Labs & Services

The IEEC academic and technological activity has motivated the creation of different laboratories to simulate the space and environment conditions:

Radiation Laboratory

The Radiation Laboratory at ICE (CSIC-IEEC) is devoted to characterise radiation detectors based on a semiconductor material such as Cadmium Zinc Telluride (Cd(Zn)Te) and Silicon (Si) in the field of High Energy Astrophysics, mainly, for space-based observatories. The available equipment allows to perform spectroscopy measurements at room and low temperature in the X and gamma-ray range (from tens of keVs to 1-2 MeV). The experimental setup consists basically of a vacuum chamber, an oil-sealed rotary vane vacuum pump, a freezer or cooling plate and a controllable high voltage power supply. The vacuum chamber is made of Aluminium and has a volume of 350mm x 300mm x 350mm. The Radiation Laboratory has been authorized by the “Consejo de Seguridad Nuclear (CSN)” with the number IRA-3137.

Available equipment:

A) Vacuum chamber

  • Dimensions: 350 x 300 x 350mm
  • Leakage vacuum: 10-6mbar

B) Vacuum pump

  • Nominal flow: 6 m3/h
  • Ultimate total pressure: 3 10-2 mbar

C) Cooling and heating control equipment. Model: Julabo FP50-HL

  • Working temperature range: -50 … 200 ºC
  • Temperature stability: ±0.01 ºC
  • Refrigerant: R404A / R507
  • Overall dimensions: 42x49x72 cm
  • Weight: 57 kg

D) Freezer. Model: Radiber GT-199

  • Capacity: 199 lts
  • Temperature: -30ºC

E) High voltage power supply. Model: Keithley 2410

F) 6U VME430 195 Mini crate series: The crate offers 9 VME slots in compact designed chassis with integrated low noise power supply and cooling fan.

G) X / gamma spectrometer:

  • Universal coincidence, Model 418 A
  • Quad timing filter amplifier, Model 863
  • Linear gate, Model 426
  • ADCAM multichanel buffer, Model 926
  • 5kV detector bias supply, Model 659
  • NIM BIN (NIM POSER SUPPLY), type 7022-7033
  • Spectroscopy amplifier and gate integrator, Model 673

H) Exempt radiation sources: 133 Ba, 57 Co, 22 Na, 137 Cs, 109 Cd

I) General purpose laboratory instrumentation:

  • 6 1/2 Digital multimeter, DMM 357, Wayne Kerr
  • Dual power supply, EL 302D, TTi
  • Four channel oscilloscope 200MHz, 2GS/s, TDS 2024B, Tektronik
  • DC power supply, IPS 3303D, ISO-TECH

View of different pitch adapters within a 4-inch Sapphire wafer

Test bench to perform spectroscopy measurements of the detector in the Planar Parallel Field configuration (hard-X and soft-gamma rays parallel to detector Electric field)

Spectroscopy measurements with a pixelated CdTe detector inside a vacuum chamber and low temperature (-10ºC)


Cushioned optical bench

Since 2010, the ICE (IEEC–CSIC) has a cushioned optical bench of about 1,5 × 2,5 m with the equipment necessary to CCD detectors calibration and characterisation. The first project that used this equipment has been PAU (Physics for the Accelerating Universe) to characterize the 18 CCD detectors that form its focal plane and the three reservation detectors. This detectors, built by the japanese company Hamamatsu K. K., are back-illuminated and measure 2048 × 4096 pixels. The complete characterization of each device consists in 9 tests that are executed in 24 hours after the device has been installed in the criostate in a white room with quality of 1000, tha vacuum has benn done and has been refrigerated to -100ºC.

One of the tests made is the determination of the spectral response of each device, from 300 tp 1.100nm. An other test consists on bombing the detector with X-rays generated by a 55Fe source, as it allows various data such as the gain or transmission of loads within the detector. With a calibrated photodiode, the filters u, g, r, i, z i Y of the PAU camera, that have been purchased to the japanese company Asahi Spectra Co. Ltd., have been verified and characterized.

The IEEC, within the IFAE (Institut de Física d’Altes Energies), is the responsible of design, build and connect the filters wheel of the infrared equipment of the ESA space project Euclid. One of the tasks to be developed is to glue filters to the assemblage and preventing tensions. A first task has been the tests by birenfringence of glue tests.

