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>> 峰峰值分辨率与有效值分辨率  翻译稿

作者:kikiki 栏目:模拟技术
研讨翻译>> 峰峰值分辨率与有效值分辨率  翻译稿
请大家指正一下,初次翻译  

___________
这篇文章很基础,对学习ADC很有用,特将整理后的文本直接帖出来,并对楼主为了一个词翻译一篇文章供大家学习表示敬意!!


Peak-to-Peak Resolution Versus Effective Resolution

ADI原文

kikiki原译

[001]
INTRODUCTION
The low bandwidth, high resolution ADCs have a resolution of 16 bits or 24 bits. However, the effective NUMBER of bits of a DEVICE is limited by noise. This varies depending on the OUTPUT word rate and the gain setting used. This parameter is specified by some companies as effective resolution. ANALOG Devices specifies peak-to-peak resolution, which is the NUMBER of flicker-free bits and is calculated differently from effective resolution. This application note distinguishes between peak-to-peak resolution and effective resolution.

[001]
导言
低带宽,高分辨率的ADC有16 位或24 位的分辨率。然而,噪声会限制器件的有效位数。这种情况 视输出字速率和增益设定应用 而改变。一些公司把这个参数作为有效值来详细说明。ADI 说明的是峰峰值分辨率,指的就是无跳动的位数,并且不同于有效值分辨率来计算。这篇应用笔记区别了峰峰值分辨率和有效值分辨率。

[002]
Noise
Figure 1 shows a typical histogram obtained from a sigma-delta ADC when the ANALOG input is grounded. Ideally, for this fixed dc ANALOG input, the OUTPUT code should be zero. However, due to noise, there will be a spread of codes for a constant ANALOG input. This noise is due to THERMAL noise within the ADC and quantization noise due to the ANALOG-to-DIGITAL conversion PROCESS.
The code spread is generally Gaussian in nature.The rms noise is calculated using the curve that results from the histogram, the width of the curve determining the rms noise. A Gaussian curve goes from –infinity to +infinity. However, 99.99% of the codes occur within 6.6 × rms noise. Therefore, the peak noise is 6.6 × rms noise.

[002]
噪声
图1表明了当一个sigma-delta ADC的模拟输入端接地的时候所获得的典型柱状图。理想情况对于固定的直流模拟输入,输出代码应该是0。然而,由于噪声对于恒定的模拟输入其输出代码会散布开数位。这个噪声是ADC内部的热噪声和模数转换处理带来的量化噪声。

代码分布通常是高斯类型(正态分布)的。有效值噪声用来自于柱状图的曲线计算,曲线宽度决定了噪声的有效值。高斯曲线存在于负无穷到正无穷。然而,99.99%的代码存在于6.6 × rms噪声 之内。因此,峰峰值噪声就是6.6 × rms噪声。

[003]
Data sheets normally specify the rms noise. The noise is dependent on the FILTER frequency used and the gain setting used. Normally, the rms noise gets smaller as the ANALOG input range decreases. However, since the full-scale ANALOG input signal is also being reduced, the effective NUMBER of bits degrades.

[003]
数据表通常指定有效值噪声。噪声由滤波频率和增益设置而定。通常的,有效值噪声随着模拟输入范围降低而减小。然而,由于满幅度模拟输入信号也被减少了,有效位数也降低了。





[004]
Peak-to-Peak Resolution Versus Effective Resolution
Most applications do not want to see code flicker on the SYSTEM OUTPUT. For example, in a weigh-scale application, the NUMBER of flicker-free bits is important. The DIGITAL word from the ADC can then be truncated so that the flickering bits are not seen on the weigh-scale MONITOR.

