Suspended conductor: a proposed configuration to enhance performance

Progress of multi-functional wireless products has led to a continuously increasing demand for high performance, low cost, miniaturized and lightweight circuits which can fit in small hand-held devices and which can work for a long time with small batteries. Low Temperature Co-fired Ceramic (LTCC) technology, with an almost arbitrary number of low-loss dielectric layers and high conductivity metals like silver and gold, is well-suited to provide solutions to such demands. The intrinsic 3D capability of LTCC technology makes it a superior choice to design compact lumped element passive components and circuits. It has already been shown in the literature that LTCC lumped element components do not suffer from the integration limitation of alternative solutions like surface mounted technology (SMT) components in ultrasmall circuits (Wu et al., 2004; Sutono et al., 2001). Furthermore, LTCC lumped components offer the possibility of providing continuous elements values rather than discrete ones since they can be properly dimensioned during the design process. The monolithic integration of different lumped LTCC components in single design can also be an advantage for size and part-count reduction.

Since the first appearance of lumped element components in LTCC (Eurskens et al., 1998), considerable research has been carried out in this area. While certain researchers focused on the realization of different RF circuits (Mousavi & Kouki, 2016; Brzezina et al., 2009; Brzezina & Roy, 2014; Zhou et al., 2014b; Joshi & Chappell, 2006), others explored the ways to improve the performance of LTCC lumped element components (Aliouane et al., 2011; Mousavi & Kouki, 2015; Eun et al., 2004; Bahl, 2003). The performance of lumped components in LTCC like any other fabrication technology is limited by loss factors and parasitic capacitors (or inductors). Since, metals with highest conductivity are already available for LTCC circuits, the main focus to enhance the performance has been on reducing dielectric loss and designing structures with lower parasitic values. In (Aliouane et al., 2011), the performance of a solenoid RF inductor has been compared to a spiral inductor, and it has been shown that the former one can provide better quality factor (Q). However, there is almost no improvement in self-resonance frequency (SRF) reported. An air-cavity has been added under a solenoid inductor in (Mousavi & Kouki, 2015) to increase SRF and Q. Using this technique, about 25 percent improvement in Q has been achieved for an inductor in UHF band. The impact of having several grounding vias on the performance of a capacitor was also investigated in (Mousavi & Kouki, 2015) and and a 23 percent improvement in SRF was reported. An air-cavity has also been used under a spiral inductor in (Eun et al., 2004) where an inductor having an inductance of 2.62 nH, a Q of 51 and a SRF of 9.1 GHz was reported. In all of the above cited studies where an air-cavity was used, an extra layer of the ceramic had been left between the air-cavity and inductor’s arms. This extra layer is unavoidable for the mechanical integrity of their design and it negatively impacts the performance improvement provided by this technique. Furthermore, to the best knowledge of the authors, all reported LTCC passive lumped elements are limited to the frequencies below 6 GHz.

Suspended conductor: a proposed configuration to enhance performance

Fabrication process

The use of air cavities in inductor design, has been tested in (Mousavi & Kouki, 2015; Eun et al., 2004) and shows promising results. However, in these works, an extra layer of ceramic was left under the inductor to mechanically support the inductor’s arm. While such air cavities help reduce Csub and G, having this extra layer limits the performance of the designed inductors. To achieve further reduction in Csub and G, this extra layer must be removed resulting in a suspended conductor. This leads to structural challenge of how to support the conductor. To this end, we propose a new technique that exploits the flexibility of LTCC technology. Indeed, by modifying in the standard fabrication process, the realization of these floating conductors becomes feasible through the following steps.

Step 1:  The suspended conductor achieves its structural integrity by being printed on both sides of the single sheet with interconnection between the two sides achieved through an array of vias. While via punching and fabrication follows the standard LTCC process, double sided printing required minor process adjustment. The resulting structure works like a virtually thick line at frequencies at which the spacing between the vias is considerably smaller than the corresponding wavelength.

Step 2: the next step is the drilling of internal cavities in each layer. After that, these layers are stacked on top of each other. All the internal cavities are filled with fugitive tape to maintain the structural integrity of green tape stack during the lamination process. The fugitive tape used for this purpose is the LTCC carbon tapes (TCS-CARB-1).

Step 3: the firing process is modified to allow the fugitive tape to completely burn out in order to obtain a clean internal cavity . This profile, compared to the standard profile, has been prolonged in three sections: (i) at 350oC for 60 minutes (ii) at 710oC for 60 minutes and (iii) at 850oC. These three plateaus in firing recipe allow burning completely out all binder components and residual carbons before LTCC pore closure and leave a clean inner cavity .

Validation

To validate the performance of the proposed solution and assess the performance gains that can be achieved, two LTCC inductors of 2.5 nH and 5 nH, were designed and fabricated in two different versions: (i) conventional and (ii) suspended, using Dupont 9K7 green tape with permittivity of 7.1 and loss tangent of 0.001. Fig. 1.5 shows photographs of the two suspended inductors. All inductors were measured using Rohde and Schwartz ZVA vector network analyzer.