Charge readout and dual-phase readout technology for large scale cryogenic liquid detectors

Charge readout in liquid argon

Liquid argon Time Projection Chambers (LAr TPC) exploit liquid argon as interaction and detection medium. Similarly as in gas TPC, a fully homogeneous detector can be built by enclosing the active LAr volume in a field shaping cage, which defines a uniform electric field of typically 0.5–1 kV/cm. This field is used in order to drift the electrons from the ionization produced in LAr by charged particles to a segmented anode. The clouds of electrons, produced by charged particles along their trajectories, keep while drifting their original shape, being affected by small diffusion effects. Two orthogonal coordinates (or more than two coordinates at different angles: typically one collection view and two induction views) can be reconstructed by reading independent views on the anode plane. The measurement of the drift time, corresponding to the travel path of the electrons at constant speed from the production point to the anode, allows for the reconstruction of the coordinate orthogonal to the anode plane. The detector can then provide an accurate 3D imaging of the original distribution of the ionization, similarly as in bubble chambers. LAr TPC provide, at the same time of tracking and particle identification capabilities excellent performance in electromagnetic and hadronic calorimetry.

The original detector concept was pioneered by C. Rubbia in 1977 (CERN EP-INT 77-8). LAr does not attach the electrons produced by the energy losses, however residual electronegative impurities, such as oxygen molecules, which would absorb the electrons must be removed from the liquid in order to avoid the absorption of the electrons along their drift path.

Several physics effects impact the behavior of the cloud of electrons produced by ionization:

  • Recombination. The fraction R of the produced ionization charge which does not recombine depends on the applied electric field E and on the density of the initial ionization. This dependence is usually expressed as a function of the mean energy losses dE/dx for the ionizing particle. Several models have been developed on the local recombination of the electrons and the ions occurring immediately after ionization but none of those describes all the experimental data in liquid argon. The Onsager theory is based on the concept of “initial recombination” of the electron-ion pairs. The limit of this model is that it does not take into account the dependence on the ionization density: it explicitly assumes independent single electron-ion pairs. A better description of the measured data is given by Jaffé: the assumption of this model is that the initial ionization charge is distributed in a column around the trajectory of the ionizing particle. Other models are based on the columnar theory, with different assumptions. All these models assume a direct proportionality between the electron drift velocity and the applied electric field: this approximation is not true in liquid argon electric fields higher than 100 V/cm. Usually, Jaffé formulation is approximated by the so called Birks law: A/(1 + k/(E * dE/dx)), where E is the electric field intensity (kV/cm), dE/dx are the ionization losses and the coefficients values are A = 0.800±0.003 and k = 0.0486±0.0006 (kV/cm)(g/cm2)/MeV, as measured by the ICARUS collaboration (see here).
  • Drift. The accuracy of the drift coordinate depends on the measurement of the drift time, i.e. the time between the absolute time of an event and the time when the drifted electrons reach the readout plane, and on the accuracy on the knowledge of the drift velocity. The drift velocity depends primarily on the electric field E; the temperature of liquid argon can influence at a lower level the drift velocity. In addition, the presence of molecules in the liquid, such as carbon hydroxides, can increase the drift velocity. On the other side, electronegative impurities like oxygen do not change the drift velocity, but reduce the number of drifting electrons. A fundamental requirement for the performance of a LAr TPC is to collect a maximum number of ionisation electrons since the output signal depends on the quantity of charge deposited on the readout planes. Thus, the purity of the medium is crucial in order to limit losses. Possible contamination has to be kept at an extremely low level. The ICARUS collaboration provided measurements (see here) of the drift velocity for electric fields as low as 60 V/cm and up to 1 kV/cm at T=89 K. The drift velocity of electrons increases less than linearly for electric fields above 100 V/cm, still there is considerable gain at higher fields. It reaches 1.55 mm/μs at 0.5 kV/cm for a temperature T=89 K, and it increases by 30% by doubling the electric field intensity from 0.5 to 1 kV/cm, and again by 30% from 1 to 2 kV/cm.
  • Diffusion. A point-like charge deposition will spread during the drift due to diffusion effects. This is particularly important for long drifts because it directly affects the accuracy of the drift time measurement and the transverse smearing of the reconstructed tracks. The magnitude of the charge spread depends on the drift time. The size of the electron cloud arriving at the anode in the longitudinal (along the drift direction) and transverse direction follows a bi-dimensional Gaussian distribution. For gases, the diffusion coefficient in the longitudinal direction differs from that in the transverse direction. For argon gas the longitudinal coefficient is substantially lower than transverse one. The ICARUS collaboration measured (see here) the longitudinal diffusion in liquid argon with a drift length of 1.5 m, giving as result D = 4.74 cm^2/s.

If the anode is immersed in the liquid, it will just collect the electrons produced by the track ionization (typically 18000 electrons every 3 mm of path of a particle at the ionization minimum). This signal has then to be amplified by low-noise electronics in order to get suitable signals, which can then be digitized. The typical S/N ratio is about 10:1. This technique has been traditionally used in LAr TPC detectors by several experiments and it can be called single-phase LAr TPC readout.

However, if the anode is located in the gas, the electrons may be amplified, as in more traditional gas detectors, by exploiting avalanches occurring in regions with high electric fields. The electrons can the be collected by the anode and the corresponding electric signals are amplified by the readout electronics. In this case the LAr TPC readout technique is called dual-phase. A higher S/N ratio (typically 100:1) can be achieved thanks to the multiplication of the electrons in the gas. The avalanches occur in confined volumes where the electric field reaches of the order of 30 kV/cm, like inside the holes of the Large Electron Multipliers (LEM) also known as Thick-GEM (Gas Electron Multipliers).

The general principles of the single-phase and dual-phase readout schemes are illustrated in the following picture. More information about the liquid argon detectors can be found here.

Single-phase and dual-phase liquid argon TPC readout

R&D activities in the framework of AIDA2020 WP8

WP8 fosters knowledge sharing and common tools in the neutrino community as regards state-of-the-art in very large cryogenic liquid detectors. The construction of liquid argon detectors at the 10 kton scale is an essential ingredient of the future international long-baseline neutrino program unifying the European and USA efforts. The DUNE detector, whose final design consists of four large LAr TPCs, each of them with a 10 kton mass, aims at precision measurements of the neutrino oscillation parameters, the study of atmospheric and solar and supernova neutrinos, as well as pushing the limit of proton decays for a variety of final states.

WP8 activities focus on the most challenging aspects related to this detector development. One of these aspects, studied within the networking activity of Task 8.3, concerns the large size scaling of the charge readout techniques including the assessment of the dual-phase techniques for charge amplification with electrons extraction from the liquid phase and amplification in micro-pattern gas detectors in absence of quenching, as well as the development of charge readout front-end cryogenic electronics and large-scale digitization systems

WP8 exploits common infrastructures for the R&D activities at CERN supported by the CERN Neutrino Platform:

  • the 3x1x1 m^3 dual-phase pilot detector
  • the 6x6x6 m^3 dual-phase demonstrator (protoDUNE-DP)
This task involves many institutes across Europe. In addition, several US groups are also strongly involved and interested by these activities via the common project DUNE.

The main aspects of the WP8 networking activity concerning the Charge Readout task include:

  • Reviewing charge readout techniques for single and dual phase readout;
  • Development of LEM detectors which include the LEM production/cleaning/tests procedures and the LEM integration on large DP readout surfaces;
  • Development of dual-phase accessible cryogenic front-end electronics;
  • Development of high bandwidth DAQ systems for giant LAr detectors;
  • Development of synchronization systems for giant LAr detector.
These activities have also synergies with other AIDA-2020 WPs such as the one on Innovative gas detectors (WP13) and the one Advanced Software (WP3) respectively on the developments on the LEM detectors and the LAr TPC reconstruction with the Pandora software.