[004]
峰峰值分辨率与有效值分辨率
大多数应用不想看系统输出的代码跳动。例如,在称重应用中,重要的是无跳动的代码数目。来自ADC 输出的数字能被截去尖端,以至于跳动位不会在称重监视器上看到。

[005]
The noise-free resolution or peak-to-peak resolution is determined from the noise VALUEs given in a data sheet. First, let’s calculate the signal-to-noise ratio. The signal-to-noise ratio equals
SNR = 20log(noise/full-scale input)

[005]
无噪声分辨率或峰峰值分辨率由数据表给定的噪声值决定。首先,来计算信噪比。信噪比(噪声信号比)等于
SNR = 20log(noise/full-scale input)

[006]
ANALOG Devices normally specifies the peak-to-peak resolution or noise-free code resolution. This is achieved by calculating the SNR using the peak noise that equals
6.6 × rms noise. From the signal-to-noise calculation, the accuracy can be determined.
SNR = 6.02N + 1.76
= 20log(peak noise/full-scale input)

[006]
ADI通常指定峰峰值分辨率和无噪声代码分辨率。根据峰峰值噪声等于6.6 倍的有效值噪声来计算信噪比。由信噪比的计算,能够决定了精度。
SNR = 6.02N + 1.76
= 20log(peak noise/full-scale input)

[007]
From the AD7719 data sheet, the rms noise equals 1.25 μV when the ANALOG input range is ±2.56 V and the data update rate equals 5.35 Hz. From this data, the signal-to-noise ratio equals

(20log((6.6  1.25E–6)/(2.56 × 2)) = –115.85 dB

From this data, the peak-to-peak resolution is
115.85 = 6.02N + 1.76 =>
N = (115.85 – 1.76)/6.02 = 19 Bits
Therefore, under the above conditions, there will be no flicker in the 19 MSBs.

[007]
AD7719 的数据表中,在数据更新率为5.35Hz,模拟输入范围是±2.56V,有效值噪声等于1.25μV。从这个数据中,信噪比等于

(20log((6.6×1.25×10-6)/2.56×2))=-115.85dB

从这个数据中,峰峰值分辨率是

115.85=6.02N+1.76  => N=(115.85-1.76)/6.02=19bit

因此,在上面条件下,无跳动码是19 最高有效位。

[008]
Some companies specify effective resolution rather than peak-to-peak resolution. The effective resolution is calculated using the rms noise rather than the peak noise. Using the rms noise leads to a signal-to-noise ratio of
(20log((1.25E–6)/(2.56×2))
= – 132.25 dB

[008]
相比于峰峰值分辨率,一些公司更愿意指定有效值分辨率。有效值分辨率用有效值噪声计算而不用峰峰值噪声。用有效值噪声得出信噪比这将得出有效值分辨率

(20log((1.25×10-6)/(2.56×2))=132.25dB

[009]
This leads to an effective resolution of
132.25 = 6.02N + 1.76 =>
N = (132.25 – 1.76)/6.02 = 21.7 Bits
Therefore, the effective resolution equals the peak-to-peak resolution + 2.7 bits.

[009]
这将得出有效值分辨率

132.25=6.02N+1.76 => N=(132.25-1.76)/6.02=21.7bit

因此,有效值分辨率等于峰峰值分辨率 +2.7 位。

[010]
It is important when evaluating ADCs to know that the effective resolution and peak-to-peak resolution are calculated differently, the effective resolution giving a VALUE that is greater than the peak-to-peak resolution by 2.7 bits. Also, the effective resolution does not highlight the amount of bits that flicker. The peak-to-peak resolution gives a better indication of performance since it indicates the NUMBER of bits that will not flicker.

[010]
评估ADC 的时候知道有效值分辨率和峰峰值分辨率的计算不同是很重要的。有效值分辨率给出的值比峰峰值分辨率大2.7 位。同样,有效值分辨率不能突出跳动位数的数量。峰峰值分辨率给出了更好的性能指示,因为它指出了无跳动的位数的数目。


[011]
Summary
The method in which the effective NUMBER of bits of low bandwidth, high resolution ADCs is calculated differs from company to company. Therefore, when comparing DEVICEs from different companies for an application, it is not valid to compare NUMBERs directly because parts that are specified using peak-to-peak resolution appear worse than par
2楼: >>参与讨论
asunmad
在哪呢?
 
