EBG Structure를 이용한 PCB에서의 SSN 저감 기술에 관한 연구

Abstract

The simultaneous switching noise (SSN) cause the degradation of signal integrity (SI). Recently, with a trend of high density packing and small size design in multilayer printed circuit boards (PCB), several methods to reduce the SSN have been investigated in many papers

decoupling and embedded capacitors, isolated power and ground plane structures, and electromagnetic band gap (EBG) structures.A decoupling capacitor between power and ground planes is shown to be effective in reducing the SSN. However, in the high frequency, using a decoupling capacitor can cause many serious problems for SI, because the capacitor is usually represented as parasitic component such as an equivalent series inductance (ESL). An alternative to the decoupling capacitor is an embedded capacitor, which is implemented by using high dielectric constant material or by reducing the dielectric thickness in the PCB, but the cost is very high. Also, dividing power or ground plane structure in PCB design are often used to protect a sensitive digital circuit from SSN without adding extra components. However, such structures are difficult to design to routing restriction by the divided planes. It is also difficult to control coupling noises due to the increment of loop inductances.For the suppression of SSN, the best known method is the electromagnetic band gap (EBG) structure, which works as the band stop filter without adding capacitive or inductive components. The most common EBG structures are mushroom EBG structure and uni-planar EBG structure. Generally, these designs can be readily designed and applied to obtain good electrical performance in a high speed layout. However, conventional EBGs have the design constraints in compact size devices such as mobile phones.To overcome the size limitation for the compact design, this thesis proposes a modified EBG structure. Proposed design has 1 mm 1 mm unit cell and a good SSN performance of at least -60 dB between 0.8 GHz and 3.4 GHz. Experimental results prove the effectiveness of this method in improving the SI performance. We also investigate two other types of modified mushroom EBG structures which are effective in reducing the number of PCB layers and suppressing the electromagnetic interference (EMI), respectively. They showed same SSN performances as the existing mushroom EBG.