Education logo MSc
(Master of Science)
Table of contents
About our MSc Education
The M.Sc. in Microelectronics is a 24 month program taught in English, which offers challenging high-level education and research opportunities to talented students with a B.Sc. degree in engineering or equivalent. The first year of study comprises theoretical study, assignments and laboratory work. The second year is devoted to a thesis project, which involves participation in one of our research projects or an assignment in collaboration with a local company.

The 2-year program starts with a number of compulsory courses that cover the entire field (the core module). Subsequently, the student can select courses from a list of specialisation courses to design a programme that suits his interest. The MSc Program in Microelectronics brings the student in close contact with the challenging and internationally recognised research areas that are covered by the Electronics Research Laboratory via its strong participation in the Delft Institute of Microelectronics and Submicrontechnology (DIMES) (see http://www.dimes.tudelft.nl). The global character of Microelectronics is reflected by the international background of both the students and the professors. Our department has its own clean rooms and extensive characterization laboratories, so that a student's circuit idea can be realized as a working silicon chip. You are invited to take a virtual tour ( http://www.dimes.tudelft.nl/labtour/index.html)!

More specific information regarding the MSc courses in Electrical Engineering and the specialisation in Microelectronics is also accessible via: http://masters.ewi.tudelft.nl/index.php?master=microelectronics/


Scholarships and Grants

There are several scholarship programs for students who cannot (fully) finance their M.Sc. studies from private resources. Although scholarship programs may have different requirements and criteria, almost all organizations that offer scholarships require admission to the university before your scholarship application is granted. Consequently, you must first apply for admission to the M.Sc. program. Following (conditional) acceptance of your application, you may apply for a scholarship from these organizations. More information can be found at: www.tudelft.nl/msc/programme/scholarships.cfm

There are a number of scholarships available specifically for the M.Sc. in Microelectronics. Companies such as Philips Semiconductors, ASML Taiwan and National Semiconductor sponsor students. Delft University of Technology also offers a limited number of scholarships. Upon (conditional) admission the Grant Awarding Committee of the Department of Microelectronics will consider your application for dedicated financial support. Please explain on your application form your potential private resources and your needs and preferences for additional financial support.


Philips Semiconductors M.Sc. Scholarship

Philips Semiconductors Candidates for Philips Semiconductors scholarships are expected to show a clear interest in a future career in microelectronics. In your essay you have to convince the Grant Awarding Committee of the Department of Microelectronics that you are interested in an education and a future career in the design of electronic circuits and systems. The Philips Semiconductors scholarships will partly cover your living expenses and includes a budget for the obligatory medical insurance. Candidates are expected to pay their tuition fees by private means or additional support. During the M.Sc. study there will be a close contact between you and Philips Semiconductors. It is ultimately intended that you will carry out your final M.Sc. graduation project in close cooperation with Philips Semiconductors in Nijmegen (The Netherlands). Upon successful graduation, Philips Semiconductors will likely offer you a permanent position as an employee, but this depends on their hiring needs. If you do not accept a job offer, the scholarship has to be refunded to Philips within 3 years. However, Philips is not obliged to offer you a position. If for whatever reason Philips cannot or does not offer you a position, you will not be obliged to repay the scholarship. If you have been offered a position at Philips one third of your scholarship will be changed into a grant after each consecutive year of employment. For detailed information about Philips Semiconductors, see: http://www.semiconductors.philips.com/index.html

Taiwanese candidates can also apply for scholarships at the Netherlands Trade & Investment Office (NTIO). This office provides information about scholarships from Philips Taiwan and ASML Taiwan. Please visit the site of NTIO: http://www.ntio.org.tw/education/grad_eindex.htm

Philips Taiwan and the Netherlands Trade & Investment Office (NTIO) jointly launched the Y.C. Lo Scholarships for Post-Graduate Studies in the Netherlands (YCLS). The scholarships support promising Taiwan graduates and young professionals intending to pursue advanced degree programs at Dutch universities.