3楼: >>参与讨论
NE5532
我就在线等着你呢。呵呵。
 
4楼: >>参与讨论
电子小虫
楼主,文章在哪儿?
 
5楼: >>参与讨论
kikiki
NE5532,你好,我已把文档发到你邮箱里,帮忙发一下,谢谢
 
6楼: >>参与讨论
ic921
谢谢你的参与 !欢迎你的参与 !
请给一下原文或原文地址,好让大家对照点评。

--在我的记录里,没有AN615(ADI),不知是哪家公司或哪里的文章。

7楼: >>参与讨论
NE5532
期待原文。
 
8楼: >>参与讨论
ic921
把首帖链接的内容帖一下

kikiki峰峰值分辨率与有效值分辨率(AN-615

导言
低带宽,高分辨率的ADC有16 位或24 位的分辨率。然而,噪声会限制器件的有效位数。这种情况 视输出字速率和增益设定应用 而改变。一些公司把这个参数作为有效值来详细说明。ADI 说明的是峰峰值分辨率,指的就是无跳动的位数,并且不同于有效值分辨率来计算。这篇应用笔记区别了峰峰值分辨率和有效值分辨率。
噪声
图1表明了当一个sigma-delta ADC的模拟输入端接地的时候所获得的典型柱状图。理想情况对于固定的直流模拟输入,输出代码应该是0。然而,由于噪声对于恒定的模拟输入其输出代码会散布开数位。这个噪声是ADC内部的热噪声和模数转换处理带来的量化噪声。代码分布通常是高斯类型(正态分布)的。有效值噪声用来自于柱状图的曲线计算,曲线宽度决定了噪声的有效值。高斯曲线存在于负无穷到正无穷。然而,99.99%的代码存在于6.6 x rms噪声 之内。因此,峰峰值噪声就是6.6 x rms噪声。
数据表通常指定有效值噪声。噪声由滤波频率和增益设置而定。通常的,有效值噪声随着模拟输入范围降低而减小。然而,由于满幅度模拟输入信号也被减少了,有效位数也降低了。

2005212244360.jpg">

峰峰值分辨率与有效值分辨率
大多数应用不想看系统输出的代码跳动。例如,在称重应用中,重要的是无跳动的代码数目。来自ADC 输出的数字能被截去尖端,以至于跳动位不会在称重监视器上看到。无噪声分辨率或峰峰值分辨率由数据表给定的噪声值决定。首先,来计算信噪比。信噪比(噪声信号比)等于

SNR=20log(noise/full-input)

ADI通常指定峰峰值分辨率和无噪声代码分辨率。根据峰峰值噪声等于6.6 倍的有效值噪声来计算信噪比。由信噪比的计算,能够决定了精度。
AD7719 的数据表中,在数据更新率为5.35Hz ,模拟输入范围是±2.56V ,有效值噪声等于1.25μV 。从这个数据中,信噪比等于

(20log((6.6×1.25×10-6)/2.56×2))=-115.85dB

从这个数据中,峰峰值分辨率是

115.85=6.02N+1.76  => N=(115.85-1.76)/6.02=19bit

因此,在上面条件下,无跳动码是19 最高有效位。
相比于峰峰值分辨率,一些公司更愿意指定有效值分辨率。有效值分辨率用有效值噪声计算而不用峰峰值噪声。用有效值噪声得出信噪比这将得出有效值分辨率