Gravitational Wave Detectors

This facility aims at test technologies and materials for gravitational wave detection in space. The proposed set-up is composed by a very high-precision (10^(-6) K/sqrt(Hz)) thermally controlled vacuum tank which allows to suppress environment fluctuations in the low frequency regime, i.e. down to 0.1 mHz. Inside the tank, an interferometer with picometer sensitivity will allow the characterization of materials used in space applications (like Carbon-Fiber Reinforced Plastics or Silicon Carbide) and opto-electronics equipment (like photodiodes or optical fibres) in a high stability environment.

This instrument will be able not only to screen the environmental fluctuations but to generate controlled perturbations to characterize the samples in a measuring bandwidth relevant for space applications. The technologies to be tested in this low-frequency test bed are of wide application in space technology. In particular, spacecraft-to-spacecraft interferometry concepts are currently being considered for geodesy missions, as in the case of the GRACE (Gravity Recovery and Climate Experiment) follow-on missions, which share similar low-frequency and high-precision requirements as the gravitational wave detection missions. Thus, this test bed can be considered a transversal test facility for space-related technologies.


The Montsec Lab is formed by the instrumentation of The Telescopi Joan Oró (TJO), a 1m-class telescope working in a completely unattended manner. The TJO is equipped with two instruments: a photometric imaging camera (MEIA) and a medium-resolution spectrograph (ARES).The construction of this robotic telescope motivated the development of the Observatori Astronòmic del Montsec (OAdM), a site devoted to host astronomical research facilities.


The MEIA instrument is the optical imager of TJO. It consists in two components: the CCD camera and the filter wheel. In addition, a system to acquire dome flats is available.

CCD camera

The CCD camera is a ProLine 4240 (model PL4240-1-B), with a back-illuminated 2k×2k chip manufactured by Finger Lakes Instrumentation (FLI):

  • Model: CCD42-40-1-B
  • Sensor Manufacturer: E2V
  • Sensor Type: Back Illuminated
  • Coating: Basic Midband
  • Number of pixels: 2048×2048
  • Pixel Size: 13.5×13.5 ?m (0.36×0.36 arcsec @ TJO)
  • Field of view at TJO: 12.3×12.3 arcmin
  • Sensor Diagonal: 39.1mm
  • Peak Quantum Efficiency: 96%
  • Typical Working Temperature: -30º C
  • Temperature Stability: 0.1º C
  • Typical Gain: 1.53 e-/ADU
  • Typical Dark Current: <1 e-/pixel/sec. @ -30º C
  • Typical Readout Noise: 8 e- RMS @ 500 kHz
  • Non linearity: <1%
  • Read-out time: 10 seconds

Filter wheel

The filter wheel is physically coupled to the telescope at the back of the primary mirror support. It can hold up to 12 3-inch filters that are placed at the optical axis of the telescope by rotating the filter wheel. Currently, 5 Johnson-Cousins photometric filters (manufactured by Custom Scientific) are installed: U, B, V, Rc, and Ic.

The limiting magnitude of MEIA for an exposure time of 300 seconds (the maximum recommended) to reach S/N=100 for each one of the five Johnson-Cousins filters is:

U B V Rc Ic
13.2 17.0 17.0 17.2 17.0

The maximum S/N that can be reached with MEIA in a single exposure strongly depends on seeing and sky brightness. However, values with a S/N=750 should ensure an optimum S/N within linerarity regime of the CCD, except in cases with very good seeing.

The time required to change from one filter to another is 30 seconds.

Dome flats

Although currently only sky flats are being obtained, dome flat-field images are expected to be routinely obtained with a homogeneity better than 1% RMS using the following instrumental set-up:

  • A white screen mounted on the inside of the dome.
  • A full-spectrum light source that illuminates the screen.
  • A control system to turn on the light when required by the observatory control system.