Admission requirements

Generally students who wish to apply for admission to the M.Sc. program must hold a B.Sc. degree (or equivalent) in Electrical Engineering or have met equivalent standard requirements. Students who hold a Bachelor's degree in a related field may apply to the program but, if admitted, they will be required to follow a specific program to make up for any deficiencies. In addition, candidates are expected to have:
  1. A Grade Point Average (GPA) for the B.Sc. of at least 75% of the scale maximum.
  2. A Graduate Record Examination (GRE) score of at least 450 verbal/550 analytical/650 quantitative proof of proficiency in the English language.
  3. A TOEFL score of at least 550 (paper-based test) or 213 (computer-based test), or an IELTS (academic version) overall Band score of at least 6.0.
Applications should include:
  1. A completed, current M.Sc. application form with a recent passport photograph.
  2. An original TOEFL or IELTS score form.
  3. An original document showing the applicants GRE general test scores.
  4. A curriculum vitae.
  5. A personal statement in English of the applicants motivation to participate in the M.Sc. program.
  6. Certified copies of the applicants academic diplomas in both the original language and translations in either English, French, German or Dutch.
  7. An original or certified copy of the full list of grades in the original language together with a certified translation of the same list of grades in either English, French, German or Dutch.
  8. Two original letters of reference; from academic staff members and if applicable, from your current employer.
An application form can be downloaded at http://www.tudelft.nl/msc/programme/applicationform.cfm


Tuition fees and other information

Tuition fees and other information about the M.Sc. in Microelectronics can be found via the following link: http://masters.ewi.tudelft.nl/index.php?master=microelectronics


Service fee for non-EU/EFTA students

TU Delft provides non EU/EFTA students with additional facilities such as accelerated visa procedure, guaranteed furnished accommodation, a one-month introductory summer course, and private health and liability insurance, which are compulsory in the Netherlands. For these services, a one-time fee is charged (not including the costs of accommodation and insurance).


About the University and Delft

The Netherlands is situated in north-west Europe, surrounded by Germany, Belgium and the North Sea. It is adjacent to Great Britain. With a population of approximately 16 million people, it is one of the most densely populated countries in the world. A large proportion of the population live in the western part of the country which is known as the Randstad. In this area you find the city of Amsterdam (capital of The Netherlands), the international airport - Schiphol, The Hague which houses the residence of the Dutch Government, and Rotterdam, which is one of the largest seaports in the world.

Delft is a historic city nestled between the seaport of Rotterdam and The Hague. Its population is approximately 100,000. Delft is just 60 km from the main international airport (Schiphol) and is well serviced by public transport. Delft has acquired the title Knowledge-based City due to the abundance of technology-based institutions and organisations setting up business in the proximity of the university campus.

Delft University of Technology (founded in 1842) is the oldest, largest, and most comprehensive technical university in the Netherlands. Today, it is a modern, forward-looking university of both national importance and significant international standing. Its 7 faculties offer 24 engineering M.Sc. degree courses, many of which are unique in the Netherlands. With a student body of more than 13,000 and almost 5,000 employees, it is an impressive establishment, renowned for its high standard and quality of innovative research and education. TU Delft has close collaborative links with international business partners, various research institutes and industry. Such ties give students the opportunity to gain valuable, relevant experience in preparation for an international career. Website: http://www.tudelft.nl
MSc projects:
RF and high-speed microelectronics
Contact: J. Long  
 W A. Serdijn   
  1. Design of a Low Noise Amplifier for dense phased arrays
    This work is related to the activity described here.
  2. Design of a CMOS RF power detector (400 MHz - 2.5 GHz)
  3. Design of a broadband low-noise amplifier for ultra-wideband communications.
    This work is related to the activity described here.
  4. Design of low-power adaptive ring oscillator for a quadrature downconversion autocorrelation receiver
  5. Ultrawideband-FM (UWB-FM) transmitter
  6. Ultrawideband-FM (UWB-FM) receiver front-end
  7. Varactorless voltage-controlled oscillator
  8. 60GHz transceiver building blocks (LNA, mixer, etc.)
  9. Collision avoidance radar transmitter
  10. High linearity mixing in deep submicron CMOS
  11. The design, implementation and measurement of a HF multiplier for application in an UWB transceiver.
  12. The design, implementation and measurement of a HF controllable delay for application in an UWB transceiver.
    Both projects will be embedded in the AIRLINK project of Delft University of Technology and TNO-FEL.
  13. Broadband Ultra Low-Noise Amplifier for SKA
    The Netherlands Foundation for Research In Astronomy is currently developing a new radio telescope that is 100 times as sensitive as the best telescopes up to now. This high sensitivity is obtained by increasing the collecting area to approximately 1 square kilometer. The project, entitled Square Kilometer Array (SKA), employs multiple antennas that are connected in a so-called phased array, together forming one larger antenna. The total amount of antennas that are envisioned is approximately 100 million(!). The figure on the right shows a SKA test setup.