(20log((1.25×10-6)/(2.56×2))=132.25dB

这将得出有效值分辨率

132.25=6.02N+1.76 => N=(132.25-1.76)/6.02=21.7bit

因此,有效值分辨率等于峰峰值分辨率 +2.7 位。
评估ADC 的时候知道有效值分辨率和峰峰值分辨率的计算不同是很重要的。有效值分辨率给出的值比峰峰值分辨率大2.7 位。同样,有效值分辨率不能突出跳动位数的数量。峰峰值分辨率给出了更好的性能指示,因为它指出了无跳动的位数的数目。
总结
不同公司的低带宽高分辨率ADC 有效位数计算方法不同。因此,对于一个应用,当比较不同公司的器件时候,直接比较位数是不正确的。因为用峰峰值分辨率指定的部分比用有效值指定的部分表现更差些。对于一块ADC,它的有效值分辨率比峰峰值分辨率好2.7 位。因此,22位有效值分辨率的器件具有22-2.7=19.3bit 的无跳动分辨率。确定指定的是峰峰值分辨率还是有效值分辨率是很重要的,以利于应用时候了解器件的真正性能。
ZFY
2005 年1月30 日 

* - 本贴最后修改时间:2005-2-1 23:35:01 修改者:IC921

9楼: >>参与讨论
电子小虫
这个原文在什么地方?
 
10楼: >>参与讨论
akaer
回kikiki、小虫、IC921
你的译稿我刚下载,粗看了一遍,感觉不错!我打算对照原文再细看看,共同提高!

小虫,IC921,原文我已经找出,附上。

* - 本贴最后修改时间:2005-2-2 21:39:47 修改者:akaer

http://www.21icsearch.com/buzi/upimage/upfile/2005222143500.pdf

11楼: >>参与讨论
asunmad
翻译得真不错!
 
12楼: >>参与讨论
电子小虫
不错,原文很有价值,翻译的也很好。
这样的文档好,短小精悍,对翻译者和读者来说都比较容易入门。
呵呵,对我这样的懒人挺合适的。

13楼: >>参与讨论
akaer
回kikiki
译文我已对照读完,没发现翻译错误,感觉你忠实原文,名词术语统一、比较准确,的确是篇好作品!

提出几个地方探讨:
1。第1段以及后面公式推导过程中出现的几个specify,建议翻译成“定义”;
2。第2段,However,due to noise...一句意思到了,但感觉不够顺;
3。第5段,...can then be truncated...一句,是否可以译为“滤掉”或干脆意译为“软件滤波”?
4。最后1段,PART做“器件”理解更贴切。



14楼: >>参与讨论
ic921
译得不错! 为方便,改一下版式,提一点意见
Peak-to-Peak Resolution Versus Effective Resolution

ADI原文

kikiki原译

[001]
INTRODUCTION
The low bandwidth, high resolution ADCs have a resolution of 16 bits or 24 bits. However, the effective NUMBER of bits of a DEVICE is limited by noise. This varies depending on the OUTPUT word rate and the gain setting used. This parameter is specified by some companies as effective resolution. ANALOG Devices specifies peak-to-peak resolution, which is the NUMBER of flicker-free bits and is calculated differently from effective resolution. This application note distinguishes between peak-to-peak resolution and effective resolution.

[001]
导言
低带宽,高分辨率的ADC有16 位或24 位的分辨率。然而,噪声会限制器件的有效位数。这种情况 视输出字速率和增益设定应用 而改变。一些公司把这个参数作为有效值来详细说明。ADI 说明的是峰峰值分辨率,指的就是无跳动的位数,并且不同于有效值分辨率来计算。这篇应用笔记区别了峰峰值分辨率和有效值分辨率。

[002]
Noise
Figure 1 shows a typical histogram obtained from a sigma-delta ADC when the ANALOG input is grounded. Ideally, for this fixed dc ANALOG input, the OUTPUT code should be zero. However, due to noise, there will be a spread of codes for a constant ANALOG input. This noise is due to THERMAL noise within the ADC and quantization noise due to the ANALOG-to-DIGITAL conversion PROCESS.
The code spread is generally Gaussian in nature.The rms noise is calculated using the curve that results from the histogram, the width of the curve determining the rms noise. A Gaussian curve goes from –infinity to +infinity. However, 99.99% of the codes occur within 6.6 × rms noise. Therefore, the peak noise is 6.6 × rms noise.

