登录 免费注册 首页 | 行业黑名单 | 帮助
维库电子市场网
技术交流 | 电路欣赏 | 工控天地 | 数字广电 | 通信技术 | 电源技术 | 测控之家 | EMC技术 | ARM技术 | EDA技术 | PCB技术 | 嵌入式系统
驱动编程 | 集成电路 | 器件替换 | 模拟技术 | 新手园地 | 单 片 机 | DSP技术 | MCU技术 | IC 设计 | IC 产业 | CAN-bus/DeviceNe

研讨翻译:大海捞针----存在于高共模电压的小差分电压测量

作者:ic921 栏目:模拟技术
研讨翻译:大海捞针----存在于高共模电压的小差分电压测量

Finding the Needle in a Haystack: Measuring small differential voltages in the presence of large common-mode voltages

大海捞针:存在于高共模电压的小差分电压测量


by Scott WAYNE

译注:本文下载地址已经不记得了,根据文内文字,可能是[ANALOG Dialogue 34-1 (2000)]的文章,如因此造成不便,请读者原谅。
导读:这篇文章描述风格独特,有的地方讲得非常好,值得一读。

Introduction ||概述


[01]

[01]

In applications such as motor CONTROL, POWER-supply current monitoring, and BATTERY cell-voltage monitoring, a small differential voltage must be sensed in the presence of a high common-mode voltage. Some of these applications require galvanic isolation, others do not. Some applications use ANALOG CONTROL, others use DIGITAL CONTROL. Four cases of such measurements will be considered, each requiring unique considerations. They are:

对象马达控制、电源电流监视、电池电压监视之类的应用,必须把一个存在于高共模电压的小差分电压提取出来。部分应用要求galvanic隔离,而其它的则没有。有的应用是模拟控制,有的则是模拟控制。这种测量有4种情况要考虑,each requiring unique considerations。它们是:
[译注]从词典上看,galvanic首先应该指意大利的医师和生理学家Luigi Galvani (1737~1798),可能是他发现的直流刺激疗法后才有这个词用到电学领域中来。在这里,我认为galvanic就是直流或直流电的意思。不知道对不对?

[02]

[02]

1) galvanic isolation with ANALOG OUTPUT;
2) galvanic isolation with DIGITAL OUTPUT;
3) no galvanic isolation, ANALOG OUTPUT;
4) no galvanic isolation, DIGITAL OUTPUT.

有模拟输出的galvanic隔离
有数字输出的galvanic隔离
无galvanic隔离,但有模拟输出
无galvanic隔离,但有数字输出

Differential Signals Versus Common-Mode Signals || 差分信号vs共模信号


[03]

[03]

Figure 1 shows the input of a measurement SYSTEM. VDIFF represents the differential voltage, the signal of interest. VCM represents the common-mode voltage, which contains no useful information about the measurement and could in fact reduce the measurement accuracy. The common-mode voltage could be an implicit PART of the measurement SYSTEM, as in a BATTERY cell-voltage monitoring application, or it could be created by a fault condition where the sensor accidentally comes in contact with a high voltage. In either case, that voltage is unwanted, and it is the job of the measurement SYSTEM to REJECT it, while responding to the differential-mode voltage.

如图1所示的测量系统的输入,VDIFF表示差分电压,它是有用信号。VCM表示共模电压,它携带的信号对测量没有用处,且会降低测量精度。象电池电压监视的应用,共模电压是测量系统所隐含的。它会增加传感器附带包括有高电压的不利条件。另外的原因是该电压没有用,是测量系统工作时所抑制的,而只响应差模电压。





Figure 1. Measurement SYSTEM with differential and common-mode voltages.
图1:有差分和共模电压的测量系统

Common-Mode Rejection (CMR) || 共模抑制


[04]

[04]