    Besides a large collecting area, a low noise temperature of the entire receiver chain is of utmost importance. To obtain this low noise temperature, better LNAs than currently available are needed. Compared to LNAs that are currently employed in, e.g., cell phones, wireless LANs and other radio-astronomy applications,
    1. its noise temperature must be lower (approximately 30 K),
    2. and must maintain low over the entire frequency band (600--1600MHz) in the presence of a complex source (antenna) impedance that varies with frequency; and
    3. the linearity must retain high
    The proposed M.Sc. project is concentrated around the following research questions and design tasks:
    1. Which IC technology (GaAs, SiGe, CMOS) is most suited to fulfill the above requirements?
    2. Which LNA topology (e.g., employing double-loop negative feedback), to be implemented in this technology, meets the above requirements?
    3. How can this topology be implemented by means of an integrated circuit?
    4. How does the performance of this integrated circuit relate to the performance of a conventional LNA?
    5. When implemented, what is the measured performance of the LNA?
    For this project, we are looking for two M.Sc. students that want to do their final project in the field of RF analog integrated circuit design. The work will be carried out at ASTRON (part time) and at the Electronics Research Laboratory (ELCA) of Delft University of Technology (part time), in close cooperation with experts in the field of radio astronomy, antennas, RF electronics and integrated circuit design. Dr. Jeroen C. Kuenen (RF system engineer, ASTRON) and Dr. Wouter A. Serdijn (Associate Professor, ELCA) will be responsible for the students' daily supervision.

    More information on this topic can be obtained from:

    Dr. W A. Serdijn
    Room: EWI H18.310
    Phone: (015) 278 1715
    Email: 

Mixed-signal electronics
Contact: J. Long   
  1. Low-IF wireless receiver backend using GNU radio
Digital electronics
Contact: J. Long   
  1. Distributed high-speed logic circuits
  2. CMOS digital design using TFT-on-glass technology
Analog electronics
Contact: J. Long   
  1. Design of a dynamic-range optimal filter for hard-disk readout, obtaining a linear phase response and a magnitude response that is maximally flat.
  2. Design of a power efficent voice coil actuator amplifier
     Supervisors:
    • Dr.ir. Wouter A. Serdijn
      (Delft University of Technology)
      Phone: 015-2781715
      e-mail:
    • Ir. D. van Loon, (SRON, Utrecht)
      Phone: 030-2535686
      e-mail:
    Location: Delft and Utrecht (parttime)

    Required discipline: Microelectronics

    Power dissipation is one of the most challenging issues in space electronics design. There are a number of reasons for this. First, the amount of power consumption directly relates to the size of the solar panels, which affect the overall spacecraft mass. Reducing the mass will lower the costs it takes to bring the spacecraft into orbit. Another reason, is that the generated heat inside the electronics is difficult to remove, due to the lack of convection in vacuum conditions. Due to the ever increasing complexity of space systems, more electronic functions are required in the same volume and same power budget, meaning that the power dissipation per electronic function should be reduced as well.

