2024年5月22日发(作者：年乐双)
The speed limitations of a digital circuit show up in four different forms:
propagation delay, setup time, rise time, and fall time. Propagation delay is
basically the time between a signal edge’s entering a device and leaving the
device.
数字电路速度的限制表现为以下四种不同的形式：传播延迟，建立时间，上升
时间和下降时间。传输延迟可以认为是一个信号边沿进入器件和离开器件的时
间。
When a number of digital devices are connected in series, their propagation
delays add up. When a similar set of digital devices are operating in parallel,
their propagation delays, because of the tolerances, is not necessary the same.
This problem is sometimes referred to as “skew”.
当大量的数字器件串联的时候，他们的传输延迟增加，当一组近似的数字器件
并行工作时，由于器件的公差，传播延迟可以不必相同，这种问题有时被称为
“时滞”
It is, of course, essential in digital circuits that signal edges occur in a known
order. It is a further absolute requirement that this order preserve a minimum
time between signal edges of concern. This called setup time. But simply, a signal
must remain at an input for a certain minimum amount of time or it will not be
recognized
当然，在数字电路中数字信号边沿按已知顺序出现是必要的。这是一个进一步
的必要条件，这个顺序在所关心的信号边沿中保持了一个最小的时间。这被称
为建立时间，但是简单来说，一个信号必须在输入端保持一段最小的时间否则
它将不会被识别
P195
In a semiconductor, on the other hand, because of the spatial overlap of their
wavefunctions, no two electrons in a crystal can be placed in the same quantum
state, i.e., possess the same eigenfunction
在半导体，另一方面，由于他们的波函数空间的重叠，在晶体没有两个电子可
以在同一个量子态，即具有相同的特征函数。
This is the so-called Pauli exclusion principle, which is one of the more important
axiomatic foundations of quantum mechanics. Each electron thus must possess a
unique spatial wavefunction and an associated eigen energy (the total energy
associated with the state).
这就是所谓的泡利不相容原理，它是量子力学更重要的基础定理之一。每个电
子因此必须具备一个独特的空间波函数和相关的本征能量（总能量与状态相关
联）.
If we plot a horizontal line, as in figure 5.1, for each allowed electron energy
(eigenenergy), we will discover that the energy levels cluster within bands that
are separated by ”energy gaps” ( “forbidden” gaps).
如果我们绘制一条水平线，如图5.1，对于每个电子能量（本征能级），我们会
发现，该能级集中在被能隙所分离的能带之中（禁带能隙）
A schematic description of the energy level spectrum of electrons in a crystals is
shown in Figure 5.1.
晶体电子光谱能级的图表描述如图5.1
P106
Diffraction losses increase as the mirror diameter a decreases and the wavelength
and resonator length increase. In order that a resonator be low in diffraction
losses at wavelength , it should be so long that
a
2
/4
L1
衍射损失增加是因为镜直径减小，波长和谐振器长度的增加。为了使一个谐振
器的波长衍射损耗降低，它应该这样长
a
2
/4
L1
The dimensionless parameter on left-hand side is called the Fresnel
number
N
F
a
2
/
4
L
. This number is not the only criterion used in
estimating diffraction losses. Other features of resonator geometry should also be
taken into account.
左侧的菲涅尔数叫做无量纲参数 。这个数不是估测衍
N
F
a
2
/4
L
射损失的唯一标准。几何谐振腔的其他特征也应该考虑。
These are determined by resonator length and mirror radii of curvature as laser
resonators normally use spherical rather than flat mirrors. So, two resonators of
the same Fresnel number may incur different diffraction losses due to different
geometry.
这些都取决于谐振腔的长度和镜曲率半径因为激光谐振腔通常使用球面镜而不
是平面镜。所以，两个有相同菲涅尔数的谐振腔可能会因为不同的几何形状而
导致不同的衍射损失。
Fig.3.2 shows a resonator produced by two spherical mirrors. Assume for
generality that the mirrors have different apertures(a1 and a2) and different
radii of curvature(r1 and r2). Such a resonator is defined by three principal
parameters
图3.2显示了一个由两个球面镜做成的谐振腔。假设镜普遍有不同的孔径（a1
和a2）和不同的曲率半径（r1和r2）。这样一个谐振腔由三个重要的参数来确
定。
P103
In order that some energy can be obtained as an output, one of the mirrors of a
laser is made slightly transmitting and is hence known as the output mirror. The
optimum transmission of the output mirror depends on the gain of the laser.