Standard calibration procedure

The TJO control system is designed to periodically run the acquisition of calibration images, including bias, darks and sky flat-fields:

  • Bias. Bias frames are usually taken in multiples of 5 after the acquisition of sky flats or between two science sequences along the night. Usually, a minimum number of ten bias frames are provided to the observers.
  • Dark. Dark current frames are usually taken just before or after the acquisition of bias frames. The exposure time for the dark frames depends on the exposure time of science images. Usually, the exposure time for the dark frames, corresponds to the maximum exposure time for the science frames.
  • Sky flat-fields. Sky flats are routinely taken by pointing to several blank fields during twilight. When sky conditions do not allow the acquisition of sky flat exposures, images taken previously can also be provided to the observers. Flat-field images are routinely obtained with an illumination that provides a homogeneity better than 1% RMS.

Calibration exposures are expected to be used for an on-the-fly evaluation of the quality of the science images taken with MEIA. All the calibration images are provided to the observers having science sequences during the night.


ARES is the optical spectrograph (designed by Fractal SLNE) currently being integrated at the TJO. It is mainly conceived to perform medium-resolution spectroscopy with the following characteristics:

  • Limit star magnitude: V < 11 mag.
  • Global trasmissivity: > 10%
  • Spectral resolution: R=12000
  • Two spectral windows:
  • Green: between 495 and 529 nm
  • Red: between 634 and 678 nm.

In order to achieve the high global transmissivity (>10%) two telescopes (manufactured by Takahashi) and two VPH are placed in Littrow configuration.


The ARES spectrograph dispersion system is composed by two VPHs developed by Wasatch Photonics, providing the two spectral windows and maintaining the high overall throughput. Both VPHs are mounted on a carriage (prepared to hold up to three VPHs), allowing a fast change from one VPH to the other.

CCD Detector

The detector used is an Andor Technology CCD Newton DU940P camera with the following features:

  • Sensor type: BV- Back Illuminated CCD, VIS optimized
  • Active pixels: 2048 x 512
  • Pixel sixe: 13.5 x 13.5 µm
  • Image area: 26.7 x 6.9 mm
  • Maximum cooling: -100ºC
  • Read-out noise: as low as 2.8 electrons
  • Dark current: as low as 0.0002 e-/pixel/sec
  • Linearity: better than 99%


El laboratorio de Instrumentación del IEEC se utiliza para el diseño, caracterización y validación de equipos destinados a misiones espaciales. Su uso está focalizado en electrónica de bajo ruido y baja frecuencia (0.1 mHz), con instrumentación especialmente dedicada a la realización de medidas magnéticas. Entre los diferentes test que se han realizado, destacan los de cualificación espacial del sistema de diagnósticos de la misión LISA Pathfinder. Algunos de los sensores/unidades validados según requerimientos de la Agencia Espacial Europea son: magnetómetros, sensores de temperatura, monitor de radiación y control de actuadores térmicos y magnéticos.Por otro lado, el laboratorio también se utiliza para la validación de software. Numerosas campañas oficiales, como el Stress Test Campaing del Software para LISA Pathfinder, han requerido la utilización de sistemas con cualificación espacial, como por ejemplo Data Management Unit. El laboratorio es el lugar idóneo para el montaje de tal infraestructura en este tipo de test.


The NanoSat Lab is a laboratory of small satellites and payloads, designed for up to 3 U CubeSats and small payloads or subsystems. It was set up to achieve educational, scientific and service objectives. Unlike other laboratories working in this area, it offers companies and institutions the possibility of qualifying components for space flight use. The NanoSat Lab has developed into a facility that is open to students and the scientific community. It also provides services for small and medium-sized enterprises and institutions or groups that wish to launch small satellites into space. The NanoSat Lab is one of the few European installations (outside Barcelona, the closest ones are in Madrid and Tolosa) that has the equipment needed to qualify components intended to be placed in orbit, both technologically and in terms of relevant standards.


hake table

Simulate launch conditions

  • Random vibrations
  • Sinusoidal vibrations
  • Shock


  • Fmax = 7,325 kN peak (sin)
  • 4,100 kN rms (rand)
  • ?xmax = 25 mmpp.
  • vmax = 1.5 m/s
  • amax = 120 g
  • amaxshock = 360 g
  • fmin = 5 Hz – fres = 2.800 Hz
  • Armor = 170 mm.
  • Table = 460 x 460 mm2

Thermal Vacuum Chamber (TVAC) & Sun simulator (partially sponsored by IEEC):

TVAC – Space simulator:

