Publication list from


HVCMOS Sensors


HVCMOS sensors are our innovation, they are pixelated sensors in CMOS technology where high voltage is used to increase the sensor volume and improve time resolution, detection efficiency and radiation hardness. HVCMOS sensors are suitable for detection of single ionizing particles. Application fields are high energy physics, photon science, medicine, electron imaging.

Development of generic HVCMOS sensors is supportet by BMBF within the project 05H21VKRD1 (Entwicklung aktiver und passiver mikrostrukturierter CMOS-Sensoren).


Latest publications


1. HVCMOS sensor prototype with counting electronics for a beam monitor at an ion therapy facility:

journal homepage:

accepted manuscript: High_Voltage_CMOS_active_pixel_sensor_chip.pdf

2. Large area HVCMOS sensor for particle physics ATLASPIX3:

journal homepage:

accepted manuscript: IEEE_JSSC_HVCMOS.pdf


Latest designs (user manuals)


1. TSI engineering run HVMAPS (HIT, LHCb, HVMAPS)

TSI engineering run v2.pdf




HVCMOS Sensors for medical applications


Radiotherapy is in important method in treatment of tumours. The most commonly used radiotherapy is x-ray and gamma radiation.

More recently, irradiation with heavy ionized particles - such as protons and carbon ions - have been introduced clinically. The source of these particles is a particle accelerator. In contrast to x-ray and gamma radiation, ions and protons depose energy is a small tissue volume. This is the result of the effect, that deposed energy is inversely proportional to the particle velocity. Consequently, particles loose most of the energy near the end of their path when they get slow (Bragg-peak). By adjusting the beam direction and particle energy it can be achieved that the largest portion of energy is delivered to the tumour, and the healthy tissue in front and behind the tumour is less affected.

Under realistic conditions, the minimum spot size is not determined by the beam properties than by the accuracy of the tumour localization.

We are developing the device – beam monitor – that precisely measures the position, size and intensity of the beam. The beam monitor could be applied at the Heidelberg Ion-Beam Therapy Center (HIT).

Motivation: The present beam monitor is made of gas-filled ionization and multiwire chambers (MWC) that provide dose, position, and spot size information. MWCs have the drawback that they cannot be operated in the magnetic field of MRI. An MRI compatible beam monitor (MR guided radiotherapy) would significantly improve the precision of the irradiation and reduce the exposure of the healthy tissue.

One advantage of the HVCMOS sensors over MWCs is that the HVCMOS sensors are not magnetic and can be operated in the field of MRI. On the other hand, HVCMOS sensors are more radiation tolerant than the standard monolithic active pixel sensors (MAPS) thanks to their large and depleted sensor diodes.

We have designed and tested a few sensor prototypes with counting and integrating electronics such as HITCNT1.

The project was supported by HeiKA.


HITCNT2 chips (courtesy of M. Pittermann)


HVCMOS Sensors for LHC Experiments


ADL develops HVCMOS sensors for experiments with hadrons (supported by BMBF), the upgrades of LHC experiments ATLAS or LHCb (supported by HeiKA) and the experiment COMPASS. We believe that these experiments could benefit from the use of HVCMOS sensors in terms of performances and construction costs. For ATLAS we have designed a series of ATLASPIX sensors.

The latest sensor ATLASPIX3 is 2cm x 2.1cm large. It contains a pixel matrix with 132 x 372 "smart" pixels that can measure the time and amplitude of particle hits. Digital electronics for signal filtering and formatting is implemented on chip. 

A similar chip called MightyPix is being developed for LHCb.

ATLASPIX sensor designed for ATLAS ITk and implemented in a 180nm HVCMOS process (courtesy of M. V. B. Pinto)


HVCMOS Sensors for Mu3e Experiment

Decays of elementary particles fulfil conservation laws, such as conservation of energy, momentum and charge. In all charged leptons decays observed so far, also the lepton numbers were conserved. The lepton number (electron-, muon- or tau number) is the difference between the number of leptons and the number of antileptons belonging to one of three flavours. The corresponding law is called lepton flavour conservation.

There is a certain evidence that the conservation of lepton number is not the fundamental conservation law resulting from a fundamental symmetry of nature.

Recently it was discovered that neutrinos oscillate from one flavour to another. This suggests that lepton flavour is not preserved in the neutral sector.
The Standard Model does not conserve quark flavour and does not explain why this is different for leptons.

Lepton flavour conservation is result of an accidental symmetry within the Standard Model. As soon as the particle model is changed, this symmetry disappears. Some theories beyond the Standard Model allow decays of charged leptons that do not conserve lepton numbers and experimental discovery of such decays would be of great importance. This is the motivation for the Mu3e experiment. 

