光子探测优化：深入理解与应用

需积分: 9 54 浏览量更新于2024-07-18 收藏 1.81MB PDF 举报

"光子探测与测量：光学系统性能最优化"这一资源深入探讨了光子探测的基本原理和方法，特别关注的是基于pn结的光电二极管，这是本书的核心内容。章节1首先介绍了光电二极管的工作原理，强调了在实际电路中有效利用这种设备的重要性。通过一个简化模型，读者可以理解光电二极管的主要特性，如输出极性判断、预期输出电流估计以及检测速度与负载的关系。模型虽然不是现代加工技术下的典型组件，但通过这个基础，可以帮助读者建立起对硅基pn结二极管的实际应用概念。章节详细解释了硅原子的四价结构，然后转向如何通过掺杂五价元素（如砷或磷）来创建一个p-n结，这在光电效应中起关键作用。在这个过程中，学习者会掌握如何利用p型和n型半导体的电子特性，使得光照射时产生电流，从而实现光信号的转换。理解光电二极管的内部工作机制，包括量子效率、暗电流、响应时间等因素，对于优化光学系统的性能至关重要。随着章节的深入，还将涉及光敏二极管的噪声分析、线性动态范围、温度依赖性等关键参数的测量和控制。此外，还会介绍几种常见的光探测器类型，如雪崩光电二极管（APD）和硅基光电倍增器，以及它们在不同应用场景中的优势和局限。通过阅读这份难得的英文资料，读者将不仅掌握基本的光子探测技术，还能了解到如何优化光学系统的性能，以满足各种高精度测量和成像应用的需求。这对于从事光学工程、物理测量、遥感技术或相关科研领域的专业人士来说，是一份极具价值的学习参考资料。

In the fourth quadrant we can extract power from the detector. This is the

region where solar cells are designed to work. For best efﬁciency we should

deﬁne the operating point through choice of the load resistance to maximize

the VI product. At this point it is possible to extract roughly 80 percent of the

product. Solar cells are pn-junction devices, usually designed with higher

doping levels than pin-photodetectors to keep series resistance low and to max-

imize the absorption of light at the short wavelength end of the silicon

absorbance spectrum. This is to best match the spectrum of sunlight. Although

solar cells are rarely considered as detectors for instrumentation, their low

cost/area can make them very attractive as such.

1.7.5 Parasitic capacitance

Let’s look now at the other elements of our equivalent circuit of Fig. 1.7. The

parasitic capacitance which appears as if across the photodiode is C

. Its origin

lies in the positive and negative space charges separated by the depletion region

of Fig. 1.2, which act like a parallel plate capacitor. Although the diode area

is rather small compared with a component capacitor, the thinness of the

depletion region can lead to high capacitance values. It increases approximately

linearly with the detector area and decreases with increasing reverse bias.

Measurement of the junction capacitance as a function of reverse voltage can

tell us the device’s doping proﬁle. A plot of 1/C

versus voltage can provide

Photodetection Basics 15

Current

Breakdown

Voltage

1st Quadrant

4th Quadrant3rd Quadrant

2nd Quadrant

Dark

Forward bias

Reverse bias

Solar cells

Max

power

Saturation

Current I

Increasing

Light

Level

Figure 1.9

Four-quadrant current/voltage characteristic. With-

out illumination this is similar to a conventional diode. Increas-

ing illumination shifts the characteristic in the negative current

direction. Detection is possible in quadrants 1, 3, and 4.

Photodetection Basics

the junction bandgap energy (Sze 1969). Different processing parameters

during manufacture can have a large effect on C

. In Chaps. 2 and 3 we will see

how important is C

to the speed and noise performance of an ampliﬁed

photodetector.

1.7.6 Leakage resistance

All photodiodes act as though they have a resistance shunting the diode (R

Fig. 1.7). Its value depends on the processes used to fabricate the diode sub-

strate and junction and decreases with increasing area. A high quality 1-mm

silicon device (e.g., PIN040 from UDT Sensors Inc.) speciﬁes a leakage resist-

ance of typically 1GW and not less than 200MW at room temperature. Tem-

perature increase reduces R

. Other manufacturers publish instead a leakage

current or dark current, which is just the current that will ﬂow through the

shunt resistance for a speciﬁed terminal voltage. If the diode voltage in the

actual circuit is zero, then no current will ﬂow and the effects of R

will be

minimized. We will see later that R

can have a signiﬁcant effect on the detec-

tion signal-to-noise ratio. For low values of R

even a few millivolts offset from

an operational ampliﬁer can lead to signiﬁcant leakage current.

Leakage can be greatly modiﬁed by process parameters, although signiﬁcant

reduction is usually at the expense of other diode characteristics. For example,

the 7.6-mm

detection area BPW34 photodiode manufactured by Siemens is

speciﬁed with a leakage of 2nA at 10-V reverse bias and capacitance of 72 pF at

zero bias. The very similar BPW33 exhibits only 20-pA dark current, but the

parasitic capacitance increases to 750pF under the same conditions.