Following the successful operation of the 3x1x1 m^3 detector, the charge readout deliverable was extended to M54 in order to include as well results from the preparation and operation of ProtoDUNE dual-phase.

The DUNE experiment

The Deep Underground Neutrino Experiment (DUNE) is an international experiment for neutrino science and proton decay studies. The DUNE experiment includes a precision near detector at the edge of the Fermilab site, in Batavia, Illinois, and a very large, modular far detector about 1.5 km underground at SURF in Lead, South Dakota, 1300 km from Fermilab.


The DUNE far detector will consist of four similar liquid TPCs, each with fiducial mass of at least 10 kt, installed about 1.5 km underground. Each detector will be installed in a cryostat with internal dimensions 14.0 m (W) × 14.1 m (H) × 62.0 m (L), and will contain a total LAr mass of about 17.5 kt. The LAr TPC technology provides excellent tracking and calorimetry performance, making it an ideal choice for the DUNE far detectors. The four identically sized cryostats give flexibility for staging and evolution of the LAr TPC technology. DUNE is planning for the single-phase (described here) and the dual-phase technology (described here).

The primary science program of DUNE focuses on fundamental open issues in neutrino and astro-particle physics:

  • Precision measurements of the parameters that govern νμ → νe oscillation: measuring the charge-parity violation, determining the neutrino mass ordering and precision tests of the three flavour neutrino oscillation paradigm;
  • Search for proton decay in several important decay modes. The observation of proton decay would represent a ground-breaking discovery in physics, providing a portal to Grand Unification of the forces;
  • Detection and measurement of the flux from core-collapse supernovae within our galaxy

Single-phase LAr TPC readout

In a single-phase TPC the electrons produced by ionization are drifted towards finely segmented anode wire planes by an intense electric field. Several wire-planes with different orientations using bias voltages, chosen for optimal field shaping, give several complimentary views of the same interaction as a function of drift time, providing the necessary information for reconstructing a three-dimensional image of the interaction. For a detector with 3 anode wire planes, the inner two planes are called induction planes, allowing drifted electrons to pass through them in a almost transparent way (see slide 3 in this presentation), inducing bipolar signals that are used for event reconstruction. The outer plane is called "collection plane" because the drifted electrons are really collected on these wires, producing a greater unipolar signal. The figure below shows a schematic representation of the signal response in a single-phase TPC with two induction planes (U and V) and the collection plane (Y).


The wires used in LAr TPC detectors have larger diameters than in gas detectors (typically 150 um diameter stainless-steel or copper-beriullium wires) since there is no need to have high electric field regions around the wires for amplification and are therefore much less fragile. Bipolar signals provided by the induction views may cancel out for very close tracks, as it may happen in large occupancy events such as electromagnetic showers.

The single-phase TPC technology has been largely demonstrated by several experiments. The largest LAr TPC ever operated is ICARUS. Filled with 760 tons of liquid Argon, ICARUS was installed at the Laboratori Nazionali del Gran Sasso (LNGS) in order to study cosmic or atmospheric neutrino and neutrino coming from the CNGS beam. The read-out chambers (two LAr TPC for each half-vessel) are mounted on the internal walls with the cathode at the center, to maximize the LAr sensitive volume (corresponding to about 480 ton in mass). The read-out chamber scheme consists of three parallel planes of wires (horizontal, +60 and -60 degrees) with a surface of 17.95 x 3.16 m^2. The wire pitch is 3 mm for all the views. ICARUS started operations in 2010 and it took data until 2012. In 2017 the ICARUS detector was moved to Fermilab, after refurbishment, for a new short-baseline neutrino experiment. The ArgoNeuT experiment was small liquid argon TPC installed at Fermilab to detect neutrinos on the NuMI beam-line. It operated 2009 and 2010.

An active experiment which uses a single-phase liquid argon TPC is MicroBooNE. It performs a short baseline search for the νμ→νe oscillation signature suggested by the Liquid Scintillator Neutrino Detector ( LSND) and by the low energy anomaly measured in MiniBoone. MicroBooNE operates at Fermilab since 2015. Two neutrino experiments are being added at Fermilab: SBND and the refurbished version of ICARUS. The Short-Baseline Near Detector (SBND) is a 112 ton active volume liquid argon time projection chamber and it will be one of the three detectors sitting on the Booster Neutrino Beam ( BNB) with MicrooBooNE and ICARUS.

The DUNE single-phase detector modules will be the largest liquid argon single-phase detectors ever constructed. One important novel feature of the DUNE single-phase is the presence of wrapped anode wires that follow a helical trajectory around the height of the Anode Plane Assembly (APA). A single APA is 6 m high by 2.3 m wide. This wire configuration provides a wire pitch of 4.7 mm. The figure below shows an illustration of the DUNE APA wire wrapping scheme:


Two APA are connected vertically, and twenty-five of these vertical stacks are linked together to define a 12 m tall by 58 m long mostly-active readout plane. This design choice was made to minimize the need to tile electronic readout around the perimeter of the APA, which would lead to dead space between neighboring anode plane assemblies. A 10 kton single-phase DUNE module includes then three of these APA readout surfaces, interleaved with cathode planes in order to partition the active volume in compartments of of 3.6 m maximum drift length, as shown in the figure below.


More information is available in the DUNE Far Detector Interim Design Report (IDR) volume on the single-phase design.

Dual-phase LAr TPC readout

The basic idea in a dual-phase LAr TPC is to drift the ionization electrons from the liquid into the vapor phase, where they can be further multiplied. The original idea of a dual-phase detector was developed in 1970s by Dolgshein. The extraction happens by applying across the liquid-vapor interface an electric field produced by two electrodes one of which is a submersed grid and the other one is in the gas phase. It is possible to observe that an electric field E > 2.5 kV/cm gives an extraction efficiency of the electrons close to 100% (see B. M. Gushchin, et al.). Once amplified inside the LEM holes, the charge is then collected on a two-dimensional segmented anode which consists of a set of strips that provide the x and y coordinates of an event with a ~3mm pitch. Both views work by collection and share evenly the produced charge. The principle of the extraction and amplification and collection of the ionization charge in a dual-phase TPC is schematically illustrated in the following figure showing the details of what happens across the LAr-gas interface:


More details on the LEM and anode design are provided in the sections below.

Large Electron Multipliers (LEM)

After the extraction of the electrons from the liquid, once in the vapor phase, the multiplication of the electrons can occur in avalanches occurring in a micropattern structure, the Large Electron Multipliers (LEM) as shown in the figure below. Differently than the applications of thick-GEM for other experiments as COMPASS, for the dual-phase LAr TPC the LEMs have to operate in pure argon vapors, and not with a gas mixture including a quenching component. The avalanches occur in confined spaces (holes in a printed circuit board) which provide a mechanical quenching of the UV photons. This limits the possibility to achieving high gains as in gas mixtures used to operate standard gas detectors.

The use of the LEMs for the charge amplification is motivated by several facts:

  • a tunable gain is achievable, of the order of a few ~10;
  • the detectors can be produced by the industry by using standard tools;
  • it is a detector capable of cryogenic operation with a compact design, large surfaces can be covered as a collection of individual ~ 0.5x0.5 m^2 elements.
The LEM consists of copper cladded epoxy plates with a thickness of a millimeter and with mechanically drilled holes. The holes have diameters of 500 μm and a pitch of 800 μm. There are ~180 holes per cm^2. The copper around the drilled hole is removed in a etched rim of 40 μm in order to avoid edge effects and sharp points which would favor discharges. To prevent high voltage discharges near or across the edge, the LEM has a clearance near to the boundary region (see figure below). LEMs are robust detectors that have been demonstrated to work in cryogenic conditions and can be economically manufactured in the Printed Circuit Board (PCB) industry.