Mu3e is an experiment at the Paul Scherrer Institute (PSI, Villigen, Switzerland). Mu3e will utilize a highly intense muon beam (108 muons decay/s in the phase 1) to search for the decay of muons in three electrons:

µ+ -> e+ e+ e-

This decay would imply change of lepton number.

The proposed Mu3e detector is based on two double-layers of HV-MAPS around a hollow double cone target, with additional timing detectors. The muons from the beam get stopped at the target and decay at the rest, their kinetic energy can be neglected.

Drawing of Mu3e detector (courtesy of N. Berger and A. Schöning)

Mu3e detector is designed to detect three electrons originating from one muon decay (the tacks have vertex) with the sum energy equal to the muon rest energy. It should be fast enough to cope with about 109 decays/s. The electron momentum (and energy) is measured by tracking electrons in the solenoidal magnetic field. Only with high spatial and momentum resolution it will be possible to distinguish the rare µ->3e decay from the common decays that could also lead to three simultaneous electron signals, although with wrong total energy or non-existing vertex.

The momentum resolution is limited by the deflection of electrons in the sensor material – multiple scattering. The average deflection angle depends on particle velocity, momentum, and detector thickness. The maximal momentum of an electron passing Mu3e detector is < 53MeV/c (53MeV is half of the muon rest energy) and the scattering angle can be large. Therefore, in order achieve required momentum resolution the detector must be ultra-thin. The detector thickness per layer should be less than 0.1% radiation lengths that corresponds to 100µm silicon. Sensor chip thickness will be only 50µm. The chips will be mounted to a low mass flexible interconnect circuit. The detector will be operated inside a dry helium atmosphere and cooled by helium gas flow to further reduce multiple scattering.

The Mu3e pixel tracker will consist of three parts, the central pixel tracker and two recurl stations. Pixel layers in the central tracker provide the main hits used for the reconstruction of tracks and of the decay vertex associated with multiple tracks. To further improve resolution, we utilise the fact that, in the case of muon decays at rest, all track momenta are below 53MeV/c and all tracks will curl back towards the magnet axis. After half a turn, effects of multiple scattering on the momentum measurement cancel to first order. To exploit this feature, the experimental design is optimised specifically for the measurement of recurling tracks by adding the recurl detectors. The total detector surface will be more than one square meter.

HVCMOS sensors are the only particle pixel senor type that allow high time resolution, construction of thin detector and large sensor area for affordable price.

All detector modules will be based on identical MuPix sensor chips, with an sensitive area off 20.48 × 20.00 mm2 equipped with pixels of size 80 × 80 µm2, corresponding to 256 × 250 pixels and small non-active area that houses peripheral digital and analogue circuitry. The chips are mounted on high density interconnect circuits (HDI), which incorporate both signal and power lines as aluminium traces on thin polyimide substrates. The HDIs provide power and bias voltage and transmit control signals and data (1.25 Gbit/s per line). Pixel modules are constructed from self-supporting sensor-HDI-polyimide ladders.

Each pixel consists of the sensor diode, a charge-sensitive amplifier and a source follower to drive the signal to the chip periphery with readout circuits. The MuPix readout circuitry provides zero suppression and generates timestamps to enable time-matching of hits from different pixel layers. A second timestamp is generated to provide time over-threshold (ToT) information. Offline, the ToT information can be used for time-walk corrections, better noise suppression, and for improving the spatial resolution by means of charge sharing. After hit detection and digitisation, an internal state machine collects all hits using an address prioritisation scheme. Data are sent via up to three serial links at a nominal data rate of 1.25 Gbit/s each using a simple protocol with time frames and 8 bit/10 bit encoding.

The experiment is currently being constructed within an international collaboration.

The latest sensor for Mu3e MUPIX11 has been submitted early 2022.

Since 2021, the Mu3e experiment is funded by DFG as a research unit with members from University of Heidelberg, KIT and University of Mainz.


HVCMOS Sensors for the Compact Linear Collider


ADL develops also the HVCMOS sensors for the future linear colliders, such as the Compact Linear Collider (CLIC). Capacitively coupled pixel detectors (CCPDs) based on a HVCMOS sensor and a readout chip are investigated. The chip to chip signal transmission is done by capacitive coupling.


Belle II


Belle II is an experiment at the Super KEKB accelerator (KEK, Tsukuba, Japan). The goal of the experiment is to measure the decay time of B mesons and anti B mesons and in this way examine the particle-antiparticle asymmetry. With our partners, ADL has developed the vertex detector for this experiment. The detector is based on the DEPFET sensor.


