Wireless systems with emerging applications are leaning towards small size, light-weight and low cost. Another trend for these wireless devices is that new applications and functionalities are being added without increasing the size of the device. To accomplish this, individual components must be miniaturized and the system should be designed to maximize the integration of the individual components. The high level of 3D integration feasible in system on package design (SoP) concept can fulfill the latter requirement.
Bandpass filters are important components on all wireless systems to reject the unwanted signals and reduce interference. Being mostly implemented with passive and distributed components, bandpass filters take considerable space in a wireless system. Moreover, with emerging bands and multiple applications encompassed in a single device, many bandpass filters are required. The miniaturization related to bandpass filters can be approached by three main ways: (1) at the component level through the miniaturization of individual bandpass filters, (2) at the system level through the use of tunable filters to reduce the overall number of filters, and (3) at the system level through the high level of integration in a 3D SoP platform. In this work we have focused on all three aspects of miniaturization of band pass filters mentioned above.
In the first part of this work, a low frequency (1.5 GHz global positioning system (GPS) band) filter implemented through 3D lumped components in two leading SoP technologies, namely low temperature co-fired ceramic (LTCC) and the liquid crystal polymers (LCP) is demonstrated. The miniaturized filter is based on a second order topology, which has been modified to improve the selectivity and out-of-band rejection without increasing the size.
Moreover, for the case of LCP, the filter is realized in an ultra-thin stack up comprising four metallization layers with an overall thickness of only 100 _m. Due to its ultra-thin structure, the LCP filter is ten times smaller size as compared to the filters reported in published work. The filter is exible and, therefore, suitable for conformal applications.
In the second part of this work, relatively higher frequency (Ku band) distributed bandpass filter is presented which can be tuned through an applied magnetic field. This has been realized in a relatively new LTCC tape with magnetic properties, known as ferrite LTCC. Traditionally, magnetically tunable filters require large external electromagnets or coils, which are non-integrable to typical planar circuit boards and are also inefficient. To demonstrate high level of integration, completely embedded windings realized in multiple layers of LTCC have been used instead of the external coils. As a result, the presented bandpass filter is several orders of magnitude smaller that the reported ones. Aside from reducing the size, the embedded windings based design is more efficient than the external coils because it can avoid the demagnetization effect (fields lost at air-ferrite interface) and thus require much smaller bias fields for tunability.
Though the embedded windings bring in a number of advantages as mentioned above,
the currents passing through these windings generate considerable heat which can inuence the performance of the microwave structure (bandpass filters in our case). This has never been studied before fro Ferrite LTCC based designs with embedded windings. In this work, the effect of the heat generated by these windings has been investigated. It has been found that this self-heating effect inuences the tunability of the filters considerably so it must be estimated at the design stage. Therefore, a strategy to simulate this effect has been developed. The resultant simulations agree well with the measurements verifying the simulation strategy. The designs presented in this work demonstrate the feasibility of realizing highly integrated, miniaturized and tunable filters in SoP platform which are very suitable for modern and futuristic small form factor and slim wireless devices.
|Date of Award||Apr 2015|
- Computer, Electrical and Mathematical Science and Engineering
|Supervisor||Atif Shamim (Supervisor)|
- Bandpass Filter