The performance of the extraction, amplification, and collection stage is characterized by a parameter called the effective gain, which takes into account the multiplication of the electrons inside the LEM holes as well as the overall electron transparency of the extraction grid, LEM, and anode. From observation on a prototype of dual-phase TPC equipped with a 10x10 cm^2 readout plan and exposed to the cosmic rays, it was observed that the effective gain of the chamber relaxes, due to charging up effects of the LEM, from an initial value of G0 to a stable value of G1 ~ G0/3, after a characteristic time of about 1.5 days for values of G0 near 100 (arXiv:1412.4402).

Charge collection with segmented anodes

The amplified charge coming out from the LEM is drifted to the anode, where it is collected by two independent charge readout views combined in a single plane providing x and y coordinates with a readout pitch of 3.1 millimeters. In the figure below it is shown the geometry of the strips which are defined by a complicated pattern of gold plated copper tracks on 0.1 mm width and 1.56 mm pitch. In order to solve the topological problem of having two crossing views electrically independent on the same plane, some continuity connections among tracks belonging to the same view are performed on the other side of the printed circuit board by using vias. The pattern of the copper tracks defines overall strips of fixed pitch. This particular pattern has been chosen after several studies in order to ensure an even sharing of the collected charge among the two views without biases related to particular track angles with respect to the strips orientation and to minimize the anode capacitance per unit of length of the strips ( arXiv:1312.6487).


Advantages of the dual-phase design. Dual-phase prototypes and future detectors

The dual-phase TPC provides several advantages with respect to a single-phase LAr TPC. They can be summarized in this way:

  • the LEM allows a tunable gain typically in a range between a few units and 100. The gain G is a function of the electric field: varying the potential difference across the two faces of the LEM around 3 kV it is possible to tune the gain as function of the purity conditions or to provide an additional safety margin with respect to the noise;
  • the charge amplification process passing from the liquid phase to the vapor and the consequent gain obtained in the LEM allows reaching a high signal-to-noise ratio typically greater than 100 and a lower detection threshold (smaller than 100 keV at 3 sigmas above the noise RMS). For a single-phase TPC like ICARUS the typical ratio S/N was ~5 for the inductions planes and ~10 for the collection plane. For a dual-phase TPC the ratio is S/N~100 in both cases;
  • the dual-phase TPC readout is based on two collection views: each plane collects the same quantity of charge, and so it is possible to have a similar reconstructed image in both views. On the contrary, single-phase TPC has induction planes and a collection plane: on the induction planes the S/N is typically worse than the one the collection plane and the signals are bipolar, resulting in a signal cancellation in case of long signals along the drift coordinate as in the case of dense electromagnetic showers. In addition, the use of two collection views limits ambiguities in the bi-dimensional localization compared to more complex schemes based on induction views like the one designed for DUNE single-phase version where one of the induction views consists of wires making a zig-zag pattern around the x and y wires by wrapping on the borders of the wire chamber;
  • high signal-to-noise ratio allows having a fully active long drift projective geometry with no dead materials embedded. It makes then possible reducing the number of readout channels needed to cover the active volume. The number of readout channels is still smaller than an equivalent single phase TPC even by adopting a finer strip pitch: for a very large detector design as in DUNE a dual-phase TPC can use a 3mm pitch, compared to the 4.7 mm size in the single-phase case still having a number of channels smaller by a factor two. This choice impacts the resolution of the detector;
  • in a dual-phase TPC the electronic readout is placed at the top part of the detector: this allows even the cryogenic electronics to be accessible and easily replaceable without contaminating the pure liquid argon volume. The digital electronics can be completely at warm temperature and based on high bandwidth and cheap technologies without the requirements of a digital electronics immersed in the liquid argon as for a SP equivalent detector;
  • a dual-phase TPC requires a smaller number of construction modules than a single-phase TPC which are simpler and less expensive to build and install. This aspect overall reduces the costs and the efforts needed to build a dual-phase module.
The following figure summarizes the dual-phase design for a 10 kton DUNE module. More information is available in the DUNE Far Detector Interim Design Report (IDR) volume on the dual-phase design.


The first large prototype a dual-phase liquid argon TPC is represented by the 3x1x1 m^3 detector, a 4.2 tonnes detector constructed in 2016 and operated in 2017 at CERN. Thanks to this prototype it has been possible to validate the concept of a non-evacuated industrial cryostat based on the GTT membrane design and to demonstrate for the first time the capabilities of the dual-phase liquid argon TPC technology on a tonne scale. The experimental apparatus is illustrated in the figure below. The LEM and anode printed circuit boards employed for the dual-phase readout are segmented in modules of the size of 50x50 cm^2. These detectors, with the extraction grid, are integrated in the Charge Readout Plane (CRP) of the size of 3x1 m^2, described below. The position of the CRP with respect to the liquid level can be adjusted in order to have the liquid level at the middle of the gap in between the submersed extraction grid and the bottom face of the LEM. A detailed description of the experimental setup, the detector components and its performances is given in this paper.


ProtoDUNE dual-phase (ProtoDUNE -DP) is a full scale demonstrator for DUNE 10 ktons dual-phase. It represents 1/20 of active area of a DUNE dual-phase detector, with an active volume of 6x6x6 m^3 (300 tons). ProtoDUNE DP started operating at CERN EHN1 - North Area from August 2019. A schematic view of the detector is given in the figure below. This detector was designed to host and test 4 CRPs of the size of 3x3 m^2, improving the experience and the knowledge acquired by the 3x1x1 m^3 detector and testing the CRP modules design foreseen for a 10 kton detector. The detector operation started in August 2019.


In order to test at cryogenic temperature each CRP assembled for the ProtoDUNE detector, a cold-box setup was installed in the same building hosting the 3x1x1 m^3 detector and operated during summer 2018. In this way it was possible to exploit all the cryogenic and safety infrastructures already put in place for the 3x1x1 m^3 detector. The cold-box has been employed to perform accurate electrical and mechanical tests on each assembled CRP in realistic cryogenic thermodynamic conditions and in reasonably pure argon vapor. A more detailed description of the tests performed in the cold box is provided in a following section. In the figure below it is shown the insertion of one of the CRP in the cold-box. The CRP is hanging from the dismountable roof of the cold-box


R&D on the design an production of LEM detectors

The LEMs were manufactured by ELTOS in Italy for both the 3x1x1 prototype and ProtoDUNE -DP. The LEM for ProtoDUNE -DP are based on a design different from the one used for the 3x1x1. The main difference with respect to the previous version is represented by a larger clearance near the boundary regions, in order to guarantee better operation stability at higher voltage, taking into account the operation feedback from the 3x1x1. The active area of a LEM for ProtoDUNE -DP cover 86% of the total 50x50 cm^2 detector surface. In the figure below it is possible to see the previous version of the LEM installed on the 3x1x1 m^3 detector (left) and the new version (right) designed for ProtoDUNE -DP.


The quite conservative choice of the LEM design for ProtoDUNE -DP was made in order to achieve stable operation conditions up to 3.5 kV. A further optimization on the current LEM design is in progress in order to find the best configuration to minimize the inactive area while still providing good operation stability. These tests are foreseen to be performed in parallel to the operation of ProtoDUNE -DP by using dedicated test benches and the cold-box setup.