The measurement SYSTEM has both a differential-mode gain and a common-mode gain. The differential-mode gain is usually greater than or equal to one, while the common-mode gain is ideally zero. RESISTOR mismatches cause the dc gain from the inverting input to differ slightly from that of the noninverting input. This, in turn, results in a dc common-mode gain that is nonzero. If the differential gain is G = R2/R1 , the common-mode gain will be (%mismatch/100)×[G/(G+1)] . The common-mode REJECTion ratio (CMRR) is the differential-mode gain divided by the common-mode gain, or (100/%mismatch )×(G + 1). The logarithmic equivalent (CMR—in dB), is 20log10[(100/%mismatch )×(G + 1)]

测量系统是共模增益和差模增益并存的。差模通常大于等于1;而理想的共模增益应当为0。电阻失配会引起反相输入和同相输入有轻微的差异,从而导致直流共模增益不为0。若差模增益为:G = R2/R1,则共模增益为(%mismatch/100)×[G/(G+1)]:。共模抑制比CMRR等于差模共模增益除以共模增益,或(100/%mismatch )×(G + 1)。CMR用对数等效表示为分贝(dB):20log10[(100/%mismatch )×(G + 1)]。

[05]

[05]

In real-world applications, external interference sources abound. Pickup will be coupled from the ac POWER LINE (50/60 Hz and its harmonics), from EQUIPMENT switching on and off, and from radio-frequency transmission sources. This type of interference is induced equally into both differential inputs, and therefore appears as a common-mode signal. So, in addition to high dc CMR, instrumentation amplifiers also require high ac CMR, especially at LINE frequencies and their harmonics. DC common-mode errors are mostly a function of RESISTOR mismatch. In contrast, ac common-mode errors are a function of differences in phase shifts or time delays between the inverting and noninverting inputs. These can be minimized by using well-matched high-speed components, and they can be trimmed with a CAPACITOR. Alternatively, in low-frequency applications, OUTPUT filtering can be used if necessary. While dc common-mode errors can usually be removed through calibration or trimming. AC common-mode errors, which can reduce the resolution of the measurement, are generally of greater concern. All ANALOG Devices instrumentation amplifiers are fully specified for both dc and low-frequency ac common-mode REJECTion.

在实际应用中,外部干扰源大量存在。电源线的50/60Hz干扰及其谐波、用电设备的开或关、无线频率发射源会耦合进系统中来。这种类型的干扰一并来到两个差分输入端,因而出现一个共模信号。可见,仪表放大器除了有高的直流共模抑制还要有高的交流共模抑制,尤其是电源频率及其谐波。直流共模误差主要是电阻失配的函数,对应地,交流共模误差则是同相反相输入端之间的相移、延时差异的函数。(这些误差),用匹配良好的高速元件可以最小化,也可以用电容进行修调,此外(Alternatively),对低频应用,如果需要,还可以进行输出滤波。直流共模增益误差通常可以通过校准或修调(trimming)技术消除,而降低测量精度的交流共模误差却更加从令人关注。AD公司所有的仪表放大器都对直流共模抑制和低频的交流共模抑制有完整的详细规定。

[译者语]Alternatively指的应该是前面说的元件匹配和后面说的滤波。

Galvanic Isolation || Galvanic隔离


[06]

[06]

Some applications require that there be no direct ELECTRICal connection between the sensor and the SYSTEM ELECTRONICS. These applications require galvanic isolation in order to protect the sensor, the SYSTEM, or both. The SYSTEM ELECTRONICS may need to be protected from high voltages at the sensor. Or, in applications requiring intrinsic safety, the sensor excitation and POWER circuitry may need to be isolated to prevent sparks or the ignition of explosive gases that could be caused by a fault condition. In medical applications, such as electrocardiograms (ECG), protection is required in both directions. The patient must be protected from accidental ELECTRIC shock. If the patient’s heart stops beating, the ECG machine must be protected from the very high voltages applied to the patient by emergency use of a defibrillator in an attempt to restore the heartbeat.