    In the field of imaging systems, and especially when the wavelength of interest is in the infrared domain, the heat generated by power dissipation can distort the measurement. One of the main contributors to the overall power dissipation in these units are the amplifiers that drive the actuators of one or multiple mirrors. At this moment these amplifiers are mainly class AB, with an efficiency of 50%. However other amplifier classes (class D) can obtain much higher efficiencies (90%).

    To reduce power dissipation in future missions requiring voice coil actuators, SRON wants to investigate the use of more power efficient amplifiers. One of these missions could be the Darwin mission, which is part of the ESA Science programme. This project aims at the detection of earth-like planets in the vicinity of a star and find an indication occurrence of life on these planets by looking at absorption bands of for example the gasses CO2 and H2O. These bands are in the infrared band (4 20 μm) which results in a stringent power dissipation requirement. The star outshines the nearby planet with a factor of a million in the infrared band, so without any measure the image would be blinded, hiding the exitence of a planet nearby. Moreover, the star will be much larger then the planet, requiring an enormous resolving power. Both these problems requires the use of interferometry; In Darwin a flotilla of six space telescopes, each of which will be at least 1.5 metres in diameter, will deflect the light coming from the planet to the central hub spacecraft, which acts as an interferometer. This way the 6 smaller mimic one larger telescope. By controlling the optical path differences in the arms of the interferometer, the system can be made sensitive to a particular distance from the star. This way the 6 beams of star light could interfere destructively, removing the light of the star as much as possible. This is known as nulling interferometry.

    Principle of nulling inferometry (Coutesy of ESA)

    The 6 optical paths can be controlled by integrating an optical delay line in each light path of the interferometer. An optical delay line is an instrument that typically consists of one fixed and one moveable mirror, which is used to in- or decrease the optical path length. The mirror could for example be moved by a voice coil actuator driven by a power efficient amplifier.

    Impression of Darwin. Flotilla of 6 telescope satellites along with a central hub and communication satellite (foreground). (Courtesy of ESA)

    Requirements on the power efficient voice coil actuator amplifier:
       Power output  : < 1W
    Efficiency  : > 80%
    Bandwidth  : 15 kHz

    The other requirements can only be derived the total system concept. This includes cabling (>10m) and voice coil load (variable inductance). This derivation will be part of the assignment.
  3. Design of a noise free oscillator with tunable frequency and amplitude
     Supervisors:
    • Dr.ir. Wouter A. Serdijn
      (Delft University of Technology)
      Phone: 015-2781715
      e-mail:
    • Ir. D. van Loon, (SRON, Utrecht)
      Phone: 030-2535686
      e-mail:
    Location: Delft and Utrecht (parttime)

    Required discipline: Microelectronics

    the National Institute for Space Research, has started its research on transition edge sensors (Tess) in 1996 and established a leading role in this field in the world. In a TES, a superconductor is voltage-biased in its narrow transition (typically 1 mK wide) from the superconducting to normal phase. Absorbed radiation increases the detector temperature leading to an increase of the TES resistance which is measured as a reduction of current with a low temperature SQUID amplifier (Superconducting Quantum Interference Device). Voltage biasing of a TES leads to so-called electro-thermal feedback (ETF) which counteracts excursions from the set point. ETF increases the thermal response speed and linearizes the output signal. Biasing can be realized using a dc-voltage source or using an ac-source (as in the case of Frequency Division Multiplexing, FDM). Typical operation temperatures are around 100 mK. TES-based detectors can be optimized for single photon microcalorimetry, for use in high energy astrophysics, or as bolometric detectors for sub-mm radiation; the latter route is pursued by SRON and Cardiff University.