Once manufactured by the industry, the LEMs were shipped to CEA/Irfu, where a post production treatment and a quality assurance campaign were set up in order to prepare the LEM for installation in the CRPs. More information about the LEM production and QA and QC can be found in this presentation and here.

LEM visual inspection and survey

First of all a thickness test is performed for all the LEMs. The LEMs undergo an optical survey for a control of the dielectric and copper thickness as well as of the hole geometry. Since the gain uniformity through the LEM holes strongly depends on the total thickness of the manufactured PCB, measurements of FR4 and copper materials of the LEMs are carried out. The precision obtained with the setup used is better than 4 μm: this result is consistent with the requirements. The figures show the distributions of the total LEM thickness and the contribution from the copper layer on both sides of the LEM printed circuit boards which were then installed on two different CRPs.



LEM high voltage connection and cleaning

After the geometry and thickness measurements, the LEMs are prepared for soldering the HV connector pins to power both sides of the LEM, gluing the MACOR insulation around the HV connectors and mounting aluminum handling plates (the handling plates are used to prevent the LEMs to be touched and contaminated during the following tasks), as shown in the figure below on the left.

The cleaning operation is an important phase of the LEM preparation. Since the cleaning requirements are higher than the standard offered by the LEM manufacturer, it has been followed a procedure defined by CERN/EP-DT-EF-MP and CEA/Irfu. This is performed by using an ultrasonic bath a 65ºC and then drying the LEMs at temperatures up to 160 °C. The figure below on the right shows the steps of the cleaning and drying procedure.


LEM high voltage tests

Since the gas amplification depends on the first Townsend coefficient α, which is a function only on the density, for a massive production and characterization of LEMs, it is more convenient to test the devices at room temperature by operating at the same density that one would have in cold argon gas, just above the LAr surface. This requires a Ar pressure of about 3.3 bar achieved in a dedicated high pressure vessel. To perform gain measurements an 214Am source of alpha particles is used. In the figure below it is possible to see a schematic view of the setup for the gain measurement with the alpha source.


The figure below shows the high-pressure vessel used at CEA/Irfu for the characterization of the ProtoDUNE -DP LEM modules. Up to nine LEM modules can be stacked inside this chamber for the HV tests.


In the following figure it is possible to see the event display of a 5.5 MeV alpha track observed in pure argon gas at the pressure of 1 bar (left) et 3 bar (right).

Ar_1bar.png Ar_3bar.png

All LEMs were validated after a training phase of several hours aiming at voltages across the 1 mm thick LEMs up to 3.4-3.5 kV, with the requirement of a maximal discharge rate of less than 3 discharges per hour.

LEM integration on large dual-phase readout surfaces

The LEMs installation has to take in consideration several technical points:

  • The size of each LEM is 50x50 cm^2. In order to cover large surfaces (3m^2 for the 3x1x1 m^3 detector or 36 m^2 for the 6x6x6 m^3 detector) it is necessary to assembly side-by-side several LEMs;
  • Once the LEM have been assembled in larger units it is crucial to guarantee that they define altogether a perfectly planar surface;
  • There should be the possibility to finely adjust the distance and the parallelism of the plane defined by the LEMs bottom surfaces with respect to the liquid argon surface;
  • The LEMs have to be positioned at fixed distances from the extraction grid and from the anode plane. These distances have to be defined in a precise way by the design on the mechanical system integrating the grid-LEM-anodes.
All these aspects were the goal of an extensive R&D program. In order to satisfy all these requirements, the extraction grid, the LEM and the anode plane are combined in an array of independent (layered) modules called Charge Readout Planes (CRPs). A CRP is composed of several 0.5×0.5m^2 units, each of which is composed of a LEM-anode sandwich. These units are embedded in a mechanically reinforced frame. Each frame is suspended from the roof of the cryostat with 3 ropes whose length can be adjusted with step motors. This design guarantees the planarity requirements over the CRP span of +-1mm, despite the temperature gradient present in the gas phase just above the liquid and possible sagging effects with respect to the three suspension points used to hang the CRP.

Charge Readout Planes (CRP)

The CRPs are the basic readout components of the future 10 kton dual-phase detector modules. The CRP mechanical structure provides the integration of the LEM-anode sandwich over a large area by minimizing the dead spaces. The planarity of the order of the mm of the active surface must be guaranteed despite possible sagging effects with respect to the three hanging points, despite differential thermal contraction effects on the various components and despite the presence of a temperature gradient in the gas phase, which could induce different thermal contractions as a function of the distance from the liquid surface. The CRP includes a submersed extraction grid at 1cm from the bottom face of the LEMs. The extraction grid and the LEMs bottom faces geometry and applied voltages define the electric field, which is exploited in order to extract the electrons from the liquid to the gas phase. The liquid argon level should be across the 1cm gap in between the grid and the LEMs. The suspension system allows adjusting the position of the gap with respect to the liquid level as well as the parallelism of the CRP plane with respect to the liquid surface. The CRP is electrically and mechanically independent from the drift cage and can be remotely adjusted to the liquid level. The 6x6x6 m^3 prototype provides a full scale test of this concept by having its anode plane segmented in 4 of these 3×3 m^2 CRP units. Each CRP integrates 36 LEM-ANODE sandwiches of 50x50 cm^2. The main design concepts of the CRP are discussed in this presentation.

Each CRP is composed of three main parts, as shown in the figure below:

  • a mechanical Invar frame which can provide a stiff supporting structure, extending vertically into the gas phase, with little sagging and very minor contraction effects, ensuring the planarity;
  • a G10 structure, which is an assembly of 9 sub-frames of 1 m^2 each. The G10 structure integrates the LEM and anodes, which are affected by a significant thermal contraction, and has a similar thermal behavior as the LEM-anode sandwiches. The G10 structure is mechanically decoupled on the horizontal plane with respect to the Invar structure (which has little thermal contraction) and it is free to slide during its thermal contraction;
  • the LEM-anode sandwiched, which are attached to this structure, are assembled with spacers that guarantee a 2 mm uniform distance between the charge amplification and collection stages.

The CRP design for the 6x6x6 m^3, based on modules of 3x3 m^2, was completed in November 2016. Due to time constraints, it was decided to instrument with LEMs and anodes only 2 of the 4 CRPs of ProtoDUNE -DP. The two other non instrumented CRP are needed in order to close the field lines and perform the extraction of the electrons from LAr. They consist essentially of the same mechanical structure equipped with the extraction grid, the LEMs are not mounted and non segmented anodes are connected to ground. One of these 2 CRP is equipped with 4 segmented anodes connected to the readout system. These anodes come from the spares of the production performed for the first 2 CRPs and are used, as a crosscheck, to read signals from electrons extracted to the gas phase but not undergoing LEMs amplification.

The figure below shows the detail of a corner of an assembled CRP: it is possible to see clearly the anode, the LEM and the extraction grid.


CRP construction

The CRPs production had to be preceded by a period devoted in the installation of the different tooling in the clean room for the structure assembly, the extraction grid weaving and the manipulation as well as the assembly of the first transport box used to transport the CRP for the cold box test and then to insert the CRP into the cryostat. This setup and the related organization procedures represent the model for a production center which can be used to the assembly of a future large detector like the DUNE 10 kton.