有的应用,传感器和系统(器件)之间要求不能有直接的电气连接。这些
2楼: >>参与讨论
ic921
研讨翻译:大海捞针----存在于高共模电压的小差分电压测量 -2/2-

High Impedance Versus Galvanic Isolation || 高阻抗vs Galvanic隔离


[09]

[09]

Many applications need the ability to sense a small differential voltage in the presence of a high common-mode voltage, but do not require the intrinsic safety or the ability to break ground loops that are provided by galvanic isolation. These applications require a high-CMR AMPLIFIER that can accept high common-mode voltage. This type of AMPLIFIER, sometimes called a “poor man’s isolation AMPLIFIER,” isolates the sensor from the SYSTEM with a high impedance, rather than with a galvanic isolation barrier. While not isolation in the true sense, it can serve the same purpose in some applications at much lower cost. In addition, a dc-to-dc converter is not required, since the whole SYSTEM is POWERed from the same POWER supply.

许多应用需要这样的能力,以感测存在于高共模电压中的小差分电压,但它并没有本质安全的要求或在提供galvani隔离的情况下允许切断地环路。此类应用就需要能接受高共模电压的高共模抑制的放大器。这种类型的放大器,有时称之为“穷人的放大器”,以高阻抗将传感器与系统隔离而不用galvanic隔离栅。如果实际的感测不隔离就在某些情况下可以以较低的费用获得同样的效果。不要DC-DC转换器的另一种情形,是因为整个系统都由同一电源供电。

[译者语]While not isolation in the true sense 中,true一词似乎有特殊的意义,那种情景/情形实在想象不出!

[10]

[10]

Figure 2 shows the AD629, a high-common-mode-voltage difference AMPLIFIER that was designed for these types of applications. It seems simple enough. It’s “just” an op amp and five resistors. Can’t users “roll their own?” Yes, but the resistors would have to be matched to better than 0.01% and would have to track to better than 3 ppm/℃. RESISTOR self-heating would degrade dc CMR, while capacitive strays would degrade ac CMR. Performance, size, and cost would all be sacrificed compared to what could be obtained in an 8-lead DIP or SOIC.

如图2所示的高共模电压差分放大器AD629就是为这种类型的应用而设计的。它显得足够简单,正好是1个运算放大器加5个电阻。Can’t users “roll their own?”是的,但电阻的匹配度优于0.01%,且有优于3ppm/℃的(温度循迹效果)。电阻自热会降低直流共模抑制DC-CMR,而杂散电容(capacitive strays)则会降低交流共模抑制AC-CMR。牺牲性能、尺寸、价格compared to what could be obtained in an 8-lead DIP or SOIC
[译者语]Can’t users “roll their own?”这句不会译。
另,杂散电容一般是stray capacitive,这里capacitive strays有什么不同吗?

[11]

[11]

Applications such as simple industrial process-CONTROL loops with ANALOG inputs and OUTPUTs that require galvanic isolation could use the AD202/AD204. These are complete isolation AMPLIFIERs with galvanic isolation between the input and OUTPUT stages. Transformer coupling means that they can also provide isolated POWER to the input stage, eliminating the need for an external dc-to-dc converter. The AD202/AD204 provide an uncommitted op amp for input signal conditioning, have CMR of 130 dB at a gain of 100, and 2000-V-peak CMV isolation. Figure 3 shows an AD202 CIRCUIT* to measure a ± 5-V full-scale signal riding on a common-mode voltage of up to 2000 V. For applications that require isolated BRIDGE excitation, cold-junction compensation, linearization, and other signal-conditioning functions, the 3B, 5B, 6B, and 7B series provide a family of complete, well-isolated signal conditioners.

象有需要模拟galvanic隔离输入输出的简单工艺控制环这样的应用,可以用AD202/AD204。这些是输入级和输出级之间完全隔离的galvanic隔离放大器。变压器耦合意味着给输入级提供电源隔离,免去了额外的DC-DC转换器。The AD202/AD204 provide an uncommitted op amp for input signal conditioning,在增益为100时有130dB的共模抑制和2000V的峰值电压隔离。如图3的AD202电路*测量±5V的满度信号所承受(riding)的共模电压高达2000V。
在需要隔离桥激励、冷端补偿、线性化和其它信号调理功能的应用,the 3B, 5B, 6B, and 7B series provide a family of complete, well-isolated signal conditioners.