1.7.7 Series resistance

Series resistance comes from the bulk resistance of the photodiode substrate,

from the ohmic contact diffusions, and from the leads. It is probably the least

troubling diode characteristic for normal laboratory detection needs but

becomes important in high bandwidth systems and in power generators using

solar cells.

1.7.8 Table of representative photodiodes

Table 1.2 shows a few representative photodiodes. The BPW34 is a common

silicon device in a clear plastic package and has been available for 30 years. The

BPW33 looks similar, with the same active area. However, as noted, the dark

leakage current has been reduced signiﬁcantly, at the expense of a tenfold

increase in junction capacitance. Items 3 and 4 are modern design silicon devices

and show the change in dark current and capacitance incurred by changing the

active area. These also offer very low temperature coefﬁcients of responsivity

at less than 100 ppm/°C. Items 5 and 6 are InGaAs detectors. The ﬁrst has an

area of 1 mm

and a capacitance of 90pF. Item 6 is similar, except that a spheri-

cal lens has been used to accept about 1mm

of beam area but focus it onto a

16 Chapter One

Photodetection Basics

much smaller diode. In the process the capacitance has dropped to 1 pF. Item 9

is another device manufactured for a very small surface area and for use with

single-mode ﬁbers. Subpicofarad capacitance has been achieved in this way.

Item 10 is a small active area and low capacitance chip device offered for 10-

Gbps telecomms applications. The last column gives the noise equivalent power

(NEP) at the wavelength of peak responsivity. This is the lowest power that

could be detected under optimum conditions in a 1-Hz bandwidth, limited by

the dark current noise.

1.8 APDs, Photomultipliers, Photon-Counting

Avalanche photodiodes (APDs) are similar to conventional pin photodiodes but

are designed to produce and withstand a high internal electric ﬁeld, which can

accelerate photoelectrons generated near the depletion region. As they are accel-

erated, additional photoelectrons are generated and a cascade process produces

10 to 1000 photoelectrons per incident photon. This high internal gain can be

of great use in receiver design. The gain is produced through application of

a higher than normal reverse terminal voltage. For example, the low-bias

Hamamatsu S2383 is a 1-mm-diameter silicon device that operates at 150 to

200V for a gain of 100. Some very large APDs also are now becoming available.

Advanced Photonix Inc. offers, for example, the 16-mm diameter 630-70-72-

Photodetection Basics 17

TABLE

1.2 Selection of Representative Photodiodes

Active Peak Respons. Capacitance NEP

Type Dimensions (A/W) Dark Current (pF) (W/Hz

)

1 Inﬁneon BPW34 2.75 ¥ 2.75 mm 0.6 @ 0.85mm2 (<30 nA) @ 72pF @ 0 V 4.2 ¥ 10

-14

Silicon 10V

2 Inﬁneon BPW33 2.75 ¥ 2.75 mm 0.6 @ 0.85mm 20 pA @ 1V 750 pF @ 0V 5.3 ¥ 10

-15

Silicon

3 Hamamatsu 1.1 ¥ 1.1mm 0.5 @ 0.95mm 20 pA @ 10mV 20 pF @ 0 V 5.7 ¥ 10

-15

S1336-18 Silicon

4 Hamamatsu 10 ¥ 10mm 0.5 @ 0.95 mm 200pA @ 10 mV 1100 pF @ 0 V 1.8 ¥ 10

-14

S1337-1010 Silicon

5 Hamamatsu 1-mm diam. 0.95 @ 1.55mm 5 nA @ 5V 90pF @ 5 V 2 ¥ 10

-14

G8370-01 InGaAs

6 Hamamatsu (Spherical lens) 0.95 @ 1.55 mm 80pA @ 5 V 1 pF @ 5V 2 ¥ 10

-15

G6854-01 InGaAs

7 AME-UDT 1 ¥ 1 mm 0.55 @ 0.90 mm 500pA @ 20 V 3 pF @ 20 V 2.3 ¥ 10

-14

BPX65 Silicon

8 AME-UDT 100-mm diam. 0.95 @ 1.55 mm 50pA @ 10 V 1.5pF —

InGaAs-100L InGaAs

9 Hamamatsu 40-mm diam. 0.95 @ 1.55mm 60 pA @ 5V 0.6 pF @ 2V 2 ¥ 10

-14

G8198-01 InGaAs

10 OSI Fibercomm 25-mm diam. 0.95 @ 1.55mm 100 pA @ 5V 0.2 pF @ 5V —

FCI-InGaAs-25C InGaAs

Photodetection Basics

5X1, providing a responsivity of 93A/W at 750nm with a bias voltage in the

range of 1700 to 2000V. APDs in InGaAs are also available for long wavelength

(1.55mm) use.