    TES-based detectors are maturing rapidly and imaging microcalorimeter arrays using TES detectors are expected to play a major role in future spectroscopic X-ray missions, aiming at spatially resolved medium- to high-energy astrophysics studies of
    • The formation and evolution of the first gravitationally bound, dark-matter-dominated systems (i.e. small galaxy groups)
    • The evolution of metal synthesis to the present epoch, through the hot Intra-Cluster medium
    • The detection of massive black holes in the earliest Active Galactic Nuclei and estimation of their mass and spin, and
    • A study of the filamentary structure of Warm Hot Intergalactic Medium (WHIM), i.e. its mass, density, temperature and metallicity.
    At this moment SRON is preparing for the ESA's X-ray Evolving Universe Spectroscopy Mission (XEUS) that will employ imaging TES-based spectrometer arrays. Development work now focuses on fabrication and readout of an imaging spectrometer. The baseline design is an array of pixels. For the readout of this array of TES based microcalorimeters a single SQUID amplifier per pixel is not attractive for several reasons. The most compelling reason is formed by the heat load of the wiring in combination with the heat load of the SQUIDs, as the available cooling power is limited. Hence, the scalability of the electronic readout of imaging arrays of microcalorimeters depends largely on the achievable level of multiplexing. Multiplexing involves reading out multiple pixels with a single SQUID current amplifier.
    Multiplexing requires separation of the individual signals in either time or frequency space. The first system, generally referred to as time domain multiplexing (TDM), is being developed at NIST in the U.S. and is optimal for the read-out of microcalorimeters with effective decay times of approximately 500 s or longer. For decay times shorter than 500 s, as is foreseen in XEUS, the second system, referred to as frequency domain multiplexing (FDM) is regarded as optimal. FDM for the read-out of microcalorimeter arrays is a novel concept; it is being developed by SRON in collaboration with VTT Microsensing in Finland.

    In FDM a TES is voltage biased with a sinusoidal bias voltage. The current through the TES is measured with the SQUID; if the resistance of the TES changes, as result of the absorption of an X-ray photon, the output signal will be amplitude modulated.

    Detector principle
     
      
    FDM readout principle Xeus artist impression (Courtesy of ESA)

    In order to establish FDM for the read-out of an array of TES microcalorimeters they are operated at different bias frequencies and the TESs modulate the bias current. The output of the TESs is connected to a single SQUID; in this way signals are separated in frequency space and can be amplified by a single SQUID. The signal from each detector can be retrieved using standard demodulation techniques. However, measurement accuracy is in this way determined by the quality of the modulation and demodulation process and is determined by the oscillator design. The oscillator requirements are derived from the XEUS overall system requirements.
  4. CMOS analog design using TFT-on-glass technology
Biomedical electronics
Contact: W.A. Serdijn   
  1. Design and implementation of a new bladder volume assessment method
    Supervisors:
    • Dr.ir. Wouter A. Serdijn
      (Delft University of Technology)
      Phone: 015-2781715
      e-mail:
    • Ir. E. Merks, (DxU / ErasmusMC, Rotterdam)
      Phone: 010-4089358
      e-mail:

    Location: Experimental Echocardiography, ErasmusMC, Rotterdam

    Required discipline: Microelectronics
    Payment: €250,-/month

    Background
    In several clinical situations, i.e. the intensive care, the recovery room, and the urology department, it is often necessary to know the bladder filling. Catheterisation is the golden standard for bladder volume assessment due to its accuracy and reliability. However, it has got major disadvantages. Besides the fact that catheterisation is not comfortable for the patient, it is invasive and might produce infections and traumas. In an attempt to reduce the number of catheterisations, thus reducing the chance of infections, alternative measurement techniques are addressed. One possibility is the use of ultrasound.

    A device called "BladderscanTM" by the company Diagnostic Ultrasound Corp. DxU) is specifically developed for bladder volume assessment with ultrasound. It is non-invasive and is claimed to be as accurate as catheterisation. Although the BladderScanTM is extremely useful within the clinic, it is less suitable for personal use due to complex handling and device cost.

    To make bladder volume assessment suitable for personal use, new ultrasound techniques are being investigated at the Experimental Echocardiography department from the ErasmusMC in Rotterdam. 

