The invention may be best understood by reading the disclosure with reference to the following drawings:
The ability of systems to generate multiple copies of the output signals from a common reference signal with low output-to-output skew and low jitter is advantageous. As described above, prior output buffering devices generate output signals at their spatially-separated output locations, thus requiring strategic design and additional reference signal route buffering to meet skew and jitter requirements. However, by generating multiple output signals at a centralized location and subsequently routing them to their respective output pads, designers gain the freedom to incorporate various output pad configurations into their systems while maintaining low output-to-output skew and jitter. Embodiments of the present invention will now be described in more detail.
By centralizing the output signal generation, the buffer unit 300 generates the output signals 215-1 to 215-N substantially in-phase, thus reducing output-to-output skew. To minimize the contributions to the output-to-output skew caused by routing the output signals 215-1 to 215-N between the buffer unit 300 and the output pads 220-1 to 220-N, designers may configure output signal routing paths to be substantially equal in length or to have substantially the same capacitance, or preferably both. For instance, tapping one or more of the routing paths, e.g., with capacitance-adjusting stubs, may substantially equalize the capacitance on the output signal routing paths. Thus, by generating the output signals 215-1 to 215-N at a centralized location and routing the output signals 215-1 to 215-N to the output pads 220-1 to 220-N, various output pad 220-1 to 220-N configurations may be realized that maintain low output-to-output skew.
In some embodiments, the buffer unit 300 generates current-driven output signals 215-1 to 215-N, i.e., their signaling is indicated by current-level changes, thus allowing the output signals 215-1 to 215-N to route from the buffer unit 300 to their respective output pads 220-1 to 220-N without requiring additional route buffering. This absence of route buffers along the output signal routing paths increases the throughput of the output buffering system 200, while decreasing jitter and system size. Furthermore, the routing of current-driven output signals 215-1 to 215-N involves lower routing impedance than its voltage-driven counterpart, which corresponds to a reduction of cross-talk with a core 230 toggling signals of the output buffering system 200.
Each of the output buffers 400-1 to 400-N receives the reference signal 205 at approximately the same phase, thus reducing output-to-output skew caused by output buffers 400-1 to 400-N generating output signals 215-1 to 215-N with minimized phase difference with respect to the reference signal 205. Furthermore, by grouping the output buffers 400-1 to 400-N within the buffer unit 300, each of the output buffers 400-1 to 400-N also receives a power supply of substantially the same voltage-level, as the IR drops associated with routing the power supply to the buffer unit 300 are approximately equal. Thus, potential variations in the voltage-level of the power supply due to IR drops are minimized, allowing the output buffers 400-1 to 400-N to generate the output signals 215-1 to 215-N substantially in-phase.
The driver 440 generates drive voltages 442A and 442B responsive to the reference signal 205 and provides them to the gate regions of the transistors 430A and 430B, respectively. The drive voltages 442A and 442B activate transistors 430A and 430B, respectively, when set to a high-level by the driver 440. Once activated, the transistors 430A and 430B steer the reference current 415 generated by the current mirror 410 to an output forming the output signal 215-1. Thus, in the differential circuit embodiments, the output buffer 400-1 generates a complimentary pair of current-driven Output signals 215-1
One of skill in the art will recognize that the concepts taught herein can be tailored to a particular application in many other advantageous ways. In particular, those skilled in the art will recognize that the illustrated embodiments are but one of many alternative implementations that will become apparent upon reading this disclosure.
The preceding embodiments are exemplary. Although the specification may refer to “an”, “one”, “another”, or “some” embodiment(s) in several locations, this does not necessarily mean that each such reference is to the same embodiment(s), or that the feature only applies to a single embodiment.
This application claims priority from U.S. Provisional Application No. 60/787,910 filed Mar. 31, 2006, which is incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
60787910 | Mar 2006 | US |