摘要:
Equalizer circuitry on an integrated circuit (“IC”) includes a plurality of NMOS equalizer stages connected in series. Each NMOS stage may include folded active inductor circuitry. Each NMOS stage may also include various circuit elements having controllably variable circuit parameters so that the equalizer can be controllably adapted to perform for any of a wide range of high-speed serial data signal bit rates and other variations of communication protocols and/or communication conditions. For example, each NMOS stage may be programmable to control at least one of bandwidth and power consumption of the equalizer circuitry. The equalizer may also have a first PMOS stage that can be used instead of the first NMOS stage in cases in which the voltage of the incoming signal to be equalized is too low for an initial NMOS stage.
摘要:
An equalizer circuitry that includes both inductor based and non-inductor based equalizer stages is provided. In one implementation, the equalizer circuitry includes a first equalizer circuitry including a first inductor based equalizer stage and a first non-inductor based equalizer stage coupled to the first inductor based equalizer stage. In one implementation, the equalizer circuitry further includes a second equalizer circuitry including a plurality of inductor based equalizer stages, where the plurality of inductor based equalizer stages includes the first inductor based equalizer stage. In one implementation, the first equalizer circuitry further includes a second inductor based equalizer stage coupled to the first inductor based equalizer stage and the first non-inductor based equalize stage.
摘要:
Devices and methods for serial loopback testing in an integrated circuit (IC) are provided. To implement loopback testing, an equalizer stage of a receiver of the IC is powered down. In addition, the common-mode voltage of the equalizer stage is reduced and/or a bulk node of the equalizer stage is connected to ground. Doing so may reduce the impact of capacitive coupling from the input pins of buffer, thereby improving the quality of the loopback output signal.
摘要:
The present disclosure provides apparatus and methods for the calibration of analog circuitry on an integrated circuit. One embodiment relates to a method of calibrating analog circuitry within an integrated circuit. A microcontroller that is embedded in the integrated circuit is booted up. A reset control signal is sent to reset an analog circuit in the integrated circuit, and a response signal for the analog circuit is monitored by the microcontroller. Based on the response signal, a calibration parameter for the analog circuit is determined, and the analog circuit is configured using the calibration parameter. Other embodiments, aspects and features are also disclosed.
摘要:
A transceiver system with reduced latency uncertainty is described. In one implementation, the transceiver system has a word aligner latency uncertainty of zero. In another implementation, the transceiver system has a receiver-to-transmitter transfer latency uncertainty of zero. In yet another implementation, the transceiver system has a word aligner latency uncertainty of zero and a receiver-to-transmitter transfer latency uncertainty of zero. In one specific implementation, the receiver-to-transmitter transfer latency uncertainty is eliminated by using the transmitter parallel clock as a feedback signal in the transmitter phase locked loop (PLL). In one implementation, this is achieved by optionally making the transmitter divider, which generates the transmitter parallel clock, part of the feedback path of the transmitter PLL. In one implementation, the word aligner latency uncertainty is eliminated by using a bit slipper to slip bits in such a way so that the total delay due to the word alignment and bit slipping is constant for all phases of the recovered clock. This allows for having a fixed and known latency between the receipt and transmission of bits for all phases of parallelization by the deserializer. In one specific implementation, the total delay due to the bit shifting by the word aligner and the bit slipping by the bit slipper is zero since the bit slipper slips bits so as to compensate for the bit shifting that was performed by the word aligner.
摘要:
In a programmable logic device with a number of different types of serial interfaces, different power supply filtering schemes are applied to different interfaces. For interfaces operating at the lowest data rates—e.g., 1 Gbps—circuit-board level filtering including one or more decoupling capacitors may be provided. For interfaces operating at somewhat higher data rates—e.g., 3 Gbps—modest on-package filtering also may be provided, which may include power-island decoupling. For interfaces operating at still higher data rates—e.g., 6 Gbps—more substantial on-package filtering, including one or more on-package decoupling capacitors, also may be provided. For interfaces operating at the highest data rates—e.g., 10 Gbps—on-die filtering, which may include one or more on-die filtering or regulating networks, may be provided. The on-die regulators may be programmably bypassable allowing a user to trade off performance for power savings.
摘要:
Techniques are provided for transmitting signals through a differential interface between circuits in different power supply domains. A driver circuit in a first power supply domain converts single-ended signals into differential signals. The driver circuit then transmits the differential signals to a receiver circuit in a second power supply domain. The receiver circuit converts the differential signals back into single-ended signals for transmission to circuit elements in the second power supply domain. The differential interface reduces the transmission of noise between circuit elements in the first power supply domain and circuit elements in the second power supply domain.
摘要:
Various methods and structures related to clock distribution for flexible channel bonding are disclosed. One embodiment provides a clock network in physical media attachment (“PMA”) circuitry, a specific type or portion of system interconnect circuitry, arranged in pairs of channel groups. In one embodiment, clock generation circuitry blocks (“CGBs”) in each pair of channel groups receives outputs of multiple phased locked loop circuits (“PLLs”) which can be selectively utilized by the CGBs to generate PMA clock signals. In another embodiment, the CGBs can also select output of a clock data recovery (“CDR”)/transmit PLL circuitry block in one of the channels of a channel group of the pair of channel groups. In one embodiment, first groups of connection lines couple circuitry in a channel group pair such that a designated CGB in each channel group pair can provide clock signals to one or more of the channels in the channel group pair. In one embodiment, second groups of connection lines connect channels in one channel group pair to channels in other channel group pairs such that one or more channels across the channel group pairs can receive a clock signal generated by a CGB in a designated channel. These and other embodiments are described more fully in the disclosure.
摘要:
Methods and circuits for automatic adjustment of equalization are presented that improve the quality of equalization for input signals with varying amplitudes. The methods and circuits may be used in Decision Feedback Equalization (DFE) circuits to maintain a constant equalization boost amplitude despite variations in input signal amplitude. The equalization circuitry measures the amplitude of the equalization input signal and computes tap coefficients to maintain a desired level of boost amplitude. Tap coefficients may be automatically adjusted by the equalization circuitry.
摘要:
Equalization of an incoming data signal can be controlled by sampling that signal at times when data values in that signal should be stable (“data samples”) and when that signal should be in transition between successive data values that are different (“transition samples”). A transition sample that has been taken between two successive differently valued data samples is compared to a reference value (which can be one of those two data samples). The result of this comparison can be used as part of a determination as to whether to increase or decrease equalization of the incoming data signal.