为了使一些能量可以作为输出来获得，一种激光镜被用作此为传输并因此被称
为输出镜。输出镜的最佳传输取决于激光增益。
If the transmission is too low, very little or no output will be achieved. If the
transmission is too high the number of passes up and down the cavity may be
insufficient to provide enough gain to overcome the losses. As a rough guide 1 or
2% transmission may be optimum for a low gain laser line, while 10% or more
may be needed if the gain is very high
如果传输太低，很少或是没有输出将会获得。如果传输过高通过上下腔的数量
可能不足以提供足够的增益以克服损失。作为一个粗略的指出1%或是2%的传
输可能是最佳的低增益激光线，然而如果增益很高的话可能会需要10%或是更
多。
P123
The end faces of the rod are polished and silvered to play the part of the optical
resonator mirrors. Therefore, in this laser the length of the resonant cavity is the
length of the rod
棒的端面被擦亮和镀银来起着光学谐振腔反射镜的作用。因此，激光谐振腔的
长度就是棒的长度。
P221
The most important feature of the buried heterostructure laser is that the active
GaAs region is surrounded on all sides by the lower index GaAlAs, so that
electro-magnetically the structure is that of a rectangular dielectric waveguide.
埋沟异质结构激光器最重要的特点是GaAs激活区域被地折射率GsAlAs在各个
方面所包围，以至于其从电磁特性来看这种结构相当于矩形介质波导。
The transverse dimensions of the active region and the index discontinuities (i.e.,
the molar fractions x, y, and z) are so chosen that only the lowest-order
transverse mode can propagate in the laser waveguide.
激活区域的横向尺寸和不连续指数（i.e,摩尔分数 x，y和z）的选取是以只有基
横模能在波导中出传播为依据的。
Another important feature of this laser is the confinement of the injected carriers
at the boundaries of the active region due to the energyband discontinuity at a
GaAs/GaAlAs interface as discussed in the last section. These act as potential
barriers inhibiting carrier escape out of the active region.
这种激光器的另一项重要特征是被注入的载流子被限制在激活区域的边界处，
正如上一章所讨论的这样，这种限制是由于GaAs和GaAlAs接缝处能带不连续
造成的。
P47
For a century or so the might of Newton’s authority held sway and the wave
theory did not obtain general acceptance, albeit with some notable exceptions;
particularly that of the mathematician Euler, who correctly associated waves of
different frequencies with different colors
在一个世纪左右牛顿的权威当道，波动理论并没有获得普遍的认可，即使有一
些著名的例外，特别是数学家欧拉，他正确的将不同频率的波和各自对应的颜
色联系在一起。
In 1801 Thomas Young in his classical two-slit experiment showed that light from
two sources could combine to form regions of brightness and darkness called
fringes.
1801年托马斯杨在他经典的双缝实验中表明从两个光源发出的光可以形成明暗
相间的干涉条纹。
These could only be explained in terms of a wave theory, a bright fringe being
formed where the two waves combine in phase so as to reinforce one another;
and dark fringes being formed where the two waves find themselves out of phase
and hence canceling each termed these phenomena constructive and
destructive interference respectively.
这些现象只能用波动理论解释为，在两列光波同相位结合处，彼此加强形成亮
纹，异相位结合处彼此相消形成暗纹。杨分别指出了这些相长和相消的干涉现
象。
The speed limitations of a digital circuit show up in four different forms:
propagation delay, setup time, rise time, and fall time. Propagation delay is
basically the time between a signal edge’s entering a device and leaving the
device.
数字电路速度的限制以四种不同的形式表现出来：传播延迟，建立时间，上升
时间和下降时间。
When a number of digital devices are connected in series, their propagation
delays add up. When a similar set of digital devices are operating in parallel,
their propagation delays, because of the tolerances, is not necessary the same.
This problem is sometimes referred to as “skew”.
当大量的数字器件串联的时候，他们的传输延迟增加，当一组近似的数字器件
并行工作时，由于器件的公差，传播延迟可以不必相同，这种问题有时被称为
“时滞”
It is, of course, essential in digital circuits that signal edges occur in a known
order. It is a further absolute requirement that this order preserve a minimum
time between signal edges of concern. This called setup time. But simply, a signal
must remain at an input for a certain minimum amount of time or it will not be
recognized
当然，在数字电路中数字信号边沿按已知顺序出现是必要的。这是一个进一步
的必要条件，这个顺序在所关心的信号边沿中保持了一个最小的时间。这被称
为建立时间，但是简单来说，一个信号必须在输入端保持一段最小的时间否则
它将不会被识别
2024年5月22日发(作者：年乐双)
The speed limitations of a digital circuit show up in four different forms:
propagation delay, setup time, rise time, and fall time. Propagation delay is
basically the time between a signal edge’s entering a device and leaving the
device.
数字电路速度的限制表现为以下四种不同的形式：传播延迟，建立时间，上升
时间和下降时间。传输延迟可以认为是一个信号边沿进入器件和离开器件的时
间。
When a number of digital devices are connected in series, their propagation
delays add up. When a similar set of digital devices are operating in parallel,
their propagation delays, because of the tolerances, is not necessary the same.
This problem is sometimes referred to as “skew”.
当大量的数字器件串联的时候，他们的传输延迟增加，当一组近似的数字器件
并行工作时，由于器件的公差，传播延迟可以不必相同，这种问题有时被称为
“时滞”
It is, of course, essential in digital circuits that signal edges occur in a known
order. It is a further absolute requirement that this order preserve a minimum
time between signal edges of concern. This called setup time. But simply, a signal
must remain at an input for a certain minimum amount of time or it will not be
recognized
当然，在数字电路中数字信号边沿按已知顺序出现是必要的。这是一个进一步
的必要条件，这个顺序在所关心的信号边沿中保持了一个最小的时间。这被称
为建立时间，但是简单来说，一个信号必须在输入端保持一段最小的时间否则
它将不会被识别
P195
In a semiconductor, on the other hand, because of the spatial overlap of their
wavefunctions, no two electrons in a crystal can be placed in the same quantum
state, i.e., possess the same eigenfunction
在半导体，另一方面，由于他们的波函数空间的重叠，在晶体没有两个电子可
以在同一个量子态，即具有相同的特征函数。
This is the so-called Pauli exclusion principle, which is one of the more important
axiomatic foundations of quantum mechanics. Each electron thus must possess a
unique spatial wavefunction and an associated eigen energy (the total energy
associated with the state).
这就是所谓的泡利不相容原理，它是量子力学更重要的基础定理之一。每个电
子因此必须具备一个独特的空间波函数和相关的本征能量（总能量与状态相关
联）.
If we plot a horizontal line, as in figure 5.1, for each allowed electron energy
(eigenenergy), we will discover that the energy levels cluster within bands that
are separated by ”energy gaps” ( “forbidden” gaps).
如果我们绘制一条水平线，如图5.1，对于每个电子能量（本征能级），我们会
发现，该能级集中在被能隙所分离的能带之中（禁带能隙）
A schematic description of the energy level spectrum of electrons in a crystals is
shown in Figure 5.1.
晶体电子光谱能级的图表描述如图5.1
P106
Diffraction losses increase as the mirror diameter a decreases and the wavelength
and resonator length increase. In order that a resonator be low in diffraction
losses at wavelength , it should be so long that
a
2
/4
L1
衍射损失增加是因为镜直径减小，波长和谐振器长度的增加。为了使一个谐振
器的波长衍射损耗降低，它应该这样长
a
2
/4
L1
The dimensionless parameter on left-hand side is called the Fresnel
number
N
F
a
2
/
4
L
. This number is not the only criterion used in
estimating diffraction losses. Other features of resonator geometry should also be
taken into account.
左侧的菲涅尔数叫做无量纲参数 。这个数不是估测衍
N
F
a
2
/4
L
射损失的唯一标准。几何谐振腔的其他特征也应该考虑。
These are determined by resonator length and mirror radii of curvature as laser
resonators normally use spherical rather than flat mirrors. So, two resonators of
the same Fresnel number may incur different diffraction losses due to different
geometry.
这些都取决于谐振腔的长度和镜曲率半径因为激光谐振腔通常使用球面镜而不
是平面镜。所以，两个有相同菲涅尔数的谐振腔可能会因为不同的几何形状而
导致不同的衍射损失。
Fig.3.2 shows a resonator produced by two spherical mirrors. Assume for
generality that the mirrors have different apertures(a1 and a2) and different
radii of curvature(r1 and r2). Such a resonator is defined by three principal
parameters
图3.2显示了一个由两个球面镜做成的谐振腔。假设镜普遍有不同的孔径（a1
和a2）和不同的曲率半径（r1和r2）。这样一个谐振腔由三个重要的参数来确
定。
P103
In order that some energy can be obtained as an output, one of the mirrors of a
laser is made slightly transmitting and is hence known as the output mirror. The
optimum transmission of the output mirror depends on the gain of the laser.
