外文文献原文基于fpga的逻辑分析仪的设计与实现中英文翻译大学论文.doc

本科毕业设计（论文）外文翻译 题 目 学生姓名 班 级 学 号 院 （系） 专 业 指导教师 职 称 2017年 月 日原文：Clock Buffer Basics author :Hamilton, Mark1( markhrennes.ucc.ie);Marnane, William P.1( liameleceng.ucc.ie) press :clock buffer with FPGAClocks are the basic building blocks for all electronics today. For every data transition in a synchronous digital system, there is a clock that controls a register. Most systems use Crystals, Frequency Timing Generators (FTGs), or inexpensive ceramic resonators to generate precision clocks for their synchronous systems. Additionally, clock buffers are used to create multiple copies, multiply and divide clock frequencies, and even move clock edges forwards or backward in time. Many clock-buffering solutions have been created over the past few years to address the many challenges required by todays high-speed logic systems. Some of these challenges include: High operating and output frequencies, propagation delays from input to output, output to output skew between pins, cycle-tocycle and long-term jitter, spread spectrum, output drive strength, I/O voltage standards, and redundancy. Because clocks are the fastest signals in a system and are usually under the heaviest loads, special consideration must be given when creating clocking trees. In this chapter, we outline the basic functions of non-PLL and PLL-based buffers and show how these devices can be used to address the high-speed logic design challenges.In todays typical synchronous designs, multiple clock signals are often needed to drive a variety of components. To create the required number of copies, a clock tree is constructed. The tree begins with a clock source such as an oscillator or an external signal and drives one or more buffers. The number of buffers is typically dependent on the number and placement of the target devices. In years past, generic logic components were used as clock buffers. These were adequate at the time, but they did little to maintain the signal integrity of the clock. In fact, they actually were a detriment to the circuit. As clock trees increased in speed and timing margins reduced, propagation delay and output skew became increasingly important. In the next several sections, we discuss the older devices and why they are inadequate to meet the needs of todays designs. The definitions of the common terms associated with modern buffers follow. Finally, we address the attributes of the modern clock buffer with and without a PLL. The FTG that is often used as a clock source is a special type of PLL clock buffer.Early BuffersA clock buffer is a device in which the output waveform follows the input waveform. The input signal propagates through the device and is re-driven by the output buffers. Hence, such devices have a propagation delay associated with them. In addition, due to differences between the propagation delay through the device on each input-output path, skew will exist between the outputs. An example of a non-PLL based clock buffer is the 74F244 that is available from several manufacturers. These devices have been available for many years and were suitable for designs where frequencies were below 20 MHz. Designers would bring in a clock and fan it out to multiple synchronous devices on a circuit card. With these slow frequencies and associated rise times, designers had suitable margins with which to meet setup and hold times for their synchronous interfaces. However, these buffers are not optimal for todays high-speed clocking requirements. The 74F244 suffers from a long propagation delay (3 to 5 ns) and long output-to-output skew delays. Non-PLL based clock buffers have improved in recent years and use more advanced I/O design techniques to improve the output-to-output skew. As the clock period gets shorter, the uncertainty or skew in the clock distribution system becomes more of a factor. Since clocks are used to drive the processors and to synchronize the transfer of data between system components, the clock distribution system is an essential part of the system design. A clock distribution system design that does not take skew into consideration may result in a system with degraded performance and reliability.Clock SkewSkew is the variation in the arrival time of two signals specified to occur at the same time. Skew is composed of the output skew of the driving device and variation in the board delays caused by the layout variation of the board traces. Since the clock signal drives many components of the system, and since all of these components should receive their clock signal at precisely the same time in order to be synchronized, any variation in the arrival of the clock signal at its destination will directly impact system performance. Skew directly affects system margins by altering the arrival of a clock edge. Because elements in a synchronized system