Joint Cryogenic Instrumentation & Slow Controls meeting 18/04/2018 Calibration of temperature sensors for ProtoDUNE-SP M. Antonova, A. Cervera, P . Fernández, M. García, A. Izmaylov, J. Llusar, M. Sorel, P . Novella, P . Bernabeu, J.V. Civera, P . León IFIC-Valencia X. Pons CERN
T-gradient monitor 175 cm • High precision (<5 mk) vertical temperature profiling LAr level with sensors cross-calibrated in the lab • 48 sensors separated 11.8 cm at top and bottom, 9 m and 23.6 cm in the middle area 7.5 m bottom cup 2 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Calibration setup • Almost converged after many different setups, including a profesional pressure vessel • Home made cryostat: a polystyrene box with 12 cm thick walls and a 3D printed PLA box with two independent volumes • Sensors are put inside a cylindrical aluminum capsule to minimise convection and avoid thermal shocks PLA Box PLA Box Aluminum capsule with 4 sensors inside Aluminum capsule inside PLA Box made in a 3D printer inside of the polystyrene box 35 cm 12 cm 5 cm 10 cm 3 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Sensor support and cable • Sensors are mounted on a PCB with an IDC-4 connector • 4-wires readout to avoid lead resistance and effect of connectors • Four sensors are put together inside the Al capsule using a PLA support to hold the sensors and keep them at a fix distance in the middle of the capsule PT102 with IDC-4 connector PLA support for 4 sensors and cables Put 4 sensors as close as possible ~5 cm 1.4 cm 4 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Readout • An accurate current source for the excitation of the temperature sensors, implemented by a compact electronic circuit using high a precision voltage reference from Texas Instruments. • A multiplexing circuit based on an ADG707 Analog Device multiplexer electronic device • A high resolution and accuracy voltage signal readout module based on National Instruments NI9238, which has 24 bits resolution over 1 V range • The current mode of operation averages over 2000 samples taken every second for each sensor 5 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Readout current source for MUX controller ADC MUX controller PT100 current source and MUX card 6 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Calibration strategy • Put several sensors as close as possible in LAr such that we can assume they are at the same temperature • Compute the temperature offset between them and refer those offsets to a single reference value • For the sensors calibrated at once (a “set”) we can compute all possible combination of offsets • But obviously the 48 sensors cannot be put together so: • either we use a single sensor as reference for all sets • or we do different sets and then correlate them cross-calibrating a sensor from each set • In fact we will do both, such that we can cross-check (next slide) 7 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Calibration strategy 1st round • Use a sensor as reference. keep that sensor in all sets 2nd round 1 2 3 3rd round • Thus, three sensors are calibrated in each set 4 5 6 7 8 9 10 11 • 3 calibration rounds 12 13 14 15 • In the first round we calibrate 18 sets of 3 sensors 16 17 18 19 each (4 measurements for each set) 20 21 22 • In 2nd round we cross-calibrate the middle sensors 23 24 25 26 of the 1st round (5 sets, 4 meas each) 27 28 • In 3rd round we cross-calibrate the middle sensors 29 30 31 32 of the 2nd round (1 set, 4 meas) 33 34 35 • This is a total of 18x4+5x4+1= 93 measurements 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 8 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Single set 1. Cool down the cryostat 2. Put four sensors in the capsule (3 + reference) 3. Close the capsule and cool it down slowly until it reached 98 K 4. Submerge it into LAr and put the PLA cover 5. Fill the cryostat with LAr 6. Record temperatures for about 40’ 7. Take out the capsule and wait until it reaches 200 K 8. Repeat steps 4-7 9 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Results • Dispersion of the measurements is below 2 mk for the middle sensor (closer to 1 4 reference) and below 4 mk for outer sensors 2 3 (ref) • Reproducibility, defined as the RMS of the mean offset between in the stable region (1000-2000 s) is below 1 mk for most sensors 1 2 4 stable region 1000-2000 s 10 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
2nd and 3rd sets 11 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
4th and 5th sets 12 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
6th and 7th sets 13 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
8th and 9th sets 14 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
mean All sets together sensor ID offset reproducibility 39616 = 9.98 +- 0.43 15 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Stability over time • Check the same sensor (middle one) after a month • NO changes (< 1 mk) after a month • Need more checks like this March 22nd + April 17th April 17th March 22nd 16 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Calibration strategy 1st round • In green sets already calibrated 2nd round 1 2 3 3rd round • We will do the second round for the first 8 sets in the 4 5 6 7 next few days 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 17 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Conclusions • We believe ~1 mk reproducibility can be achieved • Steel need to understand variations of behaviour from one set to another • Improve some elements of the calibration setup, with better insulation of the sensors • Monitor pressure and ambient temperature • There is still room for improvement: • Improve cryostat • Modify the sensor PCB such that all 4 sensors can be put even closer • Amplify the voltage signal such that we use the full range of the ADC. Currently using 20 mV, having a range of 1 V 18 Anselmo Cervera Villanueva, IFIC (UV-CSIC)—Valencia
Recommend
More recommend