为了使一些能量可以作为输出来获得，一种激光镜被用作此为传输并因此被称
为输出镜。输出镜的最佳传输取决于激光增益。
If the transmission is too low, very little or no output will be achieved. If the
transmission is too high the number of passes up and down the cavity may be
insufficient to provide enough gain to overcome the losses. As a rough guide 1 or
2% transmission may be optimum for a low gain laser line, while 10% or more
may be needed if the gain is very high
如果传输太低，很少或是没有输出将会获得。如果传输过高通过上下腔的数量
可能不足以提供足够的增益以克服损失。作为一个粗略的指出1%或是2%的传
输可能是最佳的低增益激光线，然而如果增益很高的话可能会需要10%或是更
多。
P123
The end faces of the rod are polished and silvered to play the part of the optical
resonator mirrors. Therefore, in this laser the length of the resonant cavity is the
length of the rod
棒的端面被擦亮和镀银来起着光学谐振腔反射镜的作用。因此，激光谐振腔的
长度就是棒的长度。
P221
The most important feature of the buried heterostructure laser is that the active
GaAs region is surrounded on all sides by the lower index GaAlAs, so that
electro-magnetically the structure is that of a rectangular dielectric waveguide.
埋沟异质结构激光器最重要的特点是GaAs激活区域被地折射率GsAlAs在各个
方面所包围，以至于其从电磁特性来看这种结构相当于矩形介质波导。
The transverse dimensions of the active region and the index discontinuities (i.e.,
the molar fractions x, y, and z) are so chosen that only the lowest-order
transverse mode can propagate in the laser waveguide.
激活区域的横向尺寸和不连续指数（i.e,摩尔分数 x，y和z）的选取是以只有基
横模能在波导中出传播为依据的。
Another important feature of this laser is the confinement of the injected carriers
at the boundaries of the active region due to the energyband discontinuity at a
GaAs/GaAlAs interface as discussed in the last section. These act as potential
barriers inhibiting carrier escape out of the active region.
这种激光器的另一项重要特征是被注入的载流子被限制在激活区域的边界处，
正如上一章所讨论的这样，这种限制是由于GaAs和GaAlAs接缝处能带不连续
造成的。
P47
For a century or so the might of Newton’s authority held sway and the wave
theory did not obtain general acceptance, albeit with some notable exceptions;
particularly that of the mathematician Euler, who correctly associated waves of
different frequencies with different colors
在一个世纪左右牛顿的权威当道，波动理论并没有获得普遍的认可，即使有一
些著名的例外，特别是数学家欧拉，他正确的将不同频率的波和各自对应的颜
色联系在一起。
In 1801 Thomas Young in his classical two-slit experiment showed that light from
two sources could combine to form regions of brightness and darkness called
fringes.
1801年托马斯杨在他经典的双缝实验中表明从两个光源发出的光可以形成明暗
相间的干涉条纹。
These could only be explained in terms of a wave theory, a bright fringe being
formed where the two waves combine in phase so as to reinforce one another;
and dark fringes being formed where the two waves find themselves out of phase
and hence canceling each termed these phenomena constructive and
destructive interference respectively.
这些现象只能用波动理论解释为，在两列光波同相位结合处，彼此加强形成亮
纹，异相位结合处彼此相消形成暗纹。杨分别指出了这些相长和相消的干涉现
象。
The speed limitations of a digital circuit show up in four different forms:
propagation delay, setup time, rise time, and fall time. Propagation delay is
basically the time between a signal edge’s entering a device and leaving the
device.
数字电路速度的限制以四种不同的形式表现出来：传播延迟，建立时间，上升
时间和下降时间。
When a number of digital devices are connected in series, their propagation
delays add up. When a similar set of digital devices are operating in parallel,
their propagation delays, because of the tolerances, is not necessary the same.
This problem is sometimes referred to as “skew”.
当大量的数字器件串联的时候，他们的传输延迟增加，当一组近似的数字器件
并行工作时，由于器件的公差，传播延迟可以不必相同，这种问题有时被称为
“时滞”
It is, of course, essential in digital circuits that signal edges occur in a known
order. It is a further absolute requirement that this order preserve a minimum
time between signal edges of concern. This called setup time. But simply, a signal
must remain at an input for a certain minimum amount of time or it will not be
recognized
当然，在数字电路中数字信号边沿按已知顺序出现是必要的。这是一个进一步
的必要条件，这个顺序在所关心的信号边沿中保持了一个最小的时间。这被称
为建立时间，但是简单来说，一个信号必须在输入端保持一段最小的时间否则
它将不会被识别