The CRP assembly procedure includes 8 main steps:

  • G10 frame assembly from 9 sub-frames and installation of all inserts and coupling screws;
  • Geometry survey to determine the size along the 2 directions;
  • Assembly of the G10 frame with the Invar frame by using 50 decoupling points;
  • Installation of the LEM High Voltage cables;
  • Fixation of the 36 anodes and LEMs;
  • Parallel construction of 30 sets of extraction grid wires modules;
  • Installation of the extraction grid modules;
  • Geometry survey and tuning of the planarity
The G10 frame assembly takes several days in order to guarantee the flatness and to add all the fixation elements prior to the Invar frame coupling and anodes. Before the coupling to the Invar frame, the G10 frame undergoes a precise metrology in order to know with a precision better than 0.1 mm the size of the frame along both direction. Those measurements are used to adjust the length of the grid production tool such that the grid modules have exactly the length required to get a nominal tension of about 1.5 N/wire when they are screwed on the G10 frame.

The grid production including the weaving, the wire soldering on the PCB plates and the storage can be performed at a rate of 5 modules/day. One module corresponds to 64 wires. The completion of the 30 modules of one CRP and the subsequent installation takes about 10 working days.

The whole construction period, including tests, extended over 8 months: CRP1 was built in May-June 2018, CRP2 was built in October 2018, CRP3 was built in September 2018 and CRP4 was built in January 2019. The figure below shows the sequence of operations performed in May 2018 to build the first CRP in the clean room. In general the basic time to build a complete CRP is of the order of one month. More details on the CRP integration procedures can be found in this presentation.


Full size cold-box tests experience

After its construction, each CRP has to undergo to an electrical and mechanical test in realistic cryogenic thermodynamic conditions in pure argon vapor. In order to perform these tests, a specific cold-box setup was built at CERN, in the same building where the 3x1x1 m^3 detector operated. In this way, it was possible to exploit all the cryogenic and safety infrastructures already in place for the operation of the 3x1x1 prototype. The goals of this CRP tests are the characterization of the high voltage operation of each LEM and of the extraction grid, the test of all the high voltage contacts from the feedthroughs to the LEM and grid connectors, and the measurement of the flatness of the CRP before and after the cool-down.

The external cold-box structure is a self-sustained box made out of 1 cm-thick stainless steel plates. The internal membrane is made out of 2 mm-thick stainless steel plates. The internal dimensions are 1 m in height and 3.9 m in the other directions. The thermal insulation is passive and it consists of four layers each of 10 cm-thick polyurethane foam. The insulation layers are separated by polyethylene foils to minimize the convection phenomena and the consequent increase of heat input. The insulation region is flushed with gas nitrogen, as it is done for the 3x1x1 detector.

The cold box is basically an open bath. The liquid argon level is kept constant by compensating the losses due to the evaporation with new liquid argon. The CRP is attached to a portion of the roof that can be opened. The horizontality of the CRP can be adjusted manually from the hanging system installed on the roof. The figure below shows a schematic view of the cold box.


The requirements for the environment of this test can be summarized as follows:

  • The LAr surface must be flat in order to allow the LEM to be in the vapor phase and the grid in liquid;
  • Vapor pressure and temperature must be constantly monitored;
  • Liquid argon purity must be of the order of 100 ppm, in order to be comparable with the tests at 3.3 bar and room temperature done at CEA/Irfu;
  • CRP position and planarity should be adjustable.
The monitoring instrumentation put in place for the cold box cryogenic operations consists of 7 temperature probes, a coaxial capacitive level meter, pressure meters and 3 cryogenic cameras. The 7 temperature probes were used in order to approximately determine the liquid argon level. They were positioned at 0, 2, 4, 16, 18, 20, 22 cm from the cryostat floor. During the seven different cold-box tests the level defined by the CRPs position corresponded to a liquid height from the cryostat floor in between 18 and 20 cm. The coaxial level meter was used by the cryogenic system to measure the target level to be reached by the liquid. To complete the instrumentation 3 cryogenic cameras were used to watch with a wide angle view below and above the CRP. One of the cameras was positioned to give a side view of the CRP with a fi eld of view of the order of 40 cm. This camera was the most useful as there was a clear view of the grid supporting plates, the LEM bottom surface and the liquid.

During the cold-box tests, the CRPs were operated successfully at the voltage needed to maintain the extraction field at the value corresponding to the maximum extraction efficiency. The LEM high voltage tests performed in the cold-box allowed characterizing the maximum achievable field inside the LEM holes still corresponding to stable operation conditions. It was also possible to evaluate the LEMs discharge rate as a function of the voltage and related the effective gain.

R&D on readout electronics

This aspect of the R&D focuses on designing, testing and producing cost effective and high performance analog and digital electronics for very large LAr detectors based on the dual-phase technology. The R&D activities covered in three main items covered by the WP8 Charge Readout task:

  • R&D on cryogenic readout electronics;
  • Development of high bandwidth DAQ system for giant LAr detectors;
  • Development of synchronization systems for giant LAr detector.
The electrical signals from the collected charges are brought outside the cryostat via a set of dedicated signal feedthrough chimneys. The chimneys are pipes passing through the top layer of the cryostat insulation and closed at the top and at the bottom by ultra-high-vacuum flanges. The cryogenic FE electronics cards, housed at the bottom of the chimney are plugged to the top side of the cold flange. The FE cards are based on analog cryogenic pre-amplifiers implemented in CMOS ASIC circuits for high integration and large-scale affordable production. The ASIC circuits have been especially designed, following an R&D process started in 2006, to match the signal dynamics of a DP module. Within the chimney, the cards are actively cooled to a temperature of about 110K and isolated with respect to the LAr vessel by the cold flange feedthrough. The bottom side of the cold flange is connected to the CRP via short flat cables in order to minimize the input capacitance to the pre-amplifiers. This concept guarantees having the analog cryogenic electronics very close to the detector but still accessible and replaceable from outside without risks of contamination of the pure LAr volume inside the cryostat. The cryogenic analog electronics is also completely shielded with respect to possible noise coming from the digital electronics and the outside environment.

The digital electronics for the charge digitization system, also resulting from a long R&D process started in 2006, is located on the roof the cryostat at room temperature. In this way, it is possible to use the low-cost, high-speed networking technologies used in the telecommunication industries, such as Micro Telecommunications Computing Architecture (μTCA). Digitization cards in the advanced mezzanine card ( AMC) format read 64 channels per card. Each AMC card can digitize all 64 channels at 2.5MHz and compress and transmit this continuous data stream, without zero skipping, over a network link operating at 10 Gbit/s. Lossless data compression is particularly effective thanks to the high S/N ratio of dual phase, which limits noise contributions at the level of one analog-to-digital converter (ADC) count. Each signal feedthrough chimney is coupled to a μTCA crate. The crate hosts 10 AMC digitization cards in order to read 640 channels and transmit the digitized data via the MicroTCA Carrier Hub (MCH) switch through a 10 Gbit/s optical link connected to the DAQ back-end. The figure below shows a schematic concept of the charge readout electronics architecture.


A clock distribution and synchronization system is needed in order to feed and align all the AMC digitization cards which operate independently. The clock distribution and synchronization are based on the White Rabbit (WR) standard. A specifically developed timing MCH connected to a WR network ensures the distribution of clock, absolute timing, and trigger information on the back-plane of the μTCA crates. The White Rabbit μTCA Carrier Hub (WR-MCH) are connected via 1 Gbit/s optical fibers to a system of WR switches used to interconnect the WR network. This timing network ensures that the digitization performed by the various AMC cards is completely aligned and it also refers to the absolute UTC time. The WR Grandmaster switch is connected to a GPS disciplined oscillator unit, providing absolute time and the clock frequency reference to the system.