[002]
噪声
图1表明了当一个sigma-delta ADC的模拟输入端接地的时候所获得的典型柱状图。理想情况对于固定的直流模拟输入,输出代码应该是0。然而,由于噪声对于恒定的模拟输入其输出代码会散布开数位。这个噪声是ADC内部的热噪声和模数转换处理带来的量化噪声。

代码分布通常是高斯类型(正态分布)的。有效值噪声用来自于柱状图的曲线计算,曲线宽度决定了噪声的有效值。高斯曲线存在于负无穷到正无穷。然而,99.99%的代码存在于6.6 × rms噪声 之内。因此,峰峰值噪声就是6.6 × rms噪声。

[003]
Data sheets normally specify the rms noise. The noise is dependent on the FILTER frequency used and the gain setting used. Normally, the rms noise gets smaller as the ANALOG input range decreases. However, since the full-scale ANALOG input signal is also being reduced, the effective NUMBER of bits degrades.

[003]
数据表通常指定有效值噪声。噪声由滤波频率和增益设置而定。通常的,有效值噪声随着模拟输入范围降低而减小。然而,由于满幅度模拟输入信号也被减少了,有效位数也降低了。





[004]
Peak-to-Peak Resolution Versus Effective Resolution
Most applications do not want to see code flicker on the SYSTEM OUTPUT. For example, in a weigh-scale application, the NUMBER of flicker-free bits is important. The DIGITAL word from the ADC can then be truncated so that the flickering bits are not seen on the weigh-scale MONITOR.

[004]
峰峰值分辨率与有效值分辨率
大多数应用不想看系统输出的代码跳动。例如,在称重应用中,重要的是无跳动的代码数目。来自ADC 输出的数字能被截去尖端,以至于跳动位不会在称重监视器上看到。

[005]
The noise-free resolution or peak-to-peak resolution is determined from the noise VALUEs given in a data sheet. First, let’s calculate the signal-to-noise ratio. The signal-to-noise ratio equals
SNR = 20log(noise/full-scale input)

[005]
无噪声分辨率或峰峰值分辨率由数据表给定的噪声值决定。首先,来计算信噪比。信噪比(噪声信号比)等于
SNR = 20log(noise/full-scale input)

[006]
ANALOG Devices normally specifies the peak-to-peak resolution or noise-free code resolution. This is achieved by calculating the SNR using the peak noise that equals
6.6 × rms noise. From the signal-to-noise calculation, the accuracy can be determined.
SNR = 6.02N + 1.76
= 20log(peak noise/full-scale input)

[006]
ADI通常指定峰峰值分辨率和无噪声代码分辨率。根据峰峰值噪声等于6.6 倍的有效值噪声来计算信噪比。由信噪比的计算,能够决定了精度。
SNR = 6.02N + 1.76
= 20log(peak noise/full-scale input)

[007]
From the AD7719 data sheet, the rms noise equals 1.25 μV when the ANALOG input range is ±2.56 V and the data update rate equals 5.35 Hz. From this data, the signal-to-noise ratio equals

(20log((6.6  1.25E–6)/(2.56 × 2)) = –115.85 dB

From this data, the peak-to-peak resolution is
115.85 = 6.02N + 1.76 =>
N = (115.85 – 1.76)/6.02 = 19 Bits
Therefore, under the above conditions, there will be no flicker in the 19 MSBs.

[007]
AD7719 的数据表中,在数据更新率为5.35Hz,模拟输入范围是±2.56V,有效值噪声等于1.25μV。从这个数据中,信噪比等于

(20log((6.6×1.25×10-6)/2.56×2))=-115.85dB

从这个数据中,峰峰值分辨率是

115.85=6.02N+1.76  => N=(115.85-1.76)/6.02=19bit

因此,在上面条件下,无跳动码是19 最高有效位。

[008]
Some companies specify effective resolution rather than peak-to-peak resolution. The effective resolution is calculated using the rms noise rather than the peak noise. Using the rms noise leads to a signal-to-noise ratio of
(20log((1.25E–6)/(2.56×2))
= – 132.25 dB

[008]
相比于峰峰值分辨率,一些公司更愿意指定有效值分辨率。有效值分辨率用有效值噪声计算而不用峰峰值噪声。用有效值噪声得出信噪比这将得出有效值分辨率

(20log((1.25×10-6)/(2.56×2))=132.25dB

[009]
This leads to an effective resolution of
132.25 = 6.02N + 1.76 =>
N = (132.25 – 1.76)/6.02 = 21.7 Bits
Therefore, the effective resolution equals the peak-to-peak resolution + 2.7 bits.

[009]
这将得出有效值分辨率

132.25=6.02N+1.76 => N=(132.25-1.76)/6.02=21.7bit

因此,有效值分辨率等于峰峰值分辨率 +2.7 位。

[010]
It is important when evaluating ADCs to know that the effective resolution and peak-to-peak resolution are calculated differently, the effective resolution giving a VALUE that is greater than the peak-to-peak resolution by 2.7 bits. Also, the effective resolution does not highlight the amount of bits that flicker. The peak-to-peak resolution gives a better indication of performance since it indicates the NUMBER of bits that will not flicker.

[010]
评估ADC 的时候知道有效值分辨率和峰峰值分辨率的计算不同是很重要的。有效值分辨率给出的值比峰峰值分辨率大2.7 位。同样,有效值分辨率不能突出跳动位数的数量。峰峰值分辨率给出了更好的性能指示,因为它指出了无跳动的位数的数目。


[011]
Summary
The method in which the effective NUMBER of bits of low bandwidth, high resolution ADCs is calculated differs from company to company. Therefore, when comparing DEVICEs from different companies for an application, it is not valid to compare NUMBERs directly because parts that are specified using peak-to-peak resolution appear worse than parts that are specified using effective resolution. For an ADC, its effective resolution will be better than its peak-to-peak resolution by 2.7 bits. Therefore, a DEVICE that has
15楼: >>参与讨论
asunmad
[007]there will be no flicker in the 19 MSBs
there will be no flicker in the 19 MSBs
译为“最高有效位为19位时无跳动码”不妥,这里MSB用了复数,是高19位都无跳动。所以“无跳动码是19最高有效位”和“数据在19位或19位以下无跳动码”都还比较贴愿意,但后都没有表明从哪边数19位。
也可以译为:
在19个最高有效位中无跳动。

另外:
[001]
This varies depending on the OUTPUT word rate and the gain setting used.
这种情况视输出字速率和所用的增益设定而改变。

16楼: >>参与讨论
ic921
关于 in the 19 MSBs
我认为,为让人们更加容易理解或不至于造成误会,MSB或MSBs的“最高有效位”不必译出。

大家知道,跳动码并不是最高有效位,而是最低有效位(也可以理解为BCD码的最低4位)。从这点上看,如果将MSB或MSBs的“最高有效位”译出,不是容易造成误会就是容易造成误解。---在我看来,不译比译更好。

______________
今晚回老家过年,估计上网有一定的困难。各位节后见,顺祝大家春节快乐!!

17楼: >>参与讨论
kikiki
大家的建议对我很有好处
这是我在测试AD7656过程中,和ADI应用工程师谈到直流信号转换结果时候,
他们提到的一个名词,当时我不明白,就上网翻译了一下此文,便于自己
弄清楚,大家的指正都很好,我翻译模棱两可的地方确实是自己明白了意思,
找不到合适的词汇来达到 '信‘,个人认为原文讲的还是很透彻的。

18楼: >>参与讨论
ic921
可否明确一下是哪个词?
 
19楼: >>参与讨论
xuhm
谢谢各位老师的努力。
 
参与讨论
昵称:
讨论内容:
 
 
相关帖子
这么对这个小电压测试?
谁有YXD-12864A的资料啊!急
如何从电话线上取电又不妨碍电话使用
谁知道THD的全拼?
如何用运放接成射随器
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