• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.24 no.12
• /
• pp.1120-1127
• /
• 2013

In this paper, novel chip-on-chip(CoC) flip-chip bump structures using chip-on-wafer(CoW) process technology are proposed, designed and fabricated, and their microwave frequency responses are analyzed. With conventional bumps of Cu pillar/SnAg and Cu pillar/Ni/SnAg and novel Polybenzoxazole(PBO)-passivated bumps of Cu pillar/SnAg, Cu pillar/Ni/SnAg and SnAg with the deposition option of $2^{nd}$ Polyimide(PI2) layer on the wafer, 10 kinds of CoC samples are designed and their frequency responses up to 20 GHz are investigated. The measurement results show that the bumps on the wafers with PI2 layers are better for the batch flip-chip process and have average insertion loss of 0.14 dB at 18 GHz. The developed bump structures for chips with fine-pitch pads show similar or slightly better insertion loss of 0.11~0.14 dB up to 18 GHz, compared with that of 0.13~0.17 dB of conventional bump structures in this study, and we find that they could be utilized in various microwave packages for high integration density.

• Journal of the Microelectronics and Packaging Society
• /
• v.15 no.1
• /
• pp.7-18
• /
• 2008

This paper reviews the current development of an ultra-slim SiP for Radio Frequency (RF) application, in which three flip chips, additional passive components and Surface Acoustic Wave (SAW) filters are integrated side-by-side. A systematic investigation is carried out for the design optimization, process and reliability improvement of the package, which comprises several aspects: a design study based on the 3D thermo-mechanical finite element analysis of the packaging, the determination of stress, warpage distribution, critical failure zones, and the figuration of the effects of material properties, process conditions on the reliability of package. The optimized material sets for manufacturing process were determined which can reduce the number of testing samples from 75 to 2. In addition the molded underfilling (MUF) process is proposed which not only saves one manufacturing process, but also improves the thermo-mechanical performance of the package compared with conventional epoxy underfilling process. In the end, JEDEC's moisture sensitivity test, thermal cycle test and pressure cooker tests have also been carried out for reliability evaluation. The test results show that the optimized ultra-slim SiP has a good reliability performance.

• The Transactions of the Korea Information Processing Society
• /
• v.7 no.1
• /
• pp.235-244
• /
• 2000

This paper presents a low power PLL based clock geneator circuit for microprocessors. It generates 32MHz${\sim}$1GHz clocks and can be integrated inside microprocessor chips. A high speed D Flip-Flop is designed using dynamic differential latch and a new Phase Frequency Detector(PFD) based on this FF is presented. The PFD enjoys low error characteristics in phase sensitivity and the PLL using this PFD has a low phase error. To improve the linearity of voltage controlled oscillator(VCO) in PLL, the voltage to current converter and current controlled oscillator combination is suggested. The resulting PLL provides wide lock range and extends frequency of generated clocks over 1 GHz. The clock generator is designed by using $0.65\;{\mu}m$ CMOS full custom technology and operates with $11\;{\mu}s$ lock-in time. The power consumption is less than 20mW.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.4
• /
• pp.203-211
• /
• 2013

UHF RFID tag designers usually ndde the chip impedance and read power sensitivity value obtained when a tag chip is mounted on a chip pad. The chip impedance, however, is not able to be supplied by chip manufacturer, since the chip impedance is varied according to tag designs and fabrication processes. Instead, the chip makers mostly supply the chip impedances measured on the bare dies. This study proposes a chip impedance and read power sensitivity evaluation method which requires a few simple auxiliary and some RF measuring equipment. As it is impractical to measure the chip impedance directly at mounted chip terminals, some form test fixture is employed and the effect of the fixture is modeled and de-embeded to determine the chip impedance and the read power sensitivity. Validity and accuracy of the proposed de-embed method are examined by using commercial RFID tag chips as well as a capacitor and a resistor the value of which are known.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.39 no.8
• /
• pp.74-83
• /
• 2002

In order to design a huge VLSI system, the scan testing methodology by employing scan flip-flops(cells) is a popular method to test those If chips. In this case, the connection order of scan cells are not important, and hence the order can be determined in the very final stage of physical design such as cell placement. Using this fact, we propose, in this paper, a scan cell reordering method which minimizes the length of wires for scan chain connections. Especially, our reordering method is newly proposed method in the case when the scan cells are grouped according to their clock domains. In fact, the proposed reordering method reduces the wire length about 13.6% more than that by previously proposed reordering method. Our method may also be applicable for reordering scan chains that have various constraints on the scan cell locations due to the chain grouping.

• Journal of the Microelectronics and Packaging Society
• /
• v.10 no.3
• /
• pp.29-35
• /
• 2003

The Sn-based lead-free solders with varying microstructure were prepared by changing the cooling rate from the melt. Bulky as-cast SnAg, SnAgCu, and SnCu, alloys were cold rolled and thermally stabilized before the creep tests so that there would be very small amount of microstructural change during creep (TS), and thin specimens were water quenched from the melt (WQ) to simulate microstructures of the as-reflowed solders in flip chips. Cooling rates of the WQ specimens were 140∼150 K/sec, and the resultant $\beta-Sn$ globule size was 5∼10 times smaller than that of the TS specimens. Subsequent creep tests showed that the minimum strain rate of TS specimens was about $10_2$ times higher than that of the WQ specimens. Fractographic analyses showed that creep rupture of the TS-SnAgCu specimens occurred by the nucleation of voids on the $Ag_3Sn$ Sn or $Cu_6Sn_5$ particles in the matrix, their subsequent growth by the power-law creep, and inter-linkage of microcracks to form macrocracks which led to the fast failure. On the other hand, no creep voids were found in the WQ specimens due to the mode III shear rupture coming from the thin specimens geometry.