A subset of the electronics system including all components developed in this R&D program (analog front-end electronics, digitization system and timing system), for 1280 channels (vs 7680 channels foreseen for ProtoDUNE -DP) was commissioned in the fall 2016 on the 3×1×1m^3 detector and it has been successfully operating for about one year (see here). This system included also a reduced version of the back-end online storage/processing facility which allowed validating its architecture, finalize the corresponding design for the 6×6×6m^3 (ProtoDUNE -DP) back-end system and developing the related data handling and processing software.

Details on the various steps of the R&D process are given in this presentation. More information about all the developed components of the electronics readout system are reported in the following sections and can be also found in the 3x1x1 paper and in the DUNE dual-phase IDR volume .

Cryogenic analog readout electronics

The Signal FeedThroughs (SFT) consist of 1.5-2 m (different lengths for 3x1x1 and ProtoDUNE -DP) stainless vacuum tight stainless steel tubes terminated with appropriate UHV flanges that provide the interface for routing the signals. They are designed to enable the access of the Front-End (FE) analog electronics for possible repair or exchange without contaminating the ultra-pure argon in the cryostat. In addition, their metallic structure acts as a Faraday cage, completely shielding the FE cards from the external environment and potential noise induced by the digital electronics.

The signal feedthroughs are closed at their bottom and top with vacuum tight flanges made out of 4 mm thick multilayer PCBs. The flange at the bottom (cold flange) interconnects the signals from the anodes to the analogue FE electronics cards located inside the tube and directly connected to the cold flange. The flange at the top (warm flange) acts as interface to route the analogue signals towards the digitizers and also feeds the required low voltage and control lines to the FE cards.

The FE cards are mounted on ~1.5-2m long blades made of G10, allowing for the insertion or extraction of the analogue electronics from the top of the SFT. The blades also support the flat cables that transmit the signals via differential pairs including also the low voltage lines, and other control lines. The figure below shows the details of a signal feedthrough and a μTCA crate on the 3x1x1 detector.


Each analog front-end card, as shown in the figure below, hosts four ASIC amplifier chips and a few passive discrete components for the input stage and the filtering of the low voltage power supply lines.


The input stage of every channel has a decoupling capacitor of 2.2 nF and a 1GΩ resistor that connects the anode strips to ground. An ESD device is also included into each input stage and is used to protect the amplifiers against discharges coming from the detector.

The principal component of the FE analogue cards is the cryogenic charge amplifier ASIC based on CMOS 0.35 μm technology with a large dynamic range (1200 fC) to cope with charge amplification in the CRP. The chip features 16 input channels each consisting of a charge sensitive amplifier (CSA), differential output buffer stage acting as low pass filter, and 1 pF test capacitor. The ASIC amplifier has a linear gain for input charges of up to 400 fC and then a logarithmic response in the 400-1200 fC range, as shown in the figure below. The ASIC power consumption is less than 18mW per channel in order to limit the heat dissipation inside the SFT.


High bandwidth data acquistion systems for giant LAr detectors

The AMC cards (see figure below) read and digitize the data from the front-end amplifiers and transmit them to the back-end DAQ system. The cards also include a last stage of analog shaping before the ADC input. The analog front-end cards produce differential unipolar signals defined with respect to a common positive baseline offset. Prior to the digitization, this offset is removed and the signals are subtracted in the analog input stage of the digital electronics.

Each AMC card includes eight ADC chips (Analog Devices, AD92574), two dual-port memories (Integrated Device Technology IDT70T3339), and a field programmable gate array (FPGA) (Altera Cyclone V ) on board. The FPGA provides a virtual processor ( NIOS) handling the readout and the data transmission. The choices of all components were defined in order to comply to the design requirements while optimizing at the same time other technical criteria such as: costs, chip footprints (small enough to fit on the AMC surface) and power consumption. The Cyclone V FPGA was accurately dimensioned, in order to minimize costs, following an extensive prototyping activity based on a more powerful Stratix IV GX development board ( S4-AMC).

Each AMC generates a continuous compressed stream of 2.5 MSPS 12 bit data per readout channel. The on-board ADCs operate at a rate of 25 MHz per channel. The data are down-sampled in the FPGA to 2.5 MHz by performing ten-sample averaging, which leads to further digital filtering of the noise. The data, consisting of only the 12 most significant bits from each digitized 14 bit sample, are then compressed without losses by using an optimized version of the Huffman algorithm and organized in Ethernet frames for transmission. Each frame contains the absolute timing information of the first data sample for redundancy. In the current design, each AMC has 64 channels and reads one analog front-end card.


The AMCs are hosted in μTCA crates and transmit their data via the MCH switch. Each AMC has 10 Gb/s data bandwidth. The timing synchronization of AMCs is achieved via a WR-MCH module, hosted as well in each μTCA crate, connected to the WR network. The WR-MCH is also used for the readout of AMCs triggered by external signals, by treating dedicated packets, containing trigger timestamp information, which are transmitted over the WR network.

The uTCA crates (12 crates for the charge readout system of ProtoDUNE -DP and 240 crates for the equivalent system on a DUNE 10 kton module), each one connected to the DAQ via a 10 Gb/s link, produce continuously very large data flow which must be received and treated by a very high bandwidth DAQ back-end system. The ProtoDUNE -DP back-end/storage/processing system consists of four main elements:

  • two levels of event builders machines connected to the DAQ front-end: two Level 1 machines and four Level 2 machines;
  • the network infrastructure;
  • the DAQ service machines providing the supervision and ancillary services to the system;
  • the online storage/processing facility.
The back-end architecture of ProtoDUNE-DP is modular and can be replicated by a factor 20 for a 10 kton module. The online storage system of ProtoDUNE-DP would be already sufficient at the 10 kton scale, given the low rate of events to be written on disck by the detector operating underground. The installation of the DAQ system implemented for ProtoDUNE -DP at EHN1 was completed in between summer 2017 and the fall 2018. A general scheme of the system is shown in the figure below.

The two event builders of Level 1 collect the data from the μTCA crates via private 10 Gb/s links. The task of each LV1 Event Builder is to put together data from the μTCA crates for the same drift window corresponding to half of the detector. The event data corresponding to half detector are written by each LV1 Event Builder on a RAM disk which is visible through the network to the LV2 event builders. All Event Builders have 80 Gb/s connectivity via two connections to a switch providing several 40Gb/s connections. The four LV2 Event builders assemble the half events produced by the LV1 Event Builders in full events grouped in files of 3 GB which are then pushed through the EOS distributed file system of the online storage facility where hundreds of disk units are operating in parallel in order to achieve the required bandwidth. The entire back-end system going from the LV1 events builders, the LV2 event builders, the network elements and the online storage facility is designed to guarantee a data bandwidth of 20 GB/s.

The DAQ system is an evolution of the system implemented on the 3x1x1 m^3 and performs data taking of event time windows which correspond to triggers time-tagged and handled by the white-rabbit trigger system. The 6x6x6 m^3 back-end infrastructure, given its architecture and the characteristics of the network elements and the machines employed is also a useful test-bench for the DUNE 10 kton back-end architecture.


Timing and synchronization systems for giant LAr detectors

The time synchronization system designed for the dual-phase readout exploits a White Rabbit network, which combines the synchronous 1 Gbit/s Ethernet (SyncE) technology with the exchange of PTPV2 packets in order to synchronize clocks of distant nodes to a common time base.

A high stability GPS disciplined oscillator ( GPSDO), with accuracy similar to that of an atomic clock, provides a clock reference signal to be distributed over the physical layer interface of the WR Ethernet network. The network topology is based on a cascade of specially designed switches (WR switches) that have the standard IEEE802.1x Ethernet bridge functionality with an addition of WR specific extensions to preserve the clock accuracy. The first switch at top on the network, connected to the GPSDO, is called Grandmaster switch. Time and frequency information are distributed via the switches to the end-nodes on the WR network via optical fibers. The WR protocol automatically performs dynamic self-calibrations to account for any propagation delays and keeps all connected nodes continuously synchronized to sub-ns precision.

