Patents by Inventor Sheng-Fu Chuang
Sheng-Fu Chuang has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 9973235Abstract: This invention discloses a signal receiving device for Ethernet and a control method thereof. The signal receiving device includes a gain control circuit, an alien near-end crosstalk canceller, a noise canceller, and a DFE. The gain control circuit adjusts an input signal of the signal receiving device according to a setting parameter. The alien near-end crosstalk canceller cancels an alien near-end crosstalk interference. The noise canceller uses a first filter to cancel noises. The DFE uses a second filter to cancel an inter-symbol interference of the input signal. The method includes steps of: temporarily stopping the gain control circuit from updating the setting parameter before a seed collision occurs, and temporarily stopping one of the noise canceller and the decision feedback canceller from updating the first filter coefficient of the first filter or the second filter coefficient of the second filter temporarily during the seed collision.Type: GrantFiled: July 14, 2015Date of Patent: May 15, 2018Assignee: REALTEK SEMICONDUCTOR CORPORATIONInventors: Sheng-Fu Chuang, Liang-Wei Huang, Ching-Yao Su, Hsuan-Ting Ho
-
Patent number: 9879970Abstract: A method for estimating cable length in an Ethernet system and a receiver thereof are applicable to an Ethernet system. The method for estimating cable length includes obtaining a channel tap from channel information of a feedback equalizer in the Ethernet system and estimating a cable length according to the channel tap, a first coefficient and a constant.Type: GrantFiled: July 9, 2014Date of Patent: January 30, 2018Assignee: REALTEK SEMICONDUCTOR CORP.Inventors: Shih-Wei Wang, Liang-Wei Huang, Ching-Yao Su, Sheng-Fu Chuang
-
Patent number: 9859706Abstract: A united power module having an external input and a first input is provided. The united power module includes a load block, a protection circuitry and a detection block. The detection block includes an analog circuitry. The load block and an end of the protection circuitry are electrically coupled to the external input. The load block, the first input, and another end of the protection circuitry are electrically coupled to the detection block. The load block is supplied by an external power received from the external input. The protection circuitry receives the external power and outputs an auxiliary power. The analog circuitry receives a first power from the first input under a normal power supply state, or receives the auxiliary power under an auxiliary power supply state.Type: GrantFiled: September 25, 2015Date of Patent: January 2, 2018Assignee: REALTEK SEMICONDUCTOR CORP.Inventors: Ching-Yao Su, Liang-Wei Huang, Hsuan-Ting Ho, Sheng-Fu Chuang
-
Patent number: 9824055Abstract: A control method applied to an Operating-Mode Finite-State-Machine (OPFSM) arranged for deciding a behavior of a first port of an apparatus includes: controlling the OPFSM to enter a second local state from a first local state and controlling the first port to send a signal with a wakeup pattern to a link partner of the first port when the state of the OPFSM is the first local state, and a wakeup request bit is a first local value.Type: GrantFiled: March 12, 2015Date of Patent: November 21, 2017Assignee: Realtek Semiconductor Corp.Inventors: Hsuan-Ting Ho, Liang-Wei Huang, Ching-Yao Su, Sheng-Fu Chuang
-
Patent number: 9800290Abstract: A channel detection method for an echo canceller of a communication device is provided. The method includes the following steps. A first detection signal is transmitted to an end of a channel coupled to the communication device. A plurality of taps corresponding to a reflected signal of the first detection signal are received by an echo canceller at the end of the channel. The taps corresponding to the reflected signal are compared with a reference value corresponding to each of the taps so as to determine whether each of the taps is larger than or equal to the corresponding reference value. When the tap is determined to be larger than or equal to the reference value corresponding to the tap, the tap and a position of the tap are recorded.Type: GrantFiled: January 25, 2016Date of Patent: October 24, 2017Assignee: REALTEK SEMICONDUCTOR CORP.Inventors: Shih-Wei Wang, Liang-Wei Huang, Ching-Yao Su, Sheng-Fu Chuang
-
Patent number: 9768982Abstract: A signal transmission method is provided. The method includes the following steps. First, an output signal filtered from a square wave signal by a high pass filter is received. Then, the output signal is sampled at each time interval, so as to generate the sampled values. Next, the sampled values are compared with at least one threshold to generate comparison results. Finally, a signal pattern corresponding to the output signal is identified according to the comparison results.Type: GrantFiled: October 22, 2015Date of Patent: September 19, 2017Assignee: REALTEK SEMICONDUCTOR CORP.Inventors: Sheng-Fu Chuang, Liang-wei Huang, Hsuan-Ting Ho, Ching-Yao Su
