MEMS wafer-level packagingSupervisors: Katrin Persson, Sjoerd Haasl
Contact e-mail: sjoerd.haasl@imego.com This is a study that is split up into two projects. The main goal of the first project is to measure and characterize the induced stress and package hermeticity after bonding a silicon wafer to a low-temperature co-fired ceramic (LTCC) wafer. The second project focuses on demonstrating the genericness of the principle and on evaluating an alternative, more generally applicable bonding method. READ MORE
Magnetoelastic film studySupervisors: To be determined
Contact e-mail: max.erlandsson@imego.com Imego has worked for a long time developing sensors that use the magnetoelastic resonance (MER) effect, both in corporation with companies implementing sensors and for scientific purposes to discover interesting applications. We have some important tracks that need investigation. The magnetoelastic film (MEF) as a viscous measurement tool: We need to measure the viscosity in reference fluids to determine characteristics of the MEF. Acoustic detection of MEF signal. Improvements of the range under certain conditions may be possible by detecting the acoustic wave from the MEF instead of the magnetic wave. The range may be improved by reducing eddy currents in the film. We want to study how modifications of MEFs affect range. The diploma/candidate work roughly consists of creating an experimental setup, perform experiments and analyze the results. It may be possible to combine different paths and new ideas are always welcome. Technical background Magnetoelastic resonance is a phenomena that can be used in sensors to measure temperature, humidity, water absorption, number of cells etc. A magnetostrictive material has the ability to contract or elongate due to an applied magnetic field. At a certain bias point (typically a few Gauss, 1G=0.0001T) the magnetostriction is high for some ferromagnetic materials. When an external oscillating magnetic field is applied at a certain frequency fr corresponding to the acoustic resonance frequency of a magnetoelastic film (MEF), a standing wave is formed in the film and the magnetic field in turn generating from the film becomes much higher. fr in a thin MEF can change when the surface is subjected to some changes like lying in a fluid or having a thin polymer layer on top. By measuring the change in this resonance frequency (or amplitude/Q-value), it is possible to get information about the environment.
Bonding studySupervisor: Sjoerd Haasl
Contact e-mail: Sjoerd.Haasl@imego.com Encapsulation is an essential and critical part in the fabrication of micromachined inertial sensors. One of the most important factors that affect the performance of the sensor is the gas pressure in the sensor cavity. To obtain a high Q-factor, and therefore, a low noise level, the gas damping should be controlled. The pressure inside the cavity is not necessarily defined by the ambient pressure at which the cavity was sealed, as the high surface-to-volume ratio of micromachined cavities causes the pressure to be much dependent on the outgassing of the cavity walls. This outgassing is dependent on the history of the wafers prior to bonding and the parameters of the bonding technique used. Several hermetic bonding techniques exist, but the effect of the bonding technique on the pressure inside the cavity is poorly documented. An important reason for this is the difficulty to obtain reliable pressure measurements in micromachined cavities.In a previous master thesis project, resonant structures were developed that can be used to measure the pressure inside such cavities. Though these structures were developed to evaluate the pressure in cavities obtained by anodically bonding LTCC wafers to silicon wafers, the same principle can be used to evaluate other bonding methods. The thesis consists of reproducing the pressure evaluation structures, perform eutectic and fusion bonding with varying process parameters and evaluate the pressure inside the cavities. An optional study subject would be the use of getter materials to absorb any remaining gases in the cavities. Project description Study the effect of different bond techniques upon the pressure inside cavities. Two different bond techniques will be studied: Au-Si eutectic bonding and direct silicon (fusion) bonding. Different bonding parameters such as prebaking, extensive purging and annealing will be evaluated. The project would consist of the following tasks: - Reproduce pressure evaluation structures
- Design read-out electrode wafers for foundry fabrication (using the Cadence Virtuoso layout tool)
- Develop reproducible wafer bonding procedures for fusion and Au-Si eutectic bonding
- Bond pressure evaluation structures to foundry fabricated electrode wafers
- Evaluate pressure in bonded structures
- Optional task: investigate use of getter materials
Gyroscope evaluation and automated inspectionSupervisors: Erik Aderstedt, Sjoerd Haasl
Contact e-mail: sjoerd.haasl@imego.com A set of three micromachined gyroscopes forms a crucial part in Imego’s navigation-grade inertial measurement unit (IMU). The gyroscopes used for this application are off-the-shelf roll-over detection gyros that are bought in large batches as-is from the supplier. To maximize the yield, the gyros in these batches need to be evaluated prior to assembly. The current selection criterion is very stringent but non-compliance to this criterion has not been proven to correlate with the final performance of the gyroscopes in the IMU, and much can be gained by developing an improved set of criteria that can be verified on chips before they are assembled. The large number of gyros in storage requires a semi-automated setup to contact and verify the chips. The project would consist of the following tasks: - Study the correlation between parameters measurable on unmounted chips and the performance of individual gyros mounted in Imego’s IMU.