[译者语]前一个保留未译的一句,主要是对uncommitted的含意不解。后一句可能与产品系列有关,未详查,故未译。

[12]

[12]

*These Figures are illustrative examples; they are not detailed schematics of tested applications. Please consult PRODUCT data sheets for more information. You will also find the online seminar notes, Practical ANALOG Design Techniques, and the book, Practical Design Techniques for Sensor Signal Conditioning (available from ADI), to be useful sources of design information. Use extreme caution when working with high-voltage CIRCUITs.

*这些图只是具体的实例,但不详细描述它们的测试应用原理。要获取更多更详尽的信息,请参考产品数据手册。你也可以查找在线研讨会笔记《Practical ANALOG Design Techniques》和索取《Practical Design Techniques for Sensor Signal Conditioning》一书(ADI有)的有用信息。Use extreme caution when working with high-voltage CIRCUITs.




Figure 2. AD629 High-common-mode-voltage-difference AMPLIFIER.



Figure 3. AD202/AD204 used in application requiring galvanic isolation and ANALOG OUTPUTs.


[13]

[13]

Some industrial sensor applications require galvanic isolation, combined with the DIGITAL OUTPUT of a SMART sensor. DIGITAL isolation, rather than ANALOG isolation, could be used more cost-effectively but an external dc-to-dc converter is required. An example of this sort of application is in motor CONTROL, where a fault condition in the motor could destroy the CONTROL ELECTRONICS. The AD7742 synchronous voltage-to-frequency converter could be used, together with an opto-coupler and a dc-to-dc converter, as shown in Figure 4. A remote AD7742 can be interfaced with a SYSTEM microprocessor or microCONTROLler to complete the A/D conversion. For stand-alone applications the serial-OUTPUT AD7715 ANALOG front end, a 16-bit sigma-delta A/D converter, could be used, but it has five DIGITAL lines to isolate, rather than the SINGLE DIGITAL OUTPUT from the V/F converter. However, instead of five opto-couplers and a dc-to-dc converter, an AD260 five-channel high-speed LOGIC isolator with its own on-board transformer could be used. Figure 5 shows the AD7715 and AD260.

部分工业传感器应用要求有galvanic隔离,并和智能传感器的数字输出结合到一起。用数字隔离而不用模拟隔离可以更节省但需要外加DC-DC转换器。以马达控制一类的应用为例,马达故障情况下就会损坏控制器件。如图4,将AD7742同步电压频率转换器和光耦合器、DC-DC转换器用到一起。由系统微处理器或微控制器接口就能完成A/D转换。可用有模拟前端(处理的)串行输出的16位Σ-Δ模数转换器AD7715,但它有5个数字线要隔离,不单是V/F转换那样只有一条数据信号的输出。然而,用AD260则可以代替5个光耦和一个DC-DC转换器。
[译者提示]同步VFC(AD7742)和异步VFC(如ADVFC32)是不同的。请阅读器件数据手册加以区别。





Figure 4. AD7742 used in application requiring galvanic isolation and DIGITAL OUTPUTs.



Figure 5. AD7715/AD260 used in application requiring galvanic isolation and DIGITAL OUTPUTs.



Figure 6
3楼: >>参与讨论
ic921
这是前天上班时译的
今晚整理一下发给大家评点,请不紊指教!

4楼: >>参与讨论
rfid2005
谢谢搂主,顶
 
5楼: >>参与讨论
rxl8888
请教上述翻译中
电阻失配会引起反相输入和同相输入有轻微的差异,从而导致直流共模增益不为0。若差模增益为:G = R2/R1,则共模增益为(%mismatch/100)×[G/(G+1)]。
请问此处共模增益的推导方法

6楼: >>参与讨论
IC921
自己先试推导一下看看?
 