-
Patent number: 9680673Abstract: A communication system comprises a packet stream transforming unit, a mapping unit and a transmission unit. The packet stream transforming unit is configured to receive a 4-bit packet stream and transform the 4-bit packet stream into a 6-bit packet stream. The mapping unit is configured to map the 6-bit packet stream into multiple ternary bit streams, and the mapping unit maps at least one idle symbol into the ternary bit streams according to at least one particular bit of the at least one idle symbol of the 6-bit packet stream. The transmission unit is configured to transmit the ternary bit streams to a remote communication device through a cable.Type: GrantFiled: March 4, 2016Date of Patent: June 13, 2017Assignee: REALTEK SEMICONDUCTOR CORPORATIONInventors: Heng Cheong Lao, Ta-Chin Tseng, Shieh-Hsing Kuo, Sheng-Fu Chuang
-
Patent number: 9678146Abstract: The present invention discloses a temperature insensitive testing device comprising: a transmission-end test sequence generating circuit to generate a test sequence; a transmission circuit to process the test sequence according to a transmission clock and thereby generate a test signal; a reception circuit to process an echo of the test signal and generate a digital echo signal; a correlation-value generating circuit to generate correlation values including a maximum correlation value according to the test sequence and the digital echo signal; and a decision circuit to determine whether a relation between the maximum correlation value and at least one threshold satisfies a predetermined condition and thereby generate a decision result, wherein the frequency of the transmission clock is lower than a predetermined frequency which confines the variation of the maximum correlation value to a predetermined range provided that the temperature variation of the transmission cable is within a temperature variation ranType: GrantFiled: November 3, 2014Date of Patent: June 13, 2017Assignee: REALTEK SEMICONDUCTOR CORPORATIONInventors: Ching-Yao Su, Liang-Wei Huang, Shih-Wei Wang, Sheng-Fu Chuang
-
Patent number: 9577705Abstract: This disclosure provides a network apparatus with communication ports, each connected to multiple channels, assigned a seed for eliminating the interference among the communication ports, and operable in a master or slave mode, each channel having a communication unit which comprises: an echo canceller, a near-end crosstalk (NEXT) canceller, a decision feedback equalizer, and a control unit configured for controlling the communication ports if the echo canceller and the decision feedback equalizer keep updating their filter coefficients according to the same symbol during a first pre-determined time-interval; wherein when more than two of the communication ports operate in the master mode, the control unit assigns different seeds to the more than two communication ports; and wherein when successive two of the communication ports operate in the slave mode, the control unit stops either the echo canceller or the NEXT canceller from updating their filter coefficients during a second pre-determined time-interval.Type: GrantFiled: October 29, 2014Date of Patent: February 21, 2017Assignee: REALTEK SEMICONDUCTOR CORP.Inventors: Hsuan-Ting Ho, Liang-Wei Huang, Sheng-Fu Chuang, Ching-Yao Su
-
Patent number: 9569387Abstract: A master-slave detection method includes: every single time period, utilizing a random manner for determining whether a first device is used to transmit a specific pulse signal to a second device; every single time period, utilizing a random manner for determining whether the second device is used to transmit the specific pulse signal to the first device; when the first device receives at least one portion of the specific pulse signal earlier than the second device, setting the first device as a master device, stopping the master device from sending the specific pulse signal and utilizing the master device to start transmitting a specific sequence; and setting the second device as a slave device when the second device receives the specific sequence. The at least one portion of the specific pulse signal includes continuous single pulses.Type: GrantFiled: March 10, 2014Date of Patent: February 14, 2017Assignee: Realtek Semiconductor Corp.Inventors: Ching-Yao Su, Liang-Wei Huang, Hsuan-Ting Ho, Sheng-Fu Chuang
-