The sub-ns precision on the clock synchronization is not strictly needed for aligning samples in the different dual-phase electronics AMC digitization units, since the timing granularity on the data is 400 ns. However, the WR timing system offers readily available industrial components and the necessary protocols for synchronization with automatic calibration of delay propagation. R&D on this timing distribution solution started in 2006; the final design for integrating this system, planned for the 3x1x1 prototype, ProtoDUNE -DP, and the DP module readout, was completed in 2016.

In the implementation specific to ProtoDUNE -DP, a GPS-disciplined clock unit (Meinberg LANTIME M600) feeds 10MHz and 1 PPS reference signals to a commercial WR switch (Seven Solutions WRS v3.4). The switch acts as Grandmaster of the WR network. The Grandmaster switch is then connected via 1 Gb/s optical links to the dedicated WR time-stamping node (WR-TSN) and to the WR end-node slave cards present within each μTCA crate (WR-MCH), in order to keep all these end-nodes synchronized with respect to its reference time aligned to the GPSDO. The WR Grandmaster is connected through a standard Ethernet port with the LANTIME GPSDO for its date synchronization via the NTP protocol.

The WR-MCH card developed in this R&D activity (see figure below) enables clock/timing/trigger distribution to AMCs. It communicates with them via dedicated lines in the back-plane of the μTCA crate using a customized data-frame protocol. The module contains a commercial WR slave node card, the WR Lite Embedded Node (Seven Solutions OEM WR-LEN), as mezzanine card. The WR-LEN operates with a customized firmware, which also enables it to decode the trigger time-stamp data packet received over the WR network and to transmit the data-frames on the μTCA crate back-plane.


External triggers are handled in the 3x1x1 and ProtoDUNE dual-phase by the specific WR-TSN time-stamping node producing the timestamps of the corresponding electrical trigger signals (beam triggers, cosmic counters, calibration triggers, light readout triggers). The WR-TSN module receives analog TTL-level trigger signals, generates their timestamps, expressed in the WR common time base, and transmits the data of the timestamps as Ethernet packets over the WR network to all the connected WR-MCH units. This timestamp information is then used by AMCs to find the data of the drift window starting with the trigger and transmit them to the DAQ system, by having both the trigger timestamp and the absolute time of the digitized samples expressed in the same time base.

Operation of large dual-phase prototypes: 3x1x1, cold-box, ProtoDUNE -DP

The 3x1x1 m^3 prototype was the first large dual-phase liquid argon TPC and represented an intial step towards the construction of large cryogenic liquid detectors, as DUNE. This prototype was also the first one built with a cryostat based on the GTT membrane technology, which is envisaged for the construction of the cryostats of the DUNE detector modules. The construction of the 3x1x1 prototype allowed achieving a large experience on several items related to the WP8 charge readout task :

  • Demonstration of the dual-phase concept for a LArTPC at the 3 m2 readout scale;

  • Operational feedback on the LEMs and CRP design and related high voltage distribution system, which had then design improvements implemented for ProtoDUNE -DP and tested in the cold-box setup;

  • Demonstration of the operation and performance of the cryogenic front-end electronics, high bandwidth digitization and DAQ system and timing and synchronization system
The WP8 charge readout task includes also all the definition of the procurement, assembly and test procedures for the large scale production of the LEMs, CRPs and the electronics system. All these items were developed during 2017-2018, following the feedback from the operation of the 3x1x1 prototype. The performance of the newly designed LEMs and CRPs was then assessed in the cold-box setup.

The extension of the AIDA2020 deliverable for the WP8 charge readout task allowed including as well the feedback from the first operation of ProtoDUNE -DP. For the first time, full scale charge readout planes (3x3 m^2) were built, tested in a cold-box setup and were integrated and operated together in a detector prototype representing 1/20 of the readout surface of a DUNE 10 kton module. This experience validated the design of several elements such as the CRPs, the signal feedthrough chimneys and the readout electronic and timing systems, in view of building a large scale detector module at the 10 kton scale based on the dual-phase technology.

The construction of the ProtoDUNE -DP detector inside its cryostat was completed by March 2019. The following picture represents an inside view of the cryostat at that time showing the main detector elements: the cathode at the bottom, the field cage panels all around the empty drift volume and the four CRPs at the top.


In between April and June 2019 the cryostat Temporary Construction Opening (TCO), used during construution to insert the detector elements, was closed and sealed. The external cabling was completed, as well as the installation of the cryogenic instrumentation inside the cryostat via a man-hole access path. On June 13th 2019 the man-hole was sealed and the gas purge and filling operations were started.

The filling with liquid argon of ProtoDUNE -DP was completed by August 10th 2019. After two weeks dedicated to the CRPs commissioning (level meters calibration, alignment with respect to the liquid surface and grids high voltage commissioning) first cosmic ray tracks were observed since August 29th, as soon as the LEMs were switched on.

The following picture shows a side view taken with a cryogenic camera of two CRP modules in ProtoDUNE -DP before the immersion of the extraction grids in LAr and the CRPs exact positioning with respect to the LAr surface


The figure below givens and idea of the CRPs positioning accuracy once operating across the LAr level interface. The figure shows the LAr level measurements taken by a high accuracy level meter installed on a CRP while it was operating in automatic LAr level tracking mode. The LAr level in the cryostat was increasing by about 1mm/hour. The CRP, suspended from the roof with a system of ropes controlled by step motors, was self-adjusting its distance with respect to the LAr surface in steps of 0.25 mm about every 15 minutes in order to keep its extraction grid at almost constant immersion depth.


The cryogenic front-end electronics, the high bandwidth data acquisition system and the timing system had been developed and tested during the past years and already operated successfully on the 3x1x1 prototype. However their integration at the larger scale of ProtoDUNE -DP provided the opportunity to confirm the validity of the design and the good performance of the system in terms of S/N and operation stability, making sure that there are no issues when these concepts are integrated at full-scale representative of a 10 kton detector module. The system operated also at much higher bandwidth than in the 3x1x1 detector given the number of channels and the higher trigger rate. An overall 20GB/s operation bandwidth was demonstrated.

The figure below shows a portion of the roof of the ProtoDUNE -DP cryostat with two signal feedthrough chimneys, hosting at their bottom the cryogenic front-end cards, and the associated uTCA crates of the digitization and high bandwidth DAQ system. Each chimney-crate pair handles 640 readout channels.


During the commissioning of ProtoDUNE -DP in September 2019 a full exercise of replacement of the electronics in a SFT chimney was performed while the detector was operating. The replacement procedure for 10 cryogenic front-end cards (see picture below) proceeded smoothly without any issues and took about 15 minutes. This was an important point of validation of the electronics design.


In summary the construction and commissioning of ProtoDUNE -DP offered the possibility to perform a ultimate, full-scale, assessment of the following aspects which are pertinent to the Charge readout task of AIDA2020 WP8 :

  • Validation of the CRP design, assembly, operation and tracking of the LAr surface level
  • Validation of the LEMs design and on their operation on large detection surfaces
  • Validation of the design and operation of the cryogenic front-end electronics, high bandwidth data acquisition system and timing and synchronization system
  • Validation of the signal feedthrough chimneys and of the accessibility to the front-end electronics with a complete exercise of replacement of the front-end cryogenic cards
At the time of October 2019 ProtoDUNE -DP has been operating by collecting more than 1.2 millions of cosmic ray events with the LEMs progressively operating at higher voltages up to 3.2 kV potential difference in between the two faces.

