• Journal of the Institute of Electronics and Information Engineers
• /
• v.50 no.8
• /
• pp.128-135
• /
• 2013

This paper presents a high dimming ratio Power-LED driver IC with high power which is capable of controlling LEDs. In order to accomplish a high dimming ratio LED driver, the preloading inductor current methodology is proposed for the power stage of the proposed method to achieve the fast transient response time during the Power-LED load switching. The information containing the current flowing on the LEDs can be utilized to predict the amount of the current on the inductor. The minimum LED current rising time of existing high dimming ratio Power-LED driver is limited by $3{\mu}s$, however that of the proposed high dimming ratio Power-LED driver is reduced about 1/10. The LED driver is implemented with 0.35um 60V BCDMOS 2-poly 4-metal process. The measurement results show that the proposed LED driver system features the minimum rising time as small as 240ns at the dimming frequency of 1KHz with a 12V of input voltage, nine white LEDs and 353mA of LED current. The LED rising time and power conversion efficiency of the chip are measured to be 240ns and 93.72%, respectively.

• Journal of Hydrogen and New Energy
• /
• v.10 no.1
• /
• pp.27-40
• /
• 1999

By hydrogen induced planetary ball milling process, $Mg_2NiH_x$ hydrogen absorbing materials were successfully alloyed mechanically at room temperature, using pure Mg and Ni chips. The Mg & Ni chips were mixed by 45:55 weight ratio and Mechanical Alloying(M.A.) was carried out : the hydrogen pressure induced in the jar was varied from 1 to 20 bars and the M.A. times were 24 and 48 hours. The XRD results revealed that the homogeneous $Mg_2NiH_x$ was incresed with the hydrogen pressure increasing, and that $MgH_x$ was detected by unalloyed Ni chips. The shape and size of the mechanically alloyed particles didn't depend on the induced hydrogen pressure. The results of TGA showed that the hydrogen quantities of $Mg_2NiH_x$ has 1.1~3.9 wt%.

• Korean Journal of Environmental Agriculture
• /
• v.18 no.4
• /
• pp.321-326
• /
• 1999

This study was performed to evaluate the quality of produced compost and to analyse the change of a component during the reduction compost according to the input volume. The volume of pilot scale used in this study was about 300㎥. The pile of 2m width, 20m length and 1.2m height was constructed. Woodchip was used as bulking agent to enhance pore volume of composting bay and to control water content of food waste in starting time. Food waste was turned using a mechanical tumer twice a day. The result are as follows : In these cases of input volume of $2m^3$ and $3.5m^3$, temperature of composting pile was maintained over $60^{\circ}C$ and water content was 43.6% and 47.2%, respectively. It was proved that microorganisms activity was maintained high in each input volume. After operation of step 1 and step 2, pH and organic matter in the final compost were 6.2, 6.6 and 84.3%, 79.6%, respectively. Cation concentration such as $K_2O,$ CaO and NaCl was accumulated in the compost during the composting period. NaCl concentration in the final compost was 4.62%, 4.92%, respectively. Hence, If was recommended that this compost should be applied to others expect agricultural area or mixed with a low concentration other compost. In the steps 1, input volume of $2m^3$, heavy metal concentration of Pb, Cu, Cr, Ni. Cd were 37.82㎎/㎏, 56.87㎎/㎏, 9.8㎎/㎏. 22.21㎎/㎏ and 3.69㎎/㎏, and 44.55㎎/㎏, 95.54㎎/㎏, 12.22㎎/㎏, 24.94㎎/㎏, and 3.86㎎/㎏ in the step 2.

• Analytical Science and Technology
• /
• v.11 no.2
• /
• pp.112-119
• /
• 1998

A Siemens SemiQuant (SSQ) 3000 program, a precalibrated 'standardless' analytical program handling up to 90 elements, was evaluated for the fast analysis of various types of reference materials using a wavelength dispersive X-ray spectrometer. Various types of standard reference materials such as metal discs, metal chips, and geological materials in powder form were analysed and it took 23 minutes of measuring time for 75 elements. Measurements of geological reference materials using different sampling methods were carried out and their data were interactively evaluated. The analysis of materials of a known matrix concentration such as stainless steels provided higher precision value compared to totally unknown samples. The analyses of materials prepared as pressed pellets or fused glass beads provided higher precision values compared to the measurement of loose powders with a foil on the sample surface and helium operation, though their sampling procedures were more complicate and took more time. Since very light elements such as boron, carbon, and oxygen have a strong influence on the matrix effects and also on the calculation of effective matrix corrections, the rhodium Compton check was applied to verify the reliability of the defined light element concentrations of light matrix materials and the defined major sample compounds. Failure of defining correct matrix resulted in an unoptimized matrix correction and therefore in the wrong calculation of the element concentration.

• Journal of Broadcast Engineering
• /
• v.25 no.2
• /
• pp.200-207
• /
• 2020

Recently, deep neural networks such as CNN are showing excellent performance in various fields such as image classification, object recognition, visual quality enhancement, etc. However, as the model size and computational complexity of deep learning models for most applications increases, it is hard to apply neural networks to IoT and mobile environments. Therefore, neural network compression algorithms for reducing the model size while keeping the performance have been being studied. In this paper, we apply few compression methods to CNN models and evaluate their performances in the embedded environment. For evaluate the performance, the classification performance and inference time of the original CNN models and the compressed CNN models on the image inputted by the camera are evaluated in the embedded board equipped with QCS605, which is a customized AI chip. In this paper, a few CNN models of MobileNetV2, ResNet50, and VGG-16 are compressed by applying the methods of pruning and matrix decomposition. The experimental results show that the compressed models give not only the model size reduction of 1.3~11.2 times at a classification performance loss of less than 2% compared to the original model, but also the inference time reduction of 1.2~2.21 times, and the memory reduction of 1.2~3.8 times in the embedded board.

• Asian-Australasian Journal of Animal Sciences
• /
• v.24 no.9
• /
• pp.1184-1191
• /
• 2011

We attempted to generate a linkage map using Illumina Porcine 60K SNP Beadchip genotypes of the $F_2$ offspring from Korean native pig (KNP) crossed with Yorkshire (YS) pig, and to identify quantitative trait loci (QTL) using the line-cross model. Among the genotype information of the 62,136 SNPs obtained from the high-density SNP analysis, 45,308 SNPs were used to select informative markers with allelic frequencies >0.7 between the KNP (n = 16) and YS (n = 8) F0 animals. Of the selected SNP markers, a final set of 500 SNPs with polymorphic information contents (PIC) values of >0.300 in the $F_2$ groups (n = 252) was used for detection of thirty meat quality-related QTL on chromosomes at the 5% significance level and 10 QTL at the 1% significance level. The QTL for crude protein were detected on SSC2, SSC3, SSC6, SSC9 and SSC12; for intramuscular fat and marbling on SSC2, SSC8, SSC12, SSC14 and SSC18; meat color measurements on SSC1, SSC3, SSC4, SSC5, SSC6, SSC10, SSC11, SSC12, SSC16 and SSC18; water content related measurements in pork were detected on SSC4, SSC6, SSC7, SSC10, SSC12 and SSC14. Additional QTL of pork quality traits such as texture, tenderness and pH were detected on SSC6, SSC12, SSC13 and SSC16. The most important chromosomal region of superior pork quality in KNP compared to YS was identified on SSC12. Our results demonstrated that a QTL linkage map of the $F_2$ design in the pig breed can be generated with a selected data set of high density SNP genotypes. The QTL regions detected in this study will provide useful information for identifying genetic factors related to better pork quality in KNP.

• The Korean Journal of Mycology
• /
• v.36 no.1
• /
• pp.26-30
• /
• 2008

This study was carried out to develop artificial culture method of Grifola frondosa in polypropylene bag. To find out the favorable substrate formulation of G. frondosa, physicochemical conditions were investigated. The T2 formulation (55 : 25 : 12 : 8) mixing ratio of oak sawdust, oak chip, dried bean-curd refuse and wheat bran showed the shortest time to complete the crop cycle and the highest yield (weight of fresh mushrooms harvested at maturity). Those physicochemical conditions were pH 4.3, 2.4% crude fat contents, 54.1% total carbon, 1.42% total nitrogen, 38.1 C/N ratio, 75.5% porosity and $0.21\;g/cm^3$ bulk density. Among the physicochemical factors, pH, crude fat and total nitrogen may affect mushroom yield. Therefore, development of favorable substrate would be benefit to increase production efficiency of G. frondosa, mushrooms and be commercial potential.

• The Journal of Korean Institute of Electromagnetic Engineering and Science
• /
• v.20 no.1
• /
• pp.75-82
• /
• 2009

In this paper, we proposed a U-shaped broadband RFID tag antenna with a parasitic element operating at UHF band. The proposed tag antenna consists of a U-shaped half wavelength dipole antenna and an inverse U-shaped parasitic element inside the U-shaped dipole antenna. In order to have good impedance matching, the commercial tag chip is attached to the lower center of the rectangular shaped feed. On the condition of VSWR<2, the tag antenna had the measured bandwidth of 4.96 % from 882 to 927 MHz and showed the gain deviation of less than 3.16 dB. On the condition of VSWR<5.8, the tag antenna satisfies the worldwide UHF RFID bandwidth and is showed the gain deviation of less than 5.07 dB. The minimum gain deviation characteristic appears near the center of bandwidth which minimizes variation of gain deviation characteristic with respect to the frequency.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.05a
• /
• pp.443-446
• /
• 2003

This paper implement Remote Meter Reading System which is used Bluetooth and SMS in CDMA. This System propose system which tan detect a total amount of gas, electricity or water without a meterman, at home BlueTooth is a close range wireless communication technology which uses a wireless frequency 2.4GHz and has a high trust and self error correction technology according to a low power consumption quality and a high-speed frequency hopping. This makes get a high trust concerning a data transmission than an existing modem. In addition, though wireless modem is restricted by a minimal of a wireless terminal, it will be possible to coincide with the function of the portable with the low power consumption quality by using Bluetooth. And as the system on a chip of module progresses, the possibility of the small size is present. Nowadays, SMS Protocol in CDMA for have a network function based on PPP in CDMA Phone. The proposed Remote Meter Reading System to get more nobility, efficiency, and have good function. SMS Protocol in CDMA have profits which is low power, low cost, and low microwave output.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.15 no.4
• /
• pp.913-920
• /
• 2011

In this paper, a logic process based 1-kbit EEPROM IP for RFID tag chips of 900MHz is designed. The cell array of the designed 1-kbit EEPROM IP is arranged in a form of four blocks of 16 rows x 16 columns, that is in a two-dimensional arrangement of one-word EEPROM phantom cells. We can reduce the IP size by making four memory blocks share CG (control gate) and TG (tunnel gate) driver circuits. We propose a TG switch circuit to supply respective TG bias voltages according to operational modes and to keep voltages between devices within 5.5V in terms of reliability in order to share the TG driver circuit. Also, we can reduce the power consumption in the read mode by using a partial activation method to activate just one of four memory blocks. Furthermore, we can reduce the access time by making BL (bit line) switching times faster in the read mode from reduced number of cells connected to each column. We design and compare two 1-kbit EEPROM IPs, two blocks of 32 rows ${\times}$ 16 columns and four blocks of 16 rows ${\times}$ 16 columns, which use Tower's $0.18{\mu}m$ CMOS process. The four-block IP is smaller by 11.9% in the layout size and by 51% in the power consumption in the read mode than the two-block counterpart.