Publication number: 20160301431Abstract: A signal transmission method is provided. The method includes the following steps. First, an output signal filtered from a square wave signal by a high pass filter is received. Then, the output signal is sampled at each time interval, so as to generate the sampled values. Next, the sampled values are compared with at least one threshold to generate comparison results. Finally, a signal pattern corresponding to the output signal is identified according to the comparison results.Type: ApplicationFiled: October 22, 2015Publication date: October 13, 2016Inventors: Sheng-Fu Chuang, Liang-wei Huang, Hsuan-Ting Ho, Ching-Yao Su
-
Publication number: 20160276925Abstract: A united power module having an external input and a first input is provided. The united power module includes a load block, a protection circuitry and a detection block. The detection block includes an analog circuitry. The load block and an end of the protection circuitry are electrically coupled to the external input. The load block, the first input, and another end of the protection circuitry are electrically coupled to the detection block. The load block is supplied by an external power received from the external input. The protection circuitry receives the external power and outputs an auxiliary power. The analog circuitry receives a first power from the first input under a normal power supply state, or receives the auxiliary power under an auxiliary power supply state.Type: ApplicationFiled: September 25, 2015Publication date: September 22, 2016Inventors: CHING-YAO SU, LIANG-WEI HUANG, HSUAN-TING HO, SHENG-FU CHUANG
-
Publication number: 20160269204Abstract: A communication system comprises a packet stream transforming unit, a mapping unit and a transmission unit. The packet stream transforming unit is configured to receive a 4-bit packet stream and transform the 4-bit packet stream into a 6-bit packet stream. The mapping unit is configured to map the 6-bit packet stream into multiple ternary bit streams, and the mapping unit maps at least one idle symbol into the ternary bit streams according to at least one particular bit of the at least one idle symbol of the 6-bit packet stream. The transmission unit is configured to transmit the ternary bit streams to a remote communication device through a cable.Type: ApplicationFiled: March 4, 2016Publication date: September 15, 2016Inventors: HENG CHEONG LAO, Ta-Chin TSENG, Shieh-Hsing KUO, Sheng-Fu CHUANG
-
Patent number: 9438453Abstract: The present invention discloses a network connection establishing method capable of preventing a link procedure from being strangled in a training state. An embodiment of the method comprises: a preceding step; a training step; and a following step. Said training step includes at least one of the following: counting a number, and returning to the preceding step if a local receiver is not yet ready after finishing counting the number; detecting a reception signal according to a signal detection level, and returning to the preceding step when none of the reception signal is detected; determining a number of level(s) of the reception signal, and returning to the preceding step if the number of level(s) is less than an expected level number; and comparing a signal-to-noise ratio of the reception signal with a signal-to-noise threshold, and returning to the preceding step when the signal-to-noise ratio fails to satisfy the signal-to-noise threshold.Type: GrantFiled: July 29, 2015Date of Patent: September 6, 2016Assignee: REALTEK SEMICONDUCTOR CORPORATIONInventors: Sheng-Fu Chuang, Liang-Wei Huang, Ching-Yao Su, Hsuan-Ting Ho
-
Publication number: 20160233920Abstract: A channel detection method for an echo canceller of a communication device is provided. The method includes the following steps. A first detection signal is transmitted to an end of a channel coupled to the communication device. A plurality of taps corresponding to a reflected signal of the first detection signal are received by an echo canceller at the end of the channel. The taps corresponding to the reflected signal are compared with a reference value corresponding to each of the taps so as to determine whether each of the taps is larger than or equal to the corresponding reference value. When the tap is determined to be larger than or equal to the reference value corresponding to the tap, the tap and a position of the tap are recorded.Type: ApplicationFiled: January 25, 2016Publication date: August 11, 2016Inventors: Shih-Wei Wang, Liang-Wei Huang, Ching-Yao Su, Sheng-Fu Chuang
-