require clock signals to arrive at the same time, clock skew reduces the cycle time within which information can be passed from one device to the next.As system speeds increase, clock skew becomes an increasingly large portion of the total cycle time. When cycle times were 50 ns, clock skew was rarely a design priority. Even if skew was 20% of the cycle time, it presented no problem. As cycle times dropped to 15ns and less, clock skew requires an ever-increasing amount of design resource. Now typically, these high-speed systems can have only 10% of their timing budget dedicated to clock skew, so obviously, it must be reduced.There are two types of clock skew that affect system performance. The clock driver causes intrinsic skew and the printed circuit board (PCB) layout and design is referred to as extrinsic skew. Extrinsic skew and layout procedures for clock trees will be discussed later in this book. The variation of time due to skew is defined by the following equation:tSKEW_INTRINSIC = Device Induced SkewtSKEW_EXTRINSIC = PCB + Layout + Operating Environment Induced SkewtSKEW = tSKEW_INTRINSIC + tSKEW_EXTRINSICIntrinsic clock skew is the amount of skew caused by the clock driver or buffer by itself. Board layout or any other design issues, except for the specification stated on the clock driver data sheet do not cause intrinsic skew.Output SkewOutput skew (tSK)is also referred to as pin-to-pin skew. Output skew is the difference between delays of any two outputs on the same device at identical transitions. Joint Electronic Device Engineering Council (JEDEC) defines output skew as the skew between specified outputs of a single device with all driving inputs connected together and the outputs switching in the same direction while driving identical specified loads. Figures 2.2 and 2.3 show a clock buffer with common input Cin driving outputs Co1_1 through Co1_n. The absolute maximum difference between the rising edges of the outputs will be specified as output skew. Typical output skew in todays high performance clock buffers is around 200 picoseconds (ps).Part-to-Part SkewPart-to-part skew (tDSK)is also known as package skew and device-to-device skew. Part-topart skew is similar to output skew except that it applies to two or more identical devices.Part-to-part skew is defined as the magnitude of the difference in propagation delays between any specified outputs of two separate devices operating at identical conditions. The devices must have the same input signal, supply voltage, ambient temperature, package, load, environment, etc. Figure 2.4 illustrates tDSK from the preceding example.Typical part-to-part skew for todays high performance buffers is around 500 ps. Propagation DelayPropagation delay (tPD) is the time between specified reference points on the input and output voltage waveforms with the output changing from one defined level (low) to the other (low). Propagation delay is illustrated in Figure 2.3. Non-PLL based devices in todays high performance devices range from 3 to 7 ns. PLL-based buffers are able to zero out this propagation delay with the aid of Phase Detectors, Loop Filters and Voltage Controlled Oscillators (VCOs).Uneven LoadingWhen using a high-speed clock buffer or PLL, care must be taken to equally load the outputs of the device to ensure that tight skew tolerances are maintained. Inherent in each output of the clock driver is an output impedance that is mostly resistive in nature (along with some inductance and capacitance). When each of these resistive outputs is equally loaded, the tight skew specification of the clock driver is preserved. If the loads become unbalanced, the (RC) time constants of the various outputs would be different, and the skew would be directly proportional to the variation in the loading.Input Threshold VariationAfter the low skew clock signals have been distributed, the clock receivers must accept the clock input with minimal variations. If the input threshold levels of the receivers are not uniform, the clock receivers will respond to the clock signals at different times creating clock skew. If one load device has a threshold of 1.2 volts and another load device has a threshold of 1.7 volts and the rising edge rate is 1V/ns, there will be 500 ps of skew caused by the point at which the load device switches based on the input signal. Most manufacturers center the input threshold level of their devices near 1.5 volts nominal for (TTL) input devices. This input threshold will vary slightly from manufacturer to manufacturer especially as conditions (such as voltage and temperature) change. The TTL specification for the input threshold level is guaranteed to be a logic high when the input voltage is above 2.0 volts and a logic low when the input voltage level is below 0.8 volts.This leaves a 1.2-volt window over voltage and