    Main graduation activities
    For this graduation project, the student is asked to design and implement a functional prototype for a new ultrasound bladder volume assessment technique. The main focus will be on the transducer front-end, i.e. the transducer interface. This includes the design and implementation of a transducer-specific receive-amplifier, the implementation of a transmit/receive-switch, and the design and implementation of a signal processing circuit. Depending on the progress made on these subjects and the available time left, the student will be challenged to implement a transmit circuit that is capable of generating high-voltage transmit bursts to drive the transducer. During this project the student will gain knowledge of medical ultrasound applications, perform acoustic measurements with experimental setups (in-vitro), and will attend/ perform clinical measurements at the urology department.
  2. Design of a dynamic-range optimal spatial filter for edge detection in artificial retinae.
  3. Design and implementation of an ultra low-power transconductor for very low frequency applications.
    This work is related to the activity described here.
  4. Design of a switched capacitor wavelet filter to analyze intra-cardiac electrograms.
    The filter implements the wavelet transform in an (sampled-data) analog way and will be optimized with respect to dynamic range and power consumption. This work is related to the activity described here.
  5. Pacemaker front-end
    In the field of Biomedical Electronics, our group has recently succeeded in mapping a continuous wavelet transform (CWT) on silicon. First simulation results are promising and expectations are high. Currently, we are working towards application of the WT in the sense amplifier of pacemakers, the front-end that analyses the intrinsic heart signal for abnormalities. For this project we are looking for an M.Sc. student that will help us in designing an ultra-low power (100 nW) pacemaker front-end based on the CWT. The work is done in close collaboration with Medtronic Bakken Research Centre and the University of Maastricht.
Structured Electronics Design
Contact: C. Verhoeven   
  1. Design of a negative-feedback class-B amplifier.
    For ultimate low power, the active part in amplifiers should be class-B biased, i.e., zero bias. To keep the level of distortion acceptable, negative feedback is applied. Consequently, a nonlinear dynamic feedback loop is obtained. This project focuses on the design of the amplifier and ascertaining stability measures via newly developed techniques (start after October 1, 2003).
  2. Low-distortion amplifiers.
    Predicting distortion in electronics is a tough job. Recently, a simple but relatively accurate model was developed to predict the distortion of a single amplifying stage. This project aims on generalizing the obtained result to a cascade of stages and designing test circuits for verifying the models. (start after October 1, 2003).
  3. Current sources and design for robustness.
    As a result of higher integration densities integrated circuits on the same die interfere with each other via the substrate. Within a design methodology this additional "input" should be accounted for. This project aims on designing current source with a relatively low sensitivity for these substrate signals. (start after October 1, 2003).
  4. Design of voltage-reference sources.
    Voltage-reference source are widely used basic circuits. Key issue is finding an on-chip physical reference to refer the voltage to. MOS transistors exhibit a very specific bias point, for which the temperature dependency of the gate-source voltage is zero. For this very simple characteristic an ingenious circuit needs to be developed. This project is in cooperation with Professor I. Filanovsky from the University of Alberta, Canada.
Nanoelectronics
Contact: Jaap Hoekstra   
In the area of nanoelectronics, projects are available on analyses and synthesis of circuits including tunneldiodes, the SET transistors, and Carbon nanotubes:
  1. 1/f noise in the tunneldiode.
    Nowadays tunneldiodes can be manufactured in technologies compatible with current VLSI (SiGe). Measurements on the such a tunneldiode shows a specific area of large noise. It is suspected that this is 1/f noise. Within the project you can measure the noise behavior and find an adequate description.
  2. Energy dissipation in the tunneldiode.
    It is well known that the tunneldiode dissipates energy. In the project you study the dissipating mechanism by considering the tunneling current to be formed by individual electron tunneling events.
  3. Interpretation of measurements on the SET electron pump.
    The nanoelectronics group recently developed simulation tools to analyze circuits that include both standard components and tunnel junctions. Measurements on the single-electron tunneling pump can now be analyzed for the first time.
  4. Comparison of SPICE Models of SET Transistors.
    Recently various SPICE models have been published for the SET Transistor. The models are important if you want to combine the nanoelectronic SET transistor with components in an existing technology. This project aims at understanding these models and finding their differences and usefulness.

Blackboard
Blackboard:  The central informaton site of the University, for up to date information about the courses, just like date and time of it and the examinations.