  • Pmin = 10-5 mbar
  • Tmin ~80 K (-193ºC)
  • Quartz window for Sun simulator
  • Quartz window can be covered using high emissivity material
  • Germanium window for temperature imaging (Fluke TIR camera available)

Sun simulator:

  • 4 KW Xenon lamp
  • Up to 2 Suns TOA (1 Sun = 1400 W/m²)
  • Intensity monitored and calibrated with piranometer inside TVAC

Helmholtz coils with air bearing:


  • ACS (Attitude Control Systems) design and tessting
  • Magnetic shileding testing
  • Calibration of magnetometers


Test ACS and mass properties

Amateur ground station (co-sponsored by Telecom Barcelona and the Dept. of Signal Theory and Communications):

  • Indicative: EA3RCU
  • VHF (144 MHz)/UHF(436 MHz) & C-band (to be installed)
  • Remote controlled (GENSO compatible, while GENSO was operational)

Foster university research in space technologies and help development of local SMEs

  • Unique suite of instrumentation and testing facilities all together in Europe.
  • One stop & shop point for basic testing at an affordable cost.
  • Environment that promotes creativity and entrepeneurship.




Hidra is a computing cluster designed to balance computing speed and storage capacity. It has two different physical components: a blade box (computing component) and an external array of disks (storage system), both of them mounted in the same rack.
The blade box (meaning that each computing node is not a stand-alone computer, but rather a sort of detachable motherboard with an external “face” called a blade), is filled to its maximum capacity of 8 full size nodes. Each one of this nodes has 4 processors, and every processor has 6 cores, which gives a total of 192 cores. Every node has also 48 GB of RAM, so the system has 2 GB RAM per core or 384 GB RAM in full (not shared, except for the cores within each node).
The storage system is made of four detached arrays of twelve 1 TB disks, for a total of 48 TB of raw space, which is reduced to 22 TB after configuring RAID and spare disks.

Blade box

The box which houses the blades (or computing nodes) houses too the physical elements that allow the nodes to connect between themselves (Infiniband and Ethernet switches) and to the storage system (SAS switch), external connections, the power sources and some diagnostic and management tools. This box has 6 power sources, but it can work with only 3 of them (it needs at least 4 to boot, but once booting is finished, it can keep working with 3). There is no space available to mount more blades in the box (8 full size or 16 half size blades is the maximum capacity), so if we wanted to acquire more nodes, we would need to buy a new box first.

Computing nodes

All of the blades have exactly the same hardware. Each blade has 4 hexacore processors (so there are 24 cores per node), 48 GB RAM, two mezzanine SAS cards (for disk access), a mezzanine Infiniband card and 4 Ethernet cards plus a not specially wonderful graphic card integrated into the motherboard. Each blade has two SAS 2.5_ hard disks of low capacity (147 GB), where the operative system is located. All of the nodes run a standard 64 bit Debian Squeeze GNU/Linux distribution.

Storage space

The storage space is divided in four disk trays, with 12 3.5’’ SAS 1 TB disks each one (mind that we are using TB in the sense of 1012 bytes, not 240 bytes, which we would refer as 1 TiB. This means that each disk capacity is 930 GiB, not 1024 GiB). The disk trays are paired in two groups of two trays and, within each group, every disk forms a hardware RAID1 pair with its equivalently located disk in the other tray, except for the last disk in each tray, which is set as a global “spare”. In this way, even losing an entire tray shouldn’t affect (in fact, it didn’t affect, for we did lose an entire tray when its power source broke) the cluster.
The set of 22 RAID1 disks (22 TB or 20 TiB) is directly connected to each node through a double (in the sense of redundant) SAS controller. Every RAID1 set is configured as an LVM volume group, containing a single logical volume (that is the equivalent for partitions in LVM), which is formatted using ext4, mounted locally on one node and exported through NFS to all other nodes. Only homes and the backup directory are stored in those nodes, and every logical volume is mounted in a /home subdirectory named after the volume group that contains it (that is, /home/hidra-1, home-hidra-2,.., /home/hidra-22).
Each node has too a small RAID1 pair of disks (147 GB) where the operative system is installed. The files in these disks are not visible to the other nodes, and since this space includes the /tmp directory, you should be careful not to use /tmp when executing multiprocessor jobs (unless you use those files only in one thread).