temperature. Components with Complementary Metal Oxide Semiconductor (CMOS) rail swing inputs have a typical input threshold of VCC/2 or about 2.5 volts, which is much higher than the TTL level. If the threshold levels are not uniform, clock skew will develop between components because of these variations. There are many I/O standards which have emerged and all must be taken into consideration when providing clocks to different subsystems. Table 2.1 listed below which lists the more prevalent standards along with the input threshold voltages.Non-PLL Based Clock DriversThere are two main types of modern clock driver architectures: a buffer-type device (non-PLL) and a feedback-type device (PLL).In a buffer-style (non-PLL) clock driver, the input wave propagates through the device and is “re-driven” by the output buffers. This output signal directly follows the input signal and has a propagation delay (tPD) that ranges from 5 ns to over 15 ns. These devices differ from the buffers in the past such as the 74F244 in that they are designed specifically for clock signals. On a 74F244, there are eight inputs and eight outputs. To create a one to eight buffer, all eight inputs are tied together. This causes excess loading at the inputs on the driving signal. A one to eight clock buffer has only one input and hence only one load. The output rise and fall times are also equally matched and therefore do not contribute to duty cycle error. With their improved I/O structure, the pin-to-pin skew is kept to a minimum.The output skew of this device, if it is not listed on the data sheet, can be calculated by subtracting the minimum propagation delay from the maximum propagation delay.The 10 ns tPD clock driver delay shown in Figure 2.5 does not take into account the affects of the board layout and design. These types of devices are excellent for buffering source signals such as oscillators where the output phase does not need to match the input. A variety of the non-PLL based buffers are available on the market today and typically range from as few as 4 outputs to as many as 30. Some devices also include configurable I/O and internal registers to divide the output frequencies.Among the highest performance non-PLL based Low Voltage CMOS (LVCMOS) clock buffers available today is the B9940L. The B9940L is a low-voltage clock distribution buffer with the capability to select either a differential LVPECL or a LVCMOS/LVTTL compatible input clock. The two clock sources can be used to provide for a test clock as well as the primary system clock. All other control inputs are LVCMOS/LVTTL-compatible. The eighteen outputs are 2.5V- or 3.3V-compatible and can drive two series terminated 50-Ohm transmission lines. With this capability, the B9940L has an effective fan out of 1:36. Low output-to-output skews of 150 ps, a device to device skew of 750 ps, and a high-end operating frequency of 200 MHz, makes the B9940L an ideal clock distribution buffer for nested clock trees in synchronous systems.These devices still face the problems of device propagation delay. The propagation delay through these devices is about 5 ns. This delay will cause skew in systems where both the reference clock to the buffer and the outputs of the buffer need to be aligned. These devices also have the drawback that the output waveform is directly based on the input waveform. If the input waveform is a non-50% duty-cycle clock, the output waveform will also have a less-than-ideal duty cycle. Expensive crystal oscillators with tight tolerances are needed when using this type of buffer in systems requiring near 50/50 outputs.These devices also lack the ability to phase adjust or frequency multiply their outputs. Phase adjustment allows the clock driver to compensate for trace propagation delay mismatches and setup and hold time differences, and frequency multiplication allows the distribution of high and low frequency clocks from the same common reference. Expensive components and time-consuming board routing techniques must be used to compensate for the functional shortcomings of these buffer-style clock driver devices. PLL-based devices have been incorporated to address all of these shortcomings.PLL-Based Clock DriversThe second type of clock distribution device uses a feedback input that is a function of one of its outputs. The feedback input can be connected internally or externally to the part. If its an external feedback, a trace is used to connect an output pin to the feedback pin. This type of device is usually based upon one or more PLLs that are used to align the phase and frequency of the feedback input and the reference input. Since the feedback input is a reflecion of an output pin, the propagation delay is effectively eliminated. In addition to very low device propagation delay, this type of architecture enables output signals to be phas