DEPFET sensor, SWITCHERs and DCD (courtesy of M. Koch)


To read out and control the sensors, three types of ASICs are used. The SWITCHER, to produce high-voltage signals and to turn on and off the pixel rows. The DEPFET Current Digitizer - DCD, to receive, amplify and digitize the sensor signals. The DHP-chip to process the digital amplitudes provided by DCD. ADL is responsible for the development and production of SWITCHERs and DCDs. One DCD contains 256 8-bit ADCs that take samples with 100ns period. One SWITCHER contains 32 high-voltage channels that generate fast 10ns signals with amplitudes of up to 20V.

The project is supported by BMBF within the project 05H21VKKB1 (Belle II: Pixeldetektor, Software und erste Datenanalysen).

The Belle II experiment started in March 2019.


Inner structure of DCD chip (plot made with gds2pov)


Inner structure of DCD chip (plot made with gds2pov)



3D Ultrasound Computer Tomography is a novel method for early detection of breast cancer developed at IPE. Ultra-sound computer tomography is an alternative to x-ray computer tomography (CT) or magnetic resonance imaging (MRI) for the medical breast imaging (mamography). USCT is cheaper than MRI and it does not expose patient to x-rays as CT. The USCT developed at KIT contains about 2000 ultrasound transceivers placed in a half spherical reservoir filled with water. The transducers are grouped in groups with 18 devices, each group is driven and readout with two USCT9C ASICs that are developed in our group.
Every ASIC contains nine high voltage analogue drivers and three low noise receivers. The high voltage driver generates output signals with amplitude of 120V. The bandwidth of the amplifier is about 5MHz. The receiver is implemented as a three-stage amplifier; every stage is a voltage amplifier with feedback that uses a single ended inverting amplifier as the active element. The feedback can be configured in the way to provide wide band (0.1 – 5MHz) and narrow band amplification. Frequencies and amplification can be varied. Input referred noise less than 10µV was simulated.
First 3D USCT device has been already build and successfully tested.


Nanometer-scale CMOS technologies


Nanometer-scale CMOS technologies are excellent choice for low power and high speed analogue designs and for high-resolution pixel sensors. We are developing circuits in these CMOS processes.

Our first design is a test chip in 28nm high performance RF CMOS technology of TSMC within a micro-ASIC run. One possible application would be the readout of superconducting sensors or qubits.

The submitted SAR ADC is based on an 8-bit capacitive rail to rail DAC and a switched differential comparator. The sample and hold circuit uses a bootstrapped switch. A synthesized digital part controls the DAC does 8-bit to 10-bit conversion of the ADC output and serializes the data words. First measurement results are promising.


Hybrid pixel detectors and pixel readout chips


Hybrid pixel detectors are the type of pixel sensors where the sensor and the readout part are implemented on separate substrates. In this way it is possible to implement so called smart pixels with complex in-pixel signal processing. Application fields are medical imaging, particle physics, synchrotron science, experiments at free electron lasers... We are investigating the novel concepts in hybrid pixel detectors such as capacitive signal transmission (CCPD) or simultaneous photon counting and charge integration for digital computer tomography.

The first ASIC we developed was the PHOTON readout chip with pixel size of 150µm x 150µm. The pixels contain high dynamic range integrators (26 bits) and counters (13 bits). The project is supported by Helmholtz programme Matter and Technology.

The latest design is the pixel readout chip MPROC. MPROC is implemented in 180nm CMOS technology, the pixel size is 55µm x 55µm. The chip can measure single particle signals with high energy and time resolution. One application would be detection of gold nanoparticles with a CdTe sensor (PLASMED-x).


Pixel readout chip PHOTON1 capable to count photons and measure their accumulated energy




The aim of TRISTAN experiment is to measure the spectrum of beta particles emitted by tritium source at KATRIN and to search for a signature of the sterile neutrinos. A sterile neutrino is the theoretical neutrino type that interacts only through gravitation with other particles. It may have a large mass. A discovery of sterile neutrinos could give answers to some of the most significant unresolved questions in science such as the origin of dark matter and the abundance of matter versus antimatter. ADL does detector R&D for TRISTAN:

Development of integrated readout electronics for silicon sensors. The electronics include charge sensitive amplifier and ADC. We are investigating the use of through silicon vias to allow construction of more compact detectors.

Testing of sensors with thin entrance windows of different sensor producers (AMS, FBK, HLL).

Development of detector concept based on fast particle counters in form of hybrid detectors.

The development is supported by Helmholtz programme Matter and Technology.


Very good energy resolution, Fe55 signal


CMOS detectors with high time resolution


ADL develops monolithic CMOS sensors based on standard- or avalanche photodiodes (APDs) with high time resolution.

Several prototypes have been designed in CMOS and SiGe BiCMOS technologies.

Applications: particle physics experiments, measurements of time of flight, replacement for PMTs: detection of cosmic particles and neutrinos (though measurement of Cherenkov light or UV florescence), detection of light photons emitted by scintillators.