7楼: >>参与讨论
rxl8888
我推导的结果是共模增益是1+G
推导的过程我一会发上来,ic921帮我查查错
因为是word文档做了公式编辑,粘贴不上来

8楼: >>参与讨论
rxl8888
推导
我不会贴图,就借用IC921的图1

其中定义VDIFF的负端为Vf,正端为Vz,
定义运放的反相输入端为V1,正相输入端V2,运放的输出定义为V0
列出下面连个等式:
(Vf-V1)/R1=(Vf-Vo)/R2        1)式
(Vz-V2)/R1=V2/R2            2)式
2)式-1)式得
[(Vz-Vf)-(V2-V1)]/R1=(V2-V1+Vo)/R2            3)式
由于现在是考虑V2-V1到输出的增益,所以令Vz-Vf=0即没有差分输入,并另Vm=V2-V1

(-Vm)/R1=(Vm+Vo)/R2                4)式
于是得到
Vo/Vm=-(R1+R2)/R1=-(1+G)
所以我考虑共模增益应该是-(1+G)

以上推导请指正


9楼: >>参与讨论
iC921
共模增益不能用差模电压计算VDIFF
只能计算VCM的增益。
------------------
你分数有500分(我没看)帖图的方法是
1 点击编辑框下面的“普通”
2 等待片刻后,粘贴复制好的内容进去就可以了。

我现在就这样做。




Figure 1. Measurement SYSTEM with differential and common-mode voltages.
图1:有差分和共模电压的测量系统
------------------
另一种方法是:将图片链接放到下面的“链接图片”栏中,这不受分数限制。

10楼: >>参与讨论
sheepyang
不知道上面怎么推导的
结果是正如文章所写的。
推导如下。

  考虑电阻值失配,则上面的R1和R2用R1'和R2'来替代。

故根据基尔霍夫电流方程有
  (Vcm-Vdm-V-)/R1'+(Vout-V-)/R2'=0  (1)
   V+=R2*Vcm/(R1+R2)            (2)
   V+=V-                         (3)
差模:Vcm=0,
有   
   Vout/Vdm=R2'/R1'=G

共模: Vdm=0  由上面三个方程得到

  Vout/Vcm=(R2*R1'-R2'*R1)/(R1'*(R1+R2))  (4)

从4式看出如果电阻严格匹配,则共模放大为0。

虽然存在失配,但是R2/R1~R2'/R1'=G.
所以(4)上下各除以R1,有
        (R2/R1-R2'/R1')/(1+R2/R1)   (5)
分子和分母都存在放大倍数的失配,但仍是比较小的量,在分母中可以忽略,因此有
                G*(失配率)/(1+G)
       




    

* - 本贴最后修改时间:2006-1-19 10:02:41 修改者:sheepyang

11楼: >>参与讨论
rxl8888
谢谢楼上和ic921
我推导的思路错误,我将电阻失配引起的共模增益用加在运放正负输入极上的差模电压来计算了,基本功还要加强

12楼: >>参与讨论
wahahacat
re
呵呵.


13楼: >>参与讨论
iC921
水王老乡,请进
 
14楼: >>参与讨论
hotpower
广西老乡谢了,不过现在忽悠成ADS7871
本想直接用模拟的直接采样,看来还得用ADC了...


祝新年快乐,水到渠成.

参与讨论
昵称:
讨论内容:
 
 
相关帖子
请大家帮忙看看怎么这个JFET单级放大的电压放大倍数这么低啊?
请教开关二极管疑问
不共地怎样实现电流传输?
复习一下:电阻星形联接与三角形联接的等效变换
请问HDTV上的宽带视频放大器
免费注册为维库电子开发网会员,参与电子工程师社区讨论,点此进入


Copyright © 1998-2006 www.dzsc.com 浙ICP证030469号