The following figure shows an example of recorded cosmic ray event including some electromagnetic showers and two muon decays.


-- DarioAutiero - 2018-01-28

Topic attachments
I Attachment History Action Size Date Who Comment
PNGpng 3x1x1.png r1 manage 1394.6 K 2019-07-24 - 11:52 DavideCaiulo1  
PNGpng 6x6x6.png r1 manage 650.0 K 2019-07-24 - 11:52 DavideCaiulo1  
PNGpng AMC.PNG r1 manage 2057.7 K 2019-11-01 - 12:21 DarioAutiero Advanced Mezzanine Card developed for the high bandwithd digitization system
PNGpng Ampl_response.PNG r1 manage 26.9 K 2019-10-30 - 19:04 DavideCaiulo1  
PNGpng Ar_1bar.png r1 manage 248.2 K 2019-10-04 - 18:58 DavideCaiulo1  
PNGpng Ar_3bar.png r1 manage 44.6 K 2019-10-04 - 18:58 DavideCaiulo1  
PNGpng Back-end_arch.png r1 manage 236.4 K 2019-07-24 - 15:17 DavideCaiulo1  
PNGpng CRP.png r1 manage 196.5 K 2019-10-15 - 18:36 DavideCaiulo1  
JPEGjpg CRP_PDDP.jpg r1 manage 55.7 K 2019-11-04 - 14:47 DarioAutiero CRP modules in protoDUNE-DP before immersion of the grids in LAr
PNGpng CRP_construction.png r1 manage 1277.7 K 2019-07-24 - 14:49 DavideCaiulo1  
PNGpng CRP_scheme.png r1 manage 456.2 K 2019-07-24 - 14:49 DavideCaiulo1  
PNGpng CRP_tracking.PNG r1 manage 116.6 K 2019-11-04 - 14:38 DarioAutiero CRP liquid level tracking in ProtoDUNE-DP
PNGpng Chimney_access.png r1 manage 338.9 K 2019-11-04 - 16:50 DarioAutiero Replacement of cryogenic front-end cards during the commissioning of ProtoDUNE-DP
PNGpng Cleaning_LEM.PNG r1 manage 698.9 K 2019-10-16 - 12:44 DavideCaiulo1  
PNGpng Cleaning_LEM.png r1 manage 1270.0 K 2019-10-16 - 11:39 DavideCaiulo1  
PNGpng Cryo_FE_card.PNG r1 manage 831.3 K 2019-11-01 - 12:41 DarioAutiero Picture of an analog cryogenic front-end card
PNGpng DAQ_Arch.png r1 manage 506.7 K 2019-07-24 - 15:07 DavideCaiulo1  
PNGpng DPscheme.png r1 manage 44.2 K 2019-07-18 - 22:50 DavideCaiulo1  
PNGpng DUNE-SP.PNG r1 manage 1079.0 K 2019-11-04 - 10:01 DarioAutiero APA arrangement in a DUNE single-phase detector module
PNGpng DUNE-dp.PNG r1 manage 841.8 K 2019-11-04 - 10:59 DarioAutiero Picture of a DUNE Far detector module based on the dual-phase LAr TPC technology
PNGpng Dual-Phase.png r1 manage 349.5 K 2019-07-18 - 22:43 DavideCaiulo1 illustration of the extraction, amplification and readout regions in a dual-phase LAr TPC.
PNGpng GainTest_LEM.PNG r1 manage 17.5 K 2019-10-23 - 19:38 DavideCaiulo1  
PNGpng Holes_LEM.png r1 manage 67.2 K 2019-10-23 - 17:24 DavideCaiulo1  
JPEGjpg LBNE_Graphic_061615_2016.jpg r1 manage 259.6 K 2019-10-29 - 10:29 DavideCaiulo1  
PNGpng LEM-test_1.png r1 manage 684.1 K 2019-10-04 - 18:33 DavideCaiulo1  
PNGpng LEM-test_2.png r1 manage 919.9 K 2019-10-04 - 18:33 DavideCaiulo1  
PNGpng LEM.png r1 manage 1479.1 K 2019-07-18 - 22:57 DavideCaiulo1  
PNGpng LEM_Copper_CRP1.png r1 manage 44.7 K 2019-07-24 - 14:12 DavideCaiulo1  
PNGpng LEM_Copper_CRP2.png r1 manage 45.2 K 2019-07-24 - 14:12 DavideCaiulo1  
PNGpng LEM_Thick_CRP1.png r1 manage 38.2 K 2019-07-24 - 14:12 DavideCaiulo1  
PNGpng LEM_Thick_CRP2.png r1 manage 41.0 K 2019-07-24 - 14:12 DavideCaiulo1  
PNGpng NP02_arch.PNG r1 manage 132.1 K 2019-10-25 - 19:00 DavideCaiulo1  
PNGpng ProtoDUNE-DP.png r1 manage 2859.6 K 2019-11-01 - 09:52 DarioAutiero Inside view of the ProtoDUNE-DP in March 2019 after the completion of its construction
PNGpng SFT.png r1 manage 590.5 K 2019-10-30 - 18:51 DavideCaiulo1  
PNGpng SPDP.PNG r1 manage 281.1 K 2018-01-28 - 17:19 DarioAutiero Single and dual phase LAr TPC redout
PNGpng Signal-SP.png r1 manage 345.4 K 2019-10-22 - 17:52 DavideCaiulo1  
PNGpng Soldering_LEM.PNG r1 manage 676.6 K 2019-10-16 - 12:44 DavideCaiulo1  
PNGpng WR_MCH.png r1 manage 648.9 K 2019-11-01 - 11:59 DarioAutiero Picture of a timing/synchronization end-node card (WR-MCH) based on the White-Rabbit protocol
PNGpng Wires-SP.png r2 r1 manage 769.9 K 2019-10-31 - 10:22 DavideCaiulo1  
PNGpng coldbox.png r1 manage 180.9 K 2019-07-24 - 14:49 DavideCaiulo1  
PNGpng coldbox_pic.png r1 manage 832.7 K 2019-10-30 - 16:12 DavideCaiulo1  
JPEGjpg crates.jpg r1 manage 236.9 K 2019-11-01 - 09:35 DarioAutiero View of the uTCA crates of the digitization system on the roof of the protoDUNE-DP cryostat
PNGpng electronics.png r1 manage 310.2 K 2019-07-24 - 15:07 DavideCaiulo1  
PNGpng evt.png r1 manage 415.6 K 2019-11-01 - 08:39 DarioAutiero ProtoDUNE-DP cosmic ray event
PNGpng new_LEM.png r1 manage 2538.5 K 2019-10-23 - 17:24 DavideCaiulo1  
PNGpng old_LEM.png r1 manage 2532.8 K 2019-10-16 - 11:43 DavideCaiulo1  
PNGpng pitch1.png r1 manage 34.2 K 2019-07-24 - 11:23 DavideCaiulo1  
PNGpng pitch2.png r1 manage 100.6 K 2019-07-24 - 11:23 DavideCaiulo1  
Edit | Attach | Watch | Print version | History: r57 < r56 < r55 < r54 < r53 | Backlinks | Raw View | WYSIWYG | More topic actions
Topic revision: r57 - 2019-11-04 - DarioAutiero
    • Cern Search Icon Cern Search
    • TWiki Search Icon TWiki Search
    • Google Search Icon Google Search

    AIDA2020WP8 All webs login

This site is powered by the TWiki collaboration platform Powered by PerlCopyright &© 2008-2022 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
or Ideas, requests, problems regarding TWiki? use Discourse or Send feedback