Note that these detectors are not inherently more sensitive nor do they offer

lower noise than conventional pin-diodes. They all offer about the same wave-

length response, capacitance, and quantum efﬁciency and so have essentially

the same limitations. APDs are also inherently noisier because of the presence

of excess noise caused by ﬂuctuations in the gain mechanism. However, where

the provision of sufﬁcient external electronic gain and bandwidth is problem-

atic, leading to limitations in receiver performance, the internal gain of the APD

can be very useful. This is generally the case only at higher speeds (say 1MHz

and above). They are rarely superior for low light level detection at low fre-

quencies. They are also much more difﬁcult to use than pin diodes, owing to

the precise regulation of the high bias voltage that is needed, which is itself

temperature-dependent. However, complete detection modules are available

from Hamamatsu and others to ease the design difﬁculties.

Similar tradeoffs occur with photomultipliers (PM tubes), which generate

photoelectrons in an inorganic semiconductor phosphor. Like the APDs, these

use an internal cascade multiplication process to achieve very high external

responsivities, typically 10

to 10

A/W. However, their intrinsic sensitivity is

rather low, as their photocathodes exhibit quantum efﬁciencies much less than

peak semiconductor detector values, for example, 0.1A/W. They also require a

stable, low noise high voltage supply at the kilovolt level. Advantages of PM

tubes are their wide range of sensitivity obtained through a choice of photo-

cathode materials, large detection areas compared with semiconductor photo-

diodes, and low dark currents. They are also relatively insensitive to ionizing

radiation. Both APDs and PM tubes are available with low enough dark cur-

rents to be used in photon-counting mode. Photon counting is a sensitive detec-

tion method for use at low incident powers (typically <10

photoelectrons/

second), where instead of averaging photoelectrons into a continuous current,

individual electron charges are counted. This brings the advantage of a kind of

separation between thermal noise and photoevents through the use of carefully

set up threshold discriminators. Large gains in performance are sometimes pos-

sible using this mode of detection.

1.9 Summary

In this chapter we have discussed the basics of photodetection using junction

photodiodes and shown how they can be represented by a simple current-

generator model. This is adequate in most cases to predict performance in a

wide variety of externally connected circuits. Different materials may offer

advantages in special situations, but the majority of laboratory and instru-

mentation applications are dealt with using silicon and InGaAs devices. With

this information, we will now investigate how to use photodiodes in practice to

perform useful measurements.

18 Chapter One

Photodetection Basics

Chapter

Ampliﬁed Detection Circuitry

2.1 Introduction

In this chapter we progress from photodetectors to photodetection, simple elec-

tronic circuits that allow observation and measurement of static and varying

optical signals with a voltmeter, on an oscilloscope, or as part of an optoelec-

tronic product. We saw in Chap. 1 that the signal output from a photodiode is

strongly inﬂuenced by the circuitry connected to it. In particular, unless the

load voltage is much smaller than a few tens of millivolts, simple connection to

a resistive load leads to a logarithmic or at least highly nonlinear response,

while reverse biasing into the third quadrant of the diode’s IV characteristic

generally leads to a current-output response that is linear over many orders of

magnitude of incident power. This reverse bias connection is the basis of much

optical measurement technology. In fact, the simple circuit seen in Fig. 1.8 can

be called our ﬁrst “instrument.”

2.2 The Bias Box

Figure 2.1 shows a photodiode reverse biased by a small battery, with a series

load resistor whose imposed voltage can be read with a high impedance volt-

meter or oscilloscope. With the addition of a multiway switch to change load

resistors over a wide range, say from 100W to 1 MW, and perhaps a changeover

switch to swap the detector polarity, this “equipment” should be in every optical

researcher’s kit bag. I try to keep several available. A 3.3-V lithium cell or 6-V

camera battery is unlikely to damage the majority of silicon photodiodes

through reverse breakdown, but check the photodiode’s voltage ratings ﬁrst to

be sure. However, it is possible to pass excessive current if the photodiode is

brightly illuminated on the low load resistance settings, so connect it ﬁrst,

monitor the voltage, and then illuminate. The bias box can be used with almost

any detector that is handy. It has an easily varied sensitivity and responds lin-

early to power changes. It never oscillates, and the battery voltage is almost

Source: Photodetection and Measurement

剩余246页未读，继续阅读

xiaojia729

粉丝: 2
资源: 6

光子探测优化：深入理解与应用

优化单光子测距系统性能：理论与实验验证

光子计数探测器畸变多项式校正效果实证：2.5像素与1.2像素误差

光电探测器详解：光电倍增管的工作原理与应用

TCSPC实现的单光子探测时间相关计数

基于时间相关单光子计数技术的高分辨率激光测距系统

最优化ToF飞行时间测距，需要一个硅光电倍增管-综合文档

Onsemi_最优化ToF飞行时间测距，需要一个硅光电倍增管-综合文档

MEMS材料特性检测：全光学测试与波导测试站

X射线闪烁晶体光学性能光电测试技术研究

激光雷达技术解析：相干与非相干探测

最新资源