagonal array with 128 pixels, area of a pixel is ∼ 0.8 cm2 Single pixel characterization – α 148 Gd α source, low energy peak comes from energy loss of an α in the Al mesh. Leah Broussard Single pixel characterization – β Data 2500 ∆N 100 0 -100 2000 Data Counts 1500 Simulation 1000 500 0 0.5 109 Cd 1 1.5 2 2.5 3 Voltage (V) 3.5 4 4.5 5 source, x rays at ∼ 23 keV, Auger electrons at 63 and 84 keV, PENELOPE simulation versus data Robert Pattie Single pixel characterization – low energy protons 1 Ep=26 keV Ep=30 keV Ep=34 keV Ep=38 keV Scaled counts 0.8 0.6 0.4 0.2 0 0 10 20 30 40 50 60 70 80 90 100 Channel Proton spectra taken at TUNL with accelerator built by Seth Hoedl Proton data – simulations and fits 10000 10000 1000 1000 Counts Counts Simultaneous fit of of simulations to four proton spectra 100 10 100 10 1 1 0 20 40 60 Channel 80 100 0 20 40 60 Channel 80 I Fit of simulations to Ep = 26 keV and Ep = 34 keV protons for fixed calibration with a dead layer of 80 nm I Fit FWHM=2.4 keV 100 Single pixel characterization – dead layer from protons Channels per keV 2.65 197 185 156 2.6 2.55 2.5 2.45 2.4 74 I I I 76 78 80 82 Dead layer [nm] 84 86 Global best fit with a dead layer of 80 nm yields χ2 = 155.6 for ν = 158 (P=46%). Result is 75 ≤dead layer≤ 84 nm at 99% confidence level Specification was∼100 nm, this exceeds requirements FET on detector for low noise and no pickup 50 Ω 0.1µf CF 1pf BF862 RF 1000M Ω C1 5pf P$4 P$3 B A P$2 P$1 D C IN AGND AGND 2.5V C2 50 Ω TP3 IC1 −5V R2 499 4 6 3+ C16 0.02 µf 7 +5V AD8011 2 − IC1 R1 249 B A P$2 P$1 P$4 P$3 D C +5VF AGND C3 10µf 0.1µf AGND AGND Amplifiers designed by Pat McGaughey and Jacqueline Mirabal (LANL P-25) perform first stage of amplification with FET at detector to avoid pickup before further amplification and data acquisition 8-FET card for 8 pixels toward instrumentation of 128 pixels Card with 8 FETs for 8 pixels to mount directly on detector (in testing box). Tail pulse from a function generator put through an attenuator on the left, feedback on the bottom, output on the right. We have four of these in hand right now. Leakage current frozen by cryogenic cooling 100 Leakage current [nA] 10 1 0.1 0.01 0.001 Measurement 3/2 I∝T exp(-0.543eV/kT) Amplifier requirement 0.0001 140 160 180 200 220 240 260 280 300 T[K] Leakage current monitored as a function of detector temperature, effectively zero below T = 180 K. Spectrum with FET and Si detector at 137 K 100 background best fit 90 80 Counts per 335 eV 70 60 50 40 30 20 10 0 20 30 40 50 60 70 Eγ [keV] I 139 Ce I I I spectrum, Ex (Kα) ≈ 33 keV,Ex (Kβ) ≈ 38 keV Fit gives FWHM = 3.1 keV, but in UCN experimental area with 50 ft BNC cables to ADC, no shielding, beam ON. ∆t ≈ 24 hr Detector mount – initial design Bias apparatus – initial design Detector is shielded by a shroud, which is also at voltage. Cooling apparatus – initial design This design, slightly modified, will work for an initial run (as we have demonstrated). High voltage tests Initial bias tests performed to 5 kV, discharges monitored on a scope with a 500 pf capacitor. High voltage tests with flowing LN2 No indication of any problems associated with flowing LN2 at 5 kV. Discharges appeared smaller if anything. Summary I Demonstrated ability to measure protons with energy corresponding to reasonable bias voltage I Electronics in hand for multipixel instrumentation with noise and timing capability sufficient for experimental requirements I Bias tests, R&D in progress I Next we will work on operating detectors at bias, at bias in a magnetic field, at high bias in a magnetic field with all the pixels, measuring electron-proton coincidences . . .