- Develop a measurement setup in Labview, Matlab or C# based on the mechanical and electrical properties of the gyros that can measure the criteria required to sort the sensors.
- Develop an automated interface to control an (existing) XY-table using Labview, Matlab or C#, such that measurements can be made. Basic image analysis might be necessary in case closed-loop control is required.
Glue studySupervisors: Katrin Persson, Sjoerd Haasl
Contact e-mail: sjoerd.haasl@imego.com Recommended background; mechanical engineering or material science. In the design of assembly concepts for inertial sensor elements very little is reported concerning the attachment of the sensor to the package. Most characterisation of the effect of die attach on the sensor element is performed in-house and is not openly disclosed in literature. For example, many high performance inertial sensors rely on active control of vibrating structures and the quality of the sensor output is usually very sensitive to external stress. It is also of importance to study the influence of harsh environmental factors, such as humidity, radiation, heat and vibration, on the adhesives themselves. To predict the quality of assembly concepts, the properties and influence of adhesive materials need to be investigated and simulated. To be able to do this, a number of thermal and mechanical parameters need to be experimentally extracted and represented using suitable material models. Even though such models are covered in literature, the actual parameters of commercial adhesives are not available. Therefore, one of the technical aims of the project is to bridge the gap between practical modelling of adhesives, optimal assembly design by simulation and verification by experiment. Project description Study the effect of the mechanical properties of different glues on the performance of the sensors. The glues vary in the presence of additives, conductivity, thermal curing requirements, etc. The project would consist of the following tasks: - Investigation of stress on sensors (initially bare dies).
- Develop mechanical data collection (find out what parameters are essential and how to measure them).
- Measure mechanical data on specific glues and develop a simulation model.
- Simulate (Comsol FEMlab) and evaluate glue effect on SensoNor’s SAR10 gyro with Imego’s ASIC.
- Simulate and evaluate glue effect on Imego’s third generation triaxial accelerometer.
AI in personal navigationSupervisor: Jan Wipenmyr
Contact e-mail: jan.wipenmyr@imego.com Imego is working on a navigation system to track the position and orientation of a fireman using an inertial navigation system. In this project we aim to use Artificial Intelligence algorithms to improve the performance of the navigation system as well as to provide information on how the fireman is moving. You will build on a previous study done at Imego to evaluate which AI algorithm(s) are best to use. These will be used to develop software that identifies a range of human motions in data produced by Imego’s inertial devices which output full 3D data from gyroscopes, accelerometers, magnetometers and inclinometers. The algorithm must be trained to work reliably on multiple test subjects. Finally you will help to integrate this classification algorithm into an inertial navigation engine. You will most likely be in contact with a fire department in the Göteborg region for advice and you will also demonstrate the resulting software for them. You should be interested in Artificial Intelligence problems and be familiar with the most common types of algorithms. Interest in signal processing is a bonus as is an interest in navigation, but not necessary. Most work will be done using Matlab. |