Patent number: 9407234Abstract: The present invention discloses a current balancing device and method capable of balancing an output current and an input current of a current loop. Said device comprises: a transmission circuit for outputting an output current and receiving an input current; at least one adjustable resistor set in the current loop for providing resistance according to at least one adjustment signal; and a current balancing circuit, coupled to the transmission circuit and the adjustable resistor, for determining whether the difference between the output and input currents satisfies a predetermined requirement in light of a predetermined duration and thereby generating the adjustment signal, wherein if the difference between the output and input currents fails to satisfy the predetermined requirement, the current balancing circuit will adjust the resistance of the adjustable resistor through the adjustment signal, so as to reduce the difference between the output and input currents.Type: GrantFiled: October 15, 2014Date of Patent: August 2, 2016Assignee: REALTEK SEMICONDUCTOR CORPORATIONInventors: Shih-Wei Wang, Su-Liang Liao, Liang-Wei Huang, Sheng-Fu Chuang
-
Publication number: 20160132455Abstract: A control method applied to an Operating-Mode Finite-State-Machine (OPFSM) arranged for deciding a behavior of a first port of an apparatus includes: controlling the OPFSM to enter a second local state from a first local state and controlling the first port to send a signal with a wakeup pattern to a link partner of the first port when the state of the OPFSM is the first local state, and a wakeup request bit is a first local value.Type: ApplicationFiled: March 12, 2015Publication date: May 12, 2016Inventors: Hsuan-Ting Ho, Liang-Wei Huang, Ching-Yao Su, Sheng-Fu Chuang
-
Publication number: 20160065495Abstract: The present invention discloses a network connection establishing method capable of preventing a link procedure from being strangled in a training state. An embodiment of the method comprises: a preceding step; a training step; and a following step. Said training step includes at least one of the following: counting a number, and returning to the preceding step if a local receiver is not yet ready after finishing counting the number; detecting a reception signal according to a signal detection level, and returning to the preceding step when none of the reception signal is detected; determining a number of level(s) of the reception signal, and returning to the preceding step if the number of level(s) is less than an expected level number; and comparing a signal-to-noise ratio of the reception signal with a signal-to-noise threshold, and returning to the preceding step when the signal-to-noise ratio fails to satisfy the signal-to-noise threshold.Type: ApplicationFiled: July 29, 2015Publication date: March 3, 2016Inventors: SHENG-FU CHUANG, LIANG-WEI HUANG, CHING-YAO SU, HSUAN-TING HO
-
Publication number: 20160036489Abstract: This invention discloses a signal receiving device for Ethernet and a control method thereof. The signal receiving device includes a gain control circuit, an alien near-end crosstalk canceller, a noise canceller, and a DFE. The gain control circuit adjusts an input signal of the signal receiving device according to a setting parameter. The alien near-end crosstalk canceller cancels an alien near-end crosstalk interference. The noise canceller uses a first filter to cancel noises. The DFE uses a second filter to cancel an inter-symbol interference of the input signal. The method includes steps of: temporarily stopping the gain control circuit from updating the setting parameter before a seed collision occurs, and temporarily stopping one of the noise canceller and the decision feedback canceller from updating the first filter coefficient of the first filter or the second filter coefficient of the second filter temporarily during the seed collision.Type: ApplicationFiled: July 14, 2015Publication date: February 4, 2016Inventors: SHENG-FU CHUANG, LIANG-WEI HUANG, CHING-YAO SU, HSUAN-TING HO
-
Patent number: 9219457Abstract: A method for controlling amplification factor of a receiver device includes: detecting a first variation of an equivalent length of a cable, detecting a second variation of amplification factor of a digital automatic gain controller (DAGC), determining a first update value of amplification factor of an analog automatic gain controller (AAGC) according to the first variation and the second variation, calculating a tuning ratio of the first update value to a current value of the amplification factor of the AAGC, calculating a second update value of the amplification factor of the DAGC according to the tuning ratio, updating a set value of the amplification factor of the AAGC according to the first update value, and updating a set value of the amplification factor of the DAGC according to the second update value.Type: GrantFiled: December 16, 2014Date of Patent: December 22, 2015Assignee: REALTEK SEMICONDUCTOR CORP.Inventors: Ching-Yao Su, Liang-Wei Huang, Sheng-Fu Chuang, Hsuan-Ting Ho