十一放假这两天制作的VU表驱动板，高手请进

sjh327

回复 23# stksyk


    李哥这个板子是你的那个德国调音台线路，效果也不错。

sjh327

这两天比较下驱动电路，发现对大些的表头，还是EE的驱动效果最好，就像用脚尖跳舞一样，非常轻盈敏锐，德国调音台的指针摆动就是中规中矩了。

sjh327

今天又做完了一块板子，这种色调古朴的板子真心喜欢，就是钻孔太费劲了。

不懂我就问

mark

Frank660405

赞一个，留爪学习

denggboo

网上翻到一篇VU的文章，有VU表进化和最终的方案
Introduction  引言
This article is partly based around a piece written for the old website (objectivesounds.co.uk) that was originally put online in 2015. In the interim many people e-mailed in, having read the article and built a simplified version of the improved circuit detailed in the 5th section. The original article described these circuits in a more abstract form and without certain components such as input decoupling which at the time were not shown for simplicity, and it was assumed would added by the constructor if necessary.
本文部分基于为旧网站（objectivesounds.co.uk）撰写的一篇文章，该文章最初于 2015 年上线。在此期间，许多人阅读了这篇文章，并构建了一个简化的改进电路版本，该电路版本详细描述在第 5 节。原始文章以更抽象的形式描述了这些电路，并且没有显示某些组件，如输入去耦，当时出于简化考虑没有显示，并假定构造者如有必要会添加。


VU Meter by Bevan Galbraith
Figure 1. VU meter unit built by Bevan Galbraith
图 1. 贝文·加布里厄斯制作的 VU 表单元

I was lucky enough to receive several very beautiful pictures of readers' builds, such as the one shown in Figure 1. Judging by the general feedback received in the 4 years that the article was available, most people who built the improved circuit built it as a stand-alone unit, or in conjunction with passive circuitry as in Figure 1, a guitar re-amp unit. It became clear that a lot of people were having difficulty with getting the circuitry to work as it was not apparent that these omitted elements may have been necessary. Due to this, the schematics in this article have been updated to include input decoupling, a brief description of a very simple power supply to satisfy the circuit's modest needs, and a more practical focus to ensure that they will be able to work as-is with any line level source.
我很幸运地收到了许多读者构建的非常漂亮的图片，例如图 1 中展示的。根据文章发布 4 年来收到的普遍反馈，大多数构建改进电路的人将其作为独立单元构建，或者如图 1 所示的与无源电路结合，一个吉他重放单元。很明显，很多人在电路工作方面遇到了困难，因为这些省略的元素可能是有必要的。因此，本文中的电路图已经更新，包括输入去耦、一个简单的电源描述，以满足电路的适度需求，以及更实用的焦点，以确保它们可以与任何线路级源直接工作。

There's nothing quite like sitting back and admiring the view of a warmly lit pair of elegant analogue VU meters responding in perfect synchronisation to whatever you're listening to. True, there are many great digital level indicators, even some that attempt to emulate a moving coil meter on-screen. But these can only be used for PC audio, and aesthetics wise, in the opinion of the author, don't even come close to the real thing.
没有什么能比坐下来欣赏一对温暖照亮的优雅模拟 VU 表完美同步响应你所听到的任何内容更令人陶醉了。诚然，有许多优秀的数字电平指示器，甚至有些尝试在屏幕上模仿动圈表。但它们只能用于 PC 音频，从作者的角度来看，在美学上甚至无法与真实的东西相提并论。

VU, or Volume Unit, meters are a way of measuring the level of an audio signal. They measure the an average of the signal they are presented with, not the peak level or the RMS level, which is displayed in decibels in respect to a preset level. VU meters have historically been used where careful level control is required in order to optimise signal levels to fit within a limited dynamic range, and hence can be seen in some form or another on all but the very cheapest tape recorders. Due to their average level measurement, they are much better suited to maximum level monitoring for tape recording, where distortion increases gradually as the audio level reaches and passes full scale. Briefly exceeding full scale with some mild non-linear peak compression may not yield any noticeable impairment. Digital audio recording, which is for all intents and purposes distortion free until the level exceeds the absolute maximum and gross distortion in the form of hard clipping is not well suited for absolute level monitoring as some peaky waveforms, while showing a moderate reading on a VU meter may be easily clipping to the point of audible distortion. This has lead to VU meters earning the moniker of 'virtually useless meter' in these settings.
VU，或称音量单位，是衡量音频信号电平的一种方式。它们测量的是所呈现信号的平均值，而不是峰值电平或均方根电平，后者以分贝的形式相对于预设电平显示。VU 电表在历史上被用于需要仔细控制电平以优化信号电平以适应有限动态范围的地方，因此在除了最便宜的磁带录音机之外的所有录音机中都可以以某种形式看到。由于它们的平均电平测量，它们非常适合磁带录音的最大电平监控，因为随着音频电平达到并超过满量程，失真会逐渐增加。轻微超过满量程并使用一些轻微的非线性峰值压缩可能不会产生任何明显的损害。数字音频录音，在所有意图和目的上都是无失真的，直到电平超过绝对最大值，并以硬剪辑形式出现的大失真，不适合绝对电平监控，因为一些峰值波形，虽然在一个 VU 电表上显示的读数适中，但可能已经失真到可听见的程度。这导致 VU 电表在这些设置中获得了“几乎无用的电表”的绰号。


Classic VU Meter
Figure 2. Classic moving-coil VU meter from a Teac cassette deck
图 2. 来自 Teac 卡式录音机的经典动圈 VU 表

Figure 2 shows a cheap and mass-produced style of VU meter commonly found on consumer units in the days of tape recording, dating back to the late 1970s or early 1980s. Even today, analogue VU have some very useful applications for professional and non-professional use alike. They are invaluable for monitoring average levels and give a good indication of the perceived loudness of the source they are connected to. In professional line level settings where headroom in excess of 10V will be present, the point of peak clipping may well be some 18dB or so above 0VU with typical program material, so they are still an excellent means to set levels analogue line levels by.
图 2 展示了一种廉价的、大规模生产的 VU 表样式，这种样式在磁带录音时代的消费单元中很常见，可追溯到 20 世纪 70 年代末或 80 年代初。即使今天，模拟 VU 表对于专业和非专业用户都有非常实用的应用。它们在监控平均电平方面非常有价值，并且可以很好地指示连接到的源信号的感知响度。在专业线路电平设置中，如果存在超过 10V 的余量，典型的节目材料中峰值削波点可能比 0VU 高约 18dB 左右，因此它们仍然是设置模拟线路电平的极好手段。

In a consumer listening environment, VU metes can be very effectively deployed to show how 'loud' CDs and other digital audio sources have been mastered when connected to a CD player standard audio interface. Fortuitously, a full scale output from the vast majority DACs is around 2V RMS. A VU meter configured for professional line level reads 0dB when the input voltage has an average of 1.125V meaning that a reading of +4dB will be shown when a sine wave of 2V RMS is present, the maximum undistorted level. The same can be done with phono inputs, once past a pre-amplifier with a suitable gain of 37-40dB, and the differences in levels between certain labels and formats may surprise you.
在消费级听音环境中，VU 表可以非常有效地显示 CD 和其他数字音频源通过 CD 播放器标准音频接口时的“响度”如何被处理。幸运的是，大多数 DAC 的全量程输出大约为 2V RMS。当输入电压的平均值为 1.125V 时，专业线路电平配置的 VU 表读数为 0dB，这意味着当存在 2V RMS 的正弦波时，将显示+4dB 的读数，这是最大不失真电平。同样，也可以使用唱机输入，一旦通过具有 37-40dB 合适增益的前置放大器，某些标签和格式之间的电平差异可能会让你感到惊讶。

There is also purely the aesthetic function of having them on a unit, a perfectly valid reason, as long as the meter itself does not degrade audio quality in some of the ways that will be made apparent further on. After all, the end goal of HiFi is enjoyment!
此外，它们在单元上还有纯粹的美观功能，这是一个完全合理的理由，只要仪表本身不会以以下将要说明的方式降低音频质量。毕竟，HiFi 的最终目标是享受！


The VU meter standard
VU 表标准
VU meters have been around for over 75 years now, since Bell Labs and various broadcasting companies in the United States standardised a method of measuring average levels in telephone lines and other signal carrying cables. The standard defined by the ANSI requires fairly strict tolerances in terms of the mechanical aspects of the moving coil meter itself, not withstanding the electrical components. In order to provide any meaningful indication of an AC signal a rectifier must be used to generate a proportional DC voltage across the meter coil. This rectifier consists of fairly esoteric low voltage drop elements so as not to decrease the meter's sensitivity to small signals. All this makes true VU meters rather costly. Even second hand meters can cost upwards of £70 per unit. For a new unit, expect to pay perhaps three times this price. The basic requirements of the standard are listed below.
VU 表已经存在超过 75 年了，自从贝尔实验室和美国的各种广播公司标准化了在电话线路和其他信号传输电缆中测量平均电平的方法以来。ANSI 制定的标准要求移动线圈表计的机械方面有相当严格的公差，不考虑电气元件。为了提供任何有意义的交流信号指示，必须使用整流器在表计线圈上产生成比例的直流电压。这个整流器由一些相当罕见的低电压降元件组成，以避免降低表计对微小信号的灵敏度。所有这些都使得真正的 VU 表相当昂贵。即使是二手表，每台也可能超过 70 英镑。对于新表，预计价格可能是这个价格的 3 倍。标准的基本要求如下。

The rectifier should be a low forward voltage full-wave germanium, or now obsolete copper oxide, type so as not to adversely affect the sensitivity of the meter. This is usually included inside of the meter itself.
整流器应为低正向电压的全波锗型，或现在已淘汰的氧化铜型，以避免对表计的灵敏度产生不利影响。这通常包含在表计内部。
The meter's scale should be adjusted to compensate to accommodate the forward voltage of the rectifier.
应调整表计的刻度以补偿整流器的正向电压。
The meter should display 0dB when presented with a sine wave that has an RMS voltage of either 1.228V (+4dBu) for professional line level or 300mV (-10dBV) for consumer line level.
当向表计提供具有 1.228V（+4dBu）有效值电压的 sine 波时，表计应显示 0dB，该电压为专业线路电平；当提供 300mV（-10dBV）有效值电压的 sine 波时，表计应显示 0dB，该电压为消费级线路电平。
The frequency response of the meter should be down no more than 0.5dB at 25Hz and 25kHz.
表计的频率响应在 25Hz 和 25kHz 时不应低于 0.5dB。
The meter should take 300ms to reach 99% of it's travel in any direction.
表计在任何方向上应达到其行程的 99%所需时间为 300ms。
The meter should undershoot by no less than 1% and overshoot no more than 1.5%
表计应下冲不少于 1%，超调不超过 1.5%。
The meter should read 0VU when a current of 200μA flows through the coil.
表计应在线圈中流过 200μA 电流时读数为 0VU。
The meter should present a load impedance of 7.5kΩ to the drive circuitry at 0VU. In order to achieve this, it is common practice to connect a 3.6kΩ resistor between the source and the rectifier.
表计应在 0VU 时向驱动电路提供 7.5kΩ的负载阻抗。为了实现这一点，通常在源和整流器之间连接一个 3.6kΩ的电阻。
Due to the very tight mechanical tolerances required to do so, it should be apparent that these targets will not be met in the vast majority of domestic (and sometimes even professional) cases. You can rest assured that any VU meters on domestic recording equipment such as cassette and reel to reel recorders will not conform to the specifications listed above. This isn't too great a worry as all that we really want is to see a reading of 0VU when the signal reaches a level where distortion will be likely and does so without any significant overshoot, insensitivity to low levels, or a delay of over 500ms or so.
由于实现这一目标所需的机械公差非常严格，因此很明显，在绝大多数国内（有时甚至是专业）情况下，这些目标将无法实现。您可以放心，任何国内录音设备（如卡式和开盘录音机）上的 VU 表都不会符合上述规格。这并不是一个太大的问题，因为我们真正希望看到的是，当信号达到可能导致失真的水平时，VU 表显示为 0VU，并且不会出现任何显著的过冲、对低电平不敏感或延迟超过 500 毫秒等情况。


Sifam AL20
Figure 3. Sifam AL20 VU meter
图 3. Sifam AL20 VU 表

Having said that, it's always good to try and get as close to the real thing as possible when working with lower cost non-standard meters. In many cases meters that fall even just slightly shy of the standard, but are still excellent, will be priced a factor of two lower than ones that meet it. Such is the meter shown in Figure 3 which costs £50: a most attestably well constructed piece of work, with its snappy movement, black phenolic housing and solid glass window. Further reading will show that there are even a couple of improvements to be obtained by using non-standard active drive circuitry, with low end sensitivity being a major one. A brief allusion to this problem can be seen on the face of the meter in Figure 3; the distance between the -20(dB) threshold and 0% is far smaller than the tenth of the scale that it should be to 100%. When a drive signal of 10% full scale (-20dB) is applied, the needle will only move maybe 2% of full scale.
说到这里，当使用低成本的非标准表时，总是尽量接近真实事物是很好的。在许多情况下，即使只是略微低于标准，但仍然非常好的表，其价格将是符合标准的表的一半。图 3 中所示的表就是这样，价格为 50 英镑：这是一件无可挑剔的精良之作，具有其敏捷的运动、黑色酚醛外壳和坚固的玻璃窗口。进一步阅读将表明，通过使用非标准主动驱动电路，甚至可以获得一些改进，其中低端灵敏度是一个主要因素。对这个问题的一个简短提及可以在图 3 中的仪表表面上看到；-20(dB)阈值和 0%之间的距离远小于它应该达到 100%的十分之一。当施加 10%满量程（-20dB）的驱动信号时，指针只会移动满量程的约 2%。


ANSI VU meter circuit
ANSI VU 表电路
Before having a look at more common active circuitry, it’s helpful to cast a brief glance at the standard VU meter circuit as it has appeared in professional audio equipment over the ages. In most cases the entire circuit is enclosed within the meter housing enabling the meter to be connected directly to the signal to be monitored. It’s quite a simple circuit, but some analysis here will be useful in laying out some of the basic operational fundamentals and traps of circuitry to be described later on.
在查看更常见的活动电路之前，简要地看一下标准 VU 表电路是有帮助的，因为这个电路在专业音频设备中已经存在了很长时间。在大多数情况下，整个电路都封装在仪表外壳内，使得仪表可以直接连接到要监控的信号。这是一个相当简单的电路，但在这里进行一些分析将有助于阐述一些基本操作原理和电路描述中的陷阱。


ANSI VU Meter
Figure 4. Basic ANSI VU meter schematic
图 4. 基本 ANSI VU 表原理图

Figure 4 shows the circuitry that appears in most professional VU meter circuits that meet the ANSI standard. R1 provides the necessary 3.6kΩ inline resistance in order to reduce loading on the drive source and supply the correct amount of current to allow the meter to work accurately. This current is then full-wave rectified by D1, D2, D3 and D4 in the usual arrangement and then fed into the meter, which can only respond to DC current flowing through it. The diodes that make up the rectifier must have a low forward voltage or the low end sensitivity of the meter will suffer greatly, so they are germanium types (or in the case of some very old meters copper oxide rectifiers). If they were to be standard silicon diodes such as the ubiquitous and very low cost 1N4148 types, then the meter will be quite insensitive to peak voltages under 1.2V or so, as the drive signal will have to overcome the forward voltage of two of these diodes before any current can actually make it through to the meter movement. Unfortunately germanium diodes are not the cheapest of components, or the most durable, being particularly susceptible to heat damage during soldering. It may now be possible to use Schottky diodes in such arrangements as they typically have a much lower forward voltage than standard silicon diodes, but in the estimation of the author, they will most likely give a poorer performance here than germanium devices.
图 4 展示了大多数符合 ANSI 标准的专业 VU 电路中的电路。R1 提供了必要的 3.6kΩ串联电阻，以减少对驱动源的负载并供给电表正常工作所需的正确电流。然后，这种电流通过 D1、D2、D3 和 D4 在常规布置中进行全波整流，并输入到电表中，电表只能对通过它的直流电流做出反应。组成整流器的二极管必须具有低正向电压，否则电表的低端灵敏度将大大降低，因此它们是锗型（或者在一些非常古老的电表中是氧化铜整流器）。如果使用标准的硅二极管，如普遍且成本极低的 1N4148 类型，那么电表对约 1.2V 以下的峰值电压将非常不敏感，因为驱动信号必须克服两个这些二极管的正向电压，才能有电流真正通过到电表机构。不幸的是，锗二极管并不是最便宜的元件，也不是最耐用的，尤其是在焊接过程中特别容易受到热损伤。现在可能可以使用肖特基二极管在这样的布置中，因为它们通常具有比标准硅二极管低得多的正向电压，但根据作者估计，它们在这里的性能可能比锗器件更差。

The meter itself has no electrical damping, and in VU meters that are made to conform to the ANSI standard great care is taken to ensure that the mechanical ballistics of the meter's moving assembly will meet the specifications defined in the previous section of this article when used in conjunction with the arrangement in Figure 4.
电表本身没有电气阻尼，在符合 ANSI 标准的 VU 电表中，非常注重确保电表的机械弹道在配合图 4 中的布置使用时，将满足本文前一部分定义的规格。

It is important that the meter is driven from a very low impedance source, otherwise the highly non-linear load that the rectifier presents to the driving circuitry will result in a high level of distortion being added to the audio signal. From personal experience the author can attest that even with a comparatively low line level driving impedance of only 300Ω, audible distortion was present when the meter input was monitored with this configuration. It can therefore be concluded that in order to safely use this simple passive circuit successfully, it must be driven directly from an op-amp output, which will typically be less than 1Ω, if the addition of significant distortion is to be avoided. Typical line impedances will seldom be below 70Ω which makes this unhappy scenario highly likely in practice. In almost all cases it is good practice to drive the meter from a separate output or through a suitable buffer amplifier to completely eliminate this problem, but there is still another distortion trap hiding in the meters ground current which will also be highly distorted. The meter must be grounded to a separate ground, typically a power ground, path than the audio signal otherwise this nonlinear current could potentially generate a small voltage in the ground path that could still add enough distortion to seriously compromise overall non-linearity to the tune of up to 0.1%.
仪表必须从非常低阻抗的源驱动，否则整流器对驱动电路提供的非线性负载将导致音频信号中添加高水平的失真。根据作者的个人经验，即使线驱动阻抗相对较低，仅为 300Ω，当使用此配置监控仪表输入时，仍存在可听见的失真。因此，可以得出结论，为了安全有效地使用这个简单的无源电路，必须直接从运算放大器输出驱动，这通常小于 1Ω，以避免添加显著的失真。典型线路阻抗很少低于 70Ω，这使得在实际中这种情况非常可能。在几乎所有情况下，从单独的输出或通过合适的缓冲放大器驱动仪表都是良好的实践，以完全消除这个问题，但仪表的接地电流中仍隐藏着另一个失真陷阱，这也会高度失真。仪表必须接地到一个单独的接地路径，通常是电源接地路径，而不是音频信号接地，否则这种非线性电流可能会在接地路径中产生一个小的电压，这仍然可能添加足够的失真，严重损害总体的非线性，达到高达 0.1%的程度。


A historic non-standard VU meter
历史性的非标准 VU 表
Having taken into account the great cost of producing a meter that meets the standard defined by the ANSI and various complementary standards to the same effect elsewhere in the world, it should be clear that due to cost limitations a great many of the VU meters found in consumer, and even some professional equipment, will fall short these specifications in some way or another. In practice this does not in fact matter much, and useable examples of non-standard VU meters that break the rules in some way can be found in a plethora of consumer units.
考虑到生产符合 ANSI 标准以及世界各地其他等效标准的仪表的巨大成本，应该清楚的是，由于成本限制，许多在消费类设备中，甚至一些专业设备中发现的 VU 表在某些方面可能无法满足这些规格。实际上，这并不重要，可以在大量消费类设备中找到一些违反某些规则的可用非标准 VU 表的例子。


AKAI 4000DS VU meter circuitry
Figure 5. AKAI 4000DS VU meter circuitry
图 5. AKAI 4000DS VU 表电路

Figure 5 shows just such an example, taken from the service manual of the ever-more-sought-after - for reasons that seem not to have very much to do with high fidelity recording and playback in the 21st century - Akai 4000DS reel to reel tape deck. It’s a fairly standard topology from the ‘golden-era’ of relatively large analogue VU meters on consumer with a few interesting workarounds; historical features that make it worth examining here.
图 5 展示了这样一个例子，取自越来越受欢迎的 Akai 4000DS 开盘磁带录音机的服务手册——原因似乎与 21 世纪的音质录音和播放没有太大关系。这是一个相当标准的拓扑结构，来自“黄金时代”的相对较大的模拟 VU 表，消费者中包含一些有趣的工作方案；这里值得探讨的历史特性。

It is clear at a first glance that the circuit does not use full-wave rectification, saving a total of three then even more expensive germanium parts per channel. Instead D1, a lone 1N34 germanium diode is used to provide rectification of the audio output, which is delivered to the meter through R46, with electrolytic capacitor C37 providing electrical damping to the meter which would otherwise probably exhibit overshoot, being manufactured to a cost and not a specification. At first glance it is observable that the rectifier will generate a negative DC current with respect to ground. The meter movement is therefore connected with its positive terminal to ground. There is a good reason for choosing seemingly strange positioning as the driving amplifier more able to sink current through the collector of TR8, than source it effectively through R44, thus a negative current must flow through the meter with respect to ground, and D1 points into the driver amplifier, rather than away from it.
从第一眼就能清楚地看出，该电路没有使用全波整流，每个通道节省了总共三个甚至更多的当时甚至更昂贵的锗元件。相反，使用 D1，一个单独的 1N34 锗二极管，来提供音频输出的整流，该输出通过 R46 传递到仪表，电解电容器 C37 为仪表提供电气阻尼，否则仪表可能会出现超调，因为它是按成本而不是按规格制造的。乍一看，可以观察到整流器将相对于地产生负直流电流。因此，仪表的动端连接到地，有很好的理由选择看似奇怪的定位，因为驱动放大器更擅长通过 TR8 的集电极吸收电流，而不是通过 R44 有效地提供电流，因此必须通过仪表相对于地产生负电流，而 D1 指向驱动放大器，而不是远离它。

As the meter is AC coupled through the same path as the line output, via C30, an equivalent positive DC current must flow somewhere from the negative end of C30 to ground. At first it would seem that this would have to generate a quite substantial positive offset on the line output, but another trick is employed here to reduce this ill-effect to an acceptable level. The headphone output transformer T1, necessary to couple the once ubiquitous 8Ω low impedance headphone output to the line output amplifier, allows a solution to be found. The windings on the high impedance side, connected across the same path as the VU meter rectifier circuitry, present a low enough DC resistance of around 200Ω, to prevent DC offsets and transients high enough to cause serious upset from accumulating across the line output. If this component was not present, then the VU meter rectifier would have to coupled through a separate decoupling network with a fairly low value resistor on the lower arm, passing on a much increased load to the already rather stressed line-headphone-meter amplifier.
由于电表通过 C30 与线路输出相同路径的交流耦合，必须从 C30 的负端流向地面的等效正直流电流。起初这似乎会在线路输出上产生相当大的正偏移，但这里采用了另一种技巧来将这种不良影响降低到可接受的水平。耳机输出变压器 T1，用于将曾经普遍的 8Ω低阻抗耳机输出耦合到线路输出放大器，找到了一种解决方案。高阻抗侧的绕组，连接在 VU 表整流电路相同的路径上，提供了足够低的直流电阻，大约 200Ω，以防止直流偏移和瞬态积累在线路输出上，从而引起严重的干扰。如果这个组件不存在，那么 VU 表整流器将不得不通过一个具有相当低值电阻的单独去耦网络进行耦合，这将给已经相当紧张的线路-耳机-表放大器传递更大的负载。

The driving amplifier itself somewhat satisfies the requirement of having a low output impedance, as it uses a high degree of negative feedback via VR2. The classic double NPN configuration that found its way in to most audio circuitry of this era is used, with negative feedback being delivered to the emitter of TR7 through C32. The output impedance is determined by capacitor C30 as frequency decreases, and hence the distortion generated by the rectifier, will rise quiet significantly as frequency reduces, easily degrading the distortion to over 1% or so approaching 50Hz. The spec for the whole unit states that distortion is 1.5% at full level, so it is most likely that the distortion limitations of the tape itself would have at least masked the worst of the meter distortion in the more crucial mid-range. This low frequency distortion could be mitigated by increasing the value of C30, which is already a factor of five too small to modern eyes, but would bring a cost penalty in addition to potential problems with greater turn on transients. When the unit is powered up, the positive end shoots up to approximately half the supply voltage; more capacitance means more current flow in this condition. Adding an additional muting circuit and output relay would have increased costs to an unacceptable level, and usually single supply circuits like this one were carefully designed so as to generate an acceptably low level of transient noise at switch-on, typically by adjusting various RC time constants upstream of the output through trial and error in the lab.
驱动放大器本身在一定程度上满足了低输出阻抗的要求，因为它通过 VR2 使用了高程度的负反馈。这个时代大多数音频电路中使用的经典双 NPN 配置被采用，负反馈通过 C32 传递到 TR7 的发射极。随着频率降低，输出阻抗由电容器 C30 决定，因此整流器产生的失真将显著增加，当频率降低到接近 50Hz 时，失真很容易降至 1%以上。整个单元的规格表明，在满量程时失真为 1.5%，因此，磁带的失真限制至少可以掩盖在中频更关键区域的最严重仪表失真。可以通过增加 C30 的值来减轻低频失真，这对于现代来说是太小了，但会带来成本惩罚，并可能引起更大的开启瞬态问题。当单元通电时，正极会迅速上升到约一半的供电电压；更多的电容器意味着在这种条件下有更多的电流流动。增加一个额外的静音电路和输出继电器将使成本增加到不可接受的水平，而且通常像这样的单电源电路被精心设计，以便在开关时产生可接受的低水平瞬态噪声，通常是通过在实验室中通过试验和错误调整输出上游的各种 RC 时间常数来实现。

Lastly, the line output itself is fed through an L-pad attenuation network consisting of R13 and R10, reducing the level by some 8.5dB and adding 5.66kΩ of output impedance onto the line output, which is awfully high in retrospect. The high resistor values were most likely chosen to avoid putting another heavy load on the busy line amplifier. It’s most likely that the designers were aiming for a 10dB of reduction in level when the unit was connected to the input of the average integrated amplifier of the day which would be expecting an input level of 100mV or so, again very low in comparison to the offerings of today. The reason for this quite severe reduction in level is twofold. Firstly it allows the line amplifier to develop enough voltage at its output to sufficiently drive the meter, which is already only half as sensitive as it might otherwise be due to the half wave rectification, and break through the forward voltage of D1. Secondly, the line amplifier must also drive the headphone transformer with enough level to generate an adequately ear-splitting sound in the headphones when they are plugged in, the expectation being that they will be pressed into service when otherwise antisocial amounts of noise are required.
最后，输出线路本身通过由 R13 和 R10 组成的 L 型衰减网络，降低了大约 8.5dB 的音量，并将 5.66kΩ的输出阻抗添加到线路输出中，这在事后看来相当高。高电阻值很可能是为了避免给繁忙的线路放大器增加另一个重负载。设计师们很可能是希望当该设备连接到当时平均的集成放大器输入端时，降低 10dB 的音量，因为当时的输入级别大约为 100mV，与今天的提供相比非常低。这种相当严重的音量降低有两个原因。首先，它允许线路放大器在其输出端产生足够的电压，以充分驱动仪表，该仪表的灵敏度已经是其他可能的一半，因为半波整流，并突破 D1 的正向电压。其次，线路放大器还必须以足够的音量驱动耳机变压器，以便在耳机插入时产生足够震撼的音效，预期它们将在需要其他反社会的噪声量时投入使用。

Although Figure 5 is quite unusual in the amount of function it derives from a seemingly limited number of components, this sort of topological trickery was quite common, with many different needs being served with the meter itself taking second place as somewhat of an after-thought. In most cases, these meters gave a less than stellar performance with a noticeably slow response time, and movement that seemed to be following some form of counterpoint of its own volition to the programme material itself, but were good enough to set recording level in low cost tape recorders with already uninspiring specifications. For this reason, such circuits are best repudiated to be left in the past, and not copied into the future, where components of greater flexibility such as op-amps are ready to step up to task and yield far more satisfying results.
尽管图 5 从看似有限数量的组件中衍生出大量功能，这种拓扑技巧相当常见，许多不同的需求都由电表本身来满足，而电表本身则更像是一种事后考虑。在大多数情况下，这些电表的表现并不出色，响应时间明显缓慢，动作似乎在自发地遵循某种与节目材料本身相对应的对比，但足以在具有不吸引人规格的低成本磁带录音机中设置录音电平。因此，这类电路最好被否定，留在过去，而不是复制到未来，因为具有更大灵活性的组件，如运算放大器，已经准备好承担任务并产生更令人满意的结果。


Improved VU meter driver circuit
改进的 VU 表驱动电路
The circuits described thus far have shown themselves to exhibit two major problems, that should be addressed if an excellent performance is to be obtained without using expensive parts. The first is the problem of drive, with the non-linear effect of the rectifier potentially harming overall linearity, either directly through loading, or indirectly via ground current. Secondly there is the issue of diode forward voltage, which degrades the low end sensitivity of the meter, producing a dead zone for peak signal levels 20dB or so below full scale.
到目前为止所描述的电路已经显示出两个主要问题，如果要在不使用昂贵部件的情况下获得出色的性能，这些问题应该得到解决。第一个问题是驱动问题，整流器的非线性效应可能会损害整体线性度，无论是通过负载直接还是通过地电流间接。其次，还有二极管正向电压的问题，这会降低仪表的低端灵敏度，产生一个峰值信号水平约为满量程 20dB 的死区。


Non-standard VU meter
Figure 6. 500μA VU meter seen on many online marketplaces
图 6. 在许多在线市场上看到的 500μA VU 表

The improved drive circuitry described in this section was originally designed for use with Figure 6, a comparatively current-hungry low cost meter, mass produced in Taiwan, and made available on several direct selling online sites. It's certainly not a standard meter type, but works very well as its diminutive size, and therefore low mass moving assembly, is pleasingly sprightly and responsive when driven correctly, in stark contrast with the aforementioned historic example.
本节所述的改进驱动电路最初是为与图 6 配合使用而设计的，图 6 是一种相对耗电量大但成本低廉的仪表，在台湾大量生产，并在多个直销在线网站上销售。这当然不是一种标准仪表类型，但因其小巧的尺寸和低质量移动部件，在正确驱动时表现得非常活泼和灵敏，与上述历史例子形成鲜明对比。

By calling in some active circuitry in an appropriate form, we can do away with all of these infelicities, with a single stage design, while also adding some extra functionality that will go a long way in the real world. The active device of choice here will be the TL072 op-amp, although its sister devices such as the TL052, TL062, and TL082 will suffice just as well. Although this device is much maligned for its inferior noise and distortion performance, high-level common mode misbehaviour, and load driving ability, none of these weaknesses are of particular consequence for the following application. This low cost and readily available part has been available for some 40 years now and can be reasonably foreseen to be still available in the next 40 years. This is much unlike the very latest-and-greatest audio op-amps that seem to explode onto the market every year, only to implode several years later.
通过调用适当形式的某些有源电路，我们可以消除所有这些不便，同时使用单级设计，并在实际应用中增加一些额外的功能。这里选择的有源器件将是 TL072 运算放大器，尽管其姐妹产品如 TL052、TL062 和 TL082 也完全可以满足需求。尽管这个器件因其噪声和失真性能较差、高电平共模行为不良和负载驱动能力差而受到很多批评，但在以下应用中，这些弱点并不特别重要。这个低成本且易于获得的部件已经存在了大约 40 年，可以合理预见在未来 40 年内仍然可以获得。这与每年都会出现，几年后又突然消失的最新音频运算放大器大不相同。

The TL072 has high impedance JFET inputs that will not draw any significant input bias currents compared to BJT input devices, permitting a complementary high impedance input coupling network that will present a most leisurable loading to whatever audio source we might chose to connect it with. The corresponding lack of appreciable input current noise due to the JFET input will also prevent any possibility of noise being induced onto a high impedance source. A slew rate of 20V/μs is also most suitable for purpose, fast enough to quickly punch through the rectifier when required, but not so high as to potentially cause problems with stability in a conventional layout, or a one-off type construction of strip-board. Finally, the TL072 and her sisters of minor variation issue attractively low demands when it comes to quiescent power - at a maximum of 2.5mA per amplifier, typically 1.4mA, along with a minimum power supply rejection ratio of 70dB, typically 100dB, which allows a very simple and low cost power supply to be used, as will be described further on.
TL072 具有高阻抗 JFET 输入，与 BJT 输入器件相比，不会产生任何显著的输入偏置电流，允许使用互补的高阻抗输入耦合网络，为可能连接的任何音频源提供最舒适的负载。由于 JFET 输入导致的可感知输入电流噪声的相应缺乏，也将防止任何噪声被感应到高阻抗源上。20V/μs 的转换速率也非常适合此用途，足够快，可以在需要时迅速穿透整流器，但又不至于太高，可能会在传统布局或一次性条形板构造中引起稳定性问题。最后，TL072 及其略有变化的姐妹产品在静态功耗方面要求很低——每个放大器的最大功耗为 2.5mA，通常为 1.4mA，同时具有 70dB 的最小电源抑制比，通常为 100dB，这使得可以使用非常简单且成本低的电源，这一点将在后面进一步描述。


Improved VU meter driver circuit
Figure 7. Improved active VU meter driver circuit
图 7. 改进的 VU 表驱动电路

Figure 7 shows the improved driver circuit using our op-amp of choice with, a split supply arrangement. The power rails are not shown here for simplicity, but a 100nF capacitor must be connected across the op-amps power pins to avoid any potential instability caused by stray inductance. It is not a complicated circuit, consisting of 1 potentiometer, 5 resistors, 3 capacitors and of course the op-amp and VU meter itself. It is assumed that 2 identical channels will be built out of each of the two individual op-amps contained within the TL072 IC.
图 7 展示了使用我们选择的运算放大器的改进型驱动电路，采用分立电源配置。出于简洁考虑，此处未显示电源轨，但必须在运算放大器的电源引脚之间连接一个 100nF 电容器，以避免由杂散电感引起的任何潜在的不稳定性。这不是一个复杂的电路，由 1 个电位器、5 个电阻、3 个电容器以及当然的运算放大器和 VU 表本身组成。假设将从 TL072 集成电路中包含的两个单独运算放大器中构建出 2 个相同的通道。

The audio input is referenced to the circuit ground and can range from 235mV RMS, to more than 10V RMS, with the sensitivity of the circuit being continuously adjustable between these two points via potentiometer R1, which may be a small set-and-forget preset type calibrated to any level within this range by the constructor, or possibly a panel mounted device to allow external control. The input impedance seen at the audio input by the source will be a very undemanding load of 91kΩ at the most sensitive setting and approach 100kΩ as the sensitivity is decreased to a point of total deafness when the potentiometer wiper touches ground. This sort of high impedance is suitable for practically any line level audio source that can be imagined. A potentiometer with a value of less than 100kΩ, may be used with 10kΩ being the lower limit for most line level sources, although some vintage sources such as that in Figure 5 will perceptibly remonstrate by exhibiting significant loss in level and possibly increased non-linearity. Therefore, the lowest advisable value to be used in the estimation of the author should be 20kΩ as this would still allow additional loading from another input, from say a pre-amplifier, to be connected to the source without soliciting any undue grief.
音频输入参考电路地，范围从 235mV RMS 到超过 10V RMS，电路灵敏度可通过电位器 R1 在这两点之间连续调整，该电位器可能是一个小型设置并忘记的预置类型，由制造商校准到该范围内的任何水平，或者可能是一个面板式设备，以允许外部控制。源在音频输入处看到的输入阻抗在最灵敏的设置下将是一个非常不苛刻的负载，即 91kΩ，当灵敏度降低到电位器滑片接触地面的完全失聪点时，接近 100kΩ。这种高阻抗适用于几乎可以想象到的任何线路级音频源。值小于 100kΩ的电位器可以用于大多数线路级源，尽管一些复古源，如图 5 中的源，可能会通过显示明显的电平损失和可能增加的非线性来明显地反对。因此，作者建议使用的最低值应为 20kΩ，因为这仍然允许从另一个输入，例如前置放大器，连接到源，而不会引起任何不适当的麻烦。

The level adjusted source at the potentiometer wiper then finds itself DC decoupled through C1 and onto R2. As already mentioned, the TL072 has a high impedance JFET input that draws practically zero input bias current, and can be relied on not to generate a significant DC offset across the 1MΩ of resistance that R2 places at DC between the non-inverting input and ground. This high value allows C1 to be made correspondingly small, at 100nF - usefully the same value as the not visible op-amp supply rail decoupling capacitor, without the risk of premature low frequency roll-off, which would produce an insensitivity to lower frequencies still within the range of hearing. This lower value necessitates a non polar capacitor with a polyester or possibly ceramic dielectric, resulting in an immunity to quite diabolical levels of DC offset on the audio input, greater than even the worst expected from the most ill conceived of equipment put into possession of the leakiest electrolytic output coupling capacitors. R3 performs the function of a crude RF filter in connection with the JFET input capacitance of the TL072, helpfully isolates this slightly non-linear capacitance from the source at maximum sensitivity, and finally affords robust input over voltage protection in tandem with the input clamp located inside the TL072. In high fidelity audio design, a series resistance greater than 2kΩ on the input of a TL072 is to be avoided due to the increased non-linear effect of the input capacitance, which varies with voltage, but in this instance it is of no consequence as it does not skew the average rectified measurement by any real degree.
电位器滑片处的调整级源随后通过 C1 直流耦合到 R2。如前所述，TL072 具有高阻抗 JFET 输入，几乎不吸收输入偏置电流，并且可以信赖它不会在 R2 在直流下放置的非反相输入和地之间产生显著的直流偏移，R2 的电阻为 1MΩ。这个高值使得 C1 可以相应地做得很小，为 100nF - 有用的是与不可见的运算放大器电源轨去耦电容相同的值，而不会出现低频过早衰减的风险，这会产生对听频范围内较低频率的敏感性。这个较低值需要非极性电容器，具有聚酯或可能是陶瓷介电体，从而对音频输入的相当恶毒的直流偏移水平具有免疫力，甚至超过最糟糕的预期，即使是最不切实际的设备也拥有最漏的电解质输出耦合电容器。R3 在 TL072 的 JFET 输入电容连接中执行粗略射频滤波器的功能，有助于将这种略微非线性的电容从最大灵敏度的源隔离，并最终与 TL072 内部的位置输入钳位一起提供坚固的输入过压保护。在高保真音频设计中，由于输入电容的非线性效应增加，输入电容随电压变化，因此应避免在 TL072 的输入上使用大于 2kΩ的串联电阻，但在这个例子中，它没有实际影响，因为它不会以任何实际程度歪曲平均整流测量。

The op-amp U1, then follows the AC voltage present on the non-inverting input onto R4, through the AC side of a full-wave meter current rectifier consisting of D1, D2, D3 and D4. Connected in parallel with the rectifier is C2, whose purpose is to make certain that there is no high frequency instability or an excessively non-linear high frequency reference received across R4 as the op-amps open loop gain slowly drops off at high frequency, its value being carefully chosen so as not to allow so much high frequency bypass that the sensitivity of the meter is curtailed at the top of the audio band. The rectifier diodes are all inexpensive and easily acquired 1N4148 silicon types and with a heretofore undesirable forward voltage of 0.65V. This is no problem as, the rectifier is driven by what is essentially a constant current source, which makes the forward voltage irrelevant. The op-amp will almost instantly punch through the combined rectifier forward voltage near instantaneously to deliver the current needed to replicate its input voltage across R4. The meter can now respond to low level signals, far below the threshold voltage of the rectifier diodes in the ANSI circuit of Figure 4, while simultaneously doing away with the expensive and fragile germanium diodes. Only the op-amps' DC offset threatens low end sensitivity, typically a factor of 100 less than the combined germanium diode threshold voltage in the ANSI circuit does.
运算放大器 U1 随后将非反相输入上的交流电压加到 R4 上，通过由 D1、D2、D3 和 D4 组成的全波整流器的交流侧。与整流器并联的是 C2，其目的是确保当运算放大器的开环增益在高频下逐渐降低时，R4 上不会接收到高频不稳定或过于非线性的高频参考电压，其值被仔细选择，以免允许过多的高频旁路，从而在音频频段顶部降低仪表的灵敏度。整流二极管都是价格低廉且易于获得的 1N4148 硅型，具有 0.65V 的前向电压。这没有问题，因为整流器是由一个本质上恒定电流源驱动的，这使得前向电压无关紧要。运算放大器几乎瞬间击穿组合整流器的前向电压，以在 R4 上提供复制其输入电压所需的电流。仪表现在可以响应低电平信号，远低于 ANSI 电路图 4 中整流二极管的阈值电压，同时消除昂贵且易碎的锗二极管。只有运算放大器的直流偏移威胁到低端灵敏度，通常比 ANSI 电路中组合锗二极管阈值电压低 100 倍。

The driver circuit's ground current is now perfectly clean, a linear function of input voltage, as should be the case when the input voltage is replicated across R4, a device linear by definition, by U1. When 235mV is applied across R4, 500μA flows through the meter, pulling the needle to the maximum end of the scale. As the common mode voltage on the op-amps inputs will not exceed more than about 1V peak during normal use, undesirable common mode foibles of the TL072, such as the non-linear input capacitance are not able to rear their heads above the parapet of negligibility to get a few painful shots in. One particularly aggravating effect that comes into play as the common mode voltage approaches the negative supply rail of a TL072, is that the output swings up all the way to the positive supply rail - a truly awful sound to those who have heard it.
驱动电路的地电流现在非常干净，是输入电压的线性函数，正如当输入电压通过 R4 复制时应该的那样，R4 是一个定义上的线性设备，由 U1 实现。当在 R4 上施加 235mV 时，500μA 的电流通过仪表，将指针拉到刻度的最大端。由于在正常使用中，运算放大器输入的共模电压不会超过约 1V 峰值，因此 TL072 的不希望共模缺陷，如非线性输入电容，无法在可忽略不计的壁垒之上抬头，以获得一些痛苦的打击。当共模电压接近 TL072 的负电源轨时，一个特别令人烦恼的效果是输出完全摆动到正电源轨——对于那些听过的人来说，这是一种真正可怕的声音。

A few more components are necessary to make the circuit behave itself outside of the theoretical. Resistors R5 and R6 give the op-amp output enough isolation from any stray capacitance that might exist in the wiring between the meter and the drive circuit, to prevent stability margins from evaporating into acrid fumes of oscillation. These resistors also help to limit the distorted high frequency transients from the rectifier, whose current is now undistorted, but whose differential voltage output most certainly is not, that might undesirably capacitively couple themselves onto any nearby cabling that may or may not be carrying an, up till then uncontaminated, audio signal. It is highly recommended that if more than a few centimetres of wiring is to be seen in place between the meter and the drive circuit, that it be in the form of a twisted pair, which will almost entirely eliminate this possibility. Under no circumstances should the negative end of the meter by connected to ground. Finally C3 is included to damp the movement of the low cost meter so that it does not overshoot and give stronger reading than is warranted by the programme material.
一些额外的组件是必要的，以使电路在理论之外表现良好。电阻 R5 和 R6 为运算放大器输出提供了足够的隔离，以防止任何可能存在于仪表和驱动电路之间布线中的杂散电容，从而防止稳定性余量蒸发成振荡的刺鼻烟雾。这些电阻还有助于限制整流器产生的失真高频瞬态，其电流现在是无失真的，但其差分电压输出绝对不是，可能会不希望地电容耦合到任何可能或可能不携带之前未受污染的音频信号的附近电缆。强烈建议，如果仪表和驱动电路之间可以看到超过几厘米的布线，那么它应该是双绞线形式，这几乎可以完全消除这种可能性。在任何情况下，都不应将仪表的负端连接到地。最后，C3 被包括在内，以抑制低成本仪表的运动，使其不会超出极限并给出比程序材料所保证的更强的读数。

The schematic in Figure 7 is designed to run on a split supply of anywhere from ±6V to ±17V, close to a factor of three - it is not fussy at all. A split supply arrangement is almost always far better than a single supply one where the op-amp and active circuitry is powered through a positive rail, with the negative supply currents, inevitably far less than linear for all but the most carefully designed discrete class A circuitry, flowing through the same ground as the audio. Unless more diligence than the author has repeatedly observed out in the wild is taken, these non-linear currents will stimulate corresponding non-linear voltages in the parasitic resistance of the ground path, when then aggregate themselves with the otherwise undistorted audio signal. All of a sudden distortion is a factor of 10+ higher than it should be. One particular small manufacturer here in Britain, who shall remain anonymous is a perfect example of this ill effect in practice. Several of their units exhibit rated THD up to 2000ppm. The author can attest through experience that op-amps contained in said units are capable of at least 10 times greater linearity with a more sensible supply topology. Said manufacturer obdurately refuses to accept that there is anything at all amiss, complaining vociferously through a series of convoluted and almost indecipherable statements to all who will listen, that he is maliciously persecuted by anyone who mentions this to anyone else, in any place, and at any time - such statements being comparable to mortal wounding!
图 7 中的电路图设计用于在±6V 至±17V 的分裂电源上运行，接近三倍因素——它一点也不挑剔。分裂电源配置几乎总是比单电源配置要好，在单电源配置中，运算放大器和有源电路通过正电源供电，而负电源电流，不可避免地远小于所有精心设计的离散类 A 电路的线性电流，流经与音频相同的接地路径。除非采取比作者在野外反复观察到的更勤奋的努力，否则这些非线性电流将在接地路径的寄生电阻中激发相应的非线性电压，然后与原本未受干扰的音频信号混合在一起。突然之间，失真度比应有的高 10 倍以上。英国这里的一个特别的小型制造商，我们将保持匿名，是这种不良影响的完美实践例子。他们的一些单元的额定 THD 高达 2000ppm。作者可以通过经验证明，这些单元中的运算放大器至少具有 10 倍以上的线性度，如果采用更合理的电源拓扑。该制造商固执地拒绝接受有任何问题，通过一系列复杂且几乎无法辨认的声明大声抱怨，声称他受到任何提及此事的人的恶意迫害，无论在何处，何时——这些声明相当于致命的伤害！


Single supply VU meter driver circuit
Figure 8. Single supply VU meter driver circuit
图 8. 单电源 VU 表驱动电路

Having heaped an elephantine pile of admonishment onto the single supply method, it is worth saying that there are some cases where only a single supply is available and it is most infeasible to even attempt deriving a split supply to power the circuitry of Figure 7. For this reason, the circuit shown in Figure 8 is offered to demonstrate how it can be done. Using a split supply requires the inclusion of an extra two resistors and capacitors. R7 and R8 generate half the supply voltage as bias, with C5 AC decoupling any supply noise or ripple that might be perched restlessly on the positive supply rail. It is expedient to omit R7, R8 and C5 from the second channel, if two channels are being built, and instead derive the bias from the positive end of C5 from the first. As C5 charges when power is applied, the bias level lifts up slowly for the benefit of the meter when power is applied, rather than savagely assaulting it with half the supply voltage. C4 AC couples the bottom end of R4 to ground: it cannot be directly connected as the full bias voltage, half the supply rail, would then straddle the rectifier and meter assembly, sending the needle shooting off to the right and potentially damaging the coil.
堆砌了大量的告诫于单一供电方法之上，值得指出的是，在某些情况下，只有单一供电可用，甚至尝试为图 7 的电路提供分离供电都几乎是不可能的。因此，图 8 中所示的电路被提供以展示如何实现。使用分离供电需要额外包含两个电阻和两个电容器。R7 和 R8 产生一半的供电电压作为偏置，C5 通过交流耦合去除可能静立在不稳定的正供电轨上的任何供电噪声或纹波。如果正在构建两个通道，则可以省略 R7、R8 和 C5，并从 C5 的正端获取偏置，而不是从第一个。当 C5 充电时，偏置电平会缓慢上升，这对仪表施加电源时是有益的，而不是用一半的供电电压猛烈攻击它。C4 通过交流耦合将 R4 的底部连接到地：不能直接连接，因为完整的偏置电压，即供电轨的一半，将会跨越整流器和仪表组件，使指针向右飞射，并可能损坏线圈。

The supply voltage can be anywhere between +12V and +30V and all electrolytic capacitors should be rated at or above the supply voltage. No more than 10mA of current will be drawn during normal operation, and the circuit is very tolerant of supply rail noise. A volt or maybe more of ripple, noise, or any other unpleasantness on the supply rail will not affect performance. The high degree of smoothing effected by C5, in connection with the high resistance of R2 affirms that no power supply noise can couple onto the audio source and inject electrical detritus. As with Figure 7, an op-amp supply rail capacitor of 100nF must be connected across the power pins, or instability will eventuate. If it is not evident at this point, the op-amp positive supply pin should be connected to V+ and the negative pin connected to ground (0V).
供电电压可在+12V 至+30V 之间，所有电解电容器应标称值不低于供电电压。正常工作期间，电流不会超过 10mA，电路对供电轨噪声非常容忍。供电轨上的电压波动、噪声或其他不愉快现象不会影响性能。C5 实现的平滑程度，结合 R2 的高电阻，确保没有电源噪声可以耦合到音频源并注入电气碎片。与图 7 一样，必须在电源引脚之间连接一个 100nF 的运算放大器供电轨电容器，否则将导致不稳定。如果此时还不明显，运算放大器正电源引脚应连接到 V+，负引脚连接到地（0V）。

*The value of R4 was calculated to admit a rated meter current of 500μA, +25% for meter calibration, in conjuction with a consumer line level voltage of 300mV (-10dBV). If a meter requiring less than 500μA for a full reading is to be used, then the value of R4 can be increased in a manner inversely proportional to the rated meter current relative to 500μA. It is important that this is done, as it will prevent op-amp offset voltage from deleteriously affecting the low end sensitivity of more sensitive meters. For example; a meter with a rated current of 200μA, 2.5 times smaller, would warrant changing R4 to 1.2kΩ. A 100μA meter, would be best suited to an R4 value of 2.2kΩ, a 250μA part 1kΩ, and so on and so forth...
R4 的值计算为允许额定表电流为 500μA，表校准+25%，与 300mV（-10dBV）的消费者线路电压相结合。如果需要使用小于 500μA 才能完全读数的表，则 R4 的值可以按与 500μA 相对的额定表电流成反比的方式增加。这样做很重要，因为它将防止运算放大器偏移电压对更灵敏表的低端灵敏度产生有害影响。例如；额定电流为 200μA 的表，比 500μA 小 2.5 倍，则应将 R4 改为 1.2kΩ。100μA 的表最适合 2.2kΩ的 R4 值，250μA 的部件 1kΩ，等等。


Adding peak detection  添加峰值检测
The VU meter's unofficial appellation of ‘virtually useless meter’ is not entirely unearned. While VU meters do give excellent indications of general loudness and average levels, they are quite unable to reveal anything more than an implied approximation of peak level, which may be significantly higher than the average level might imply either due to transient peaks or highly asymmetrical waveforms such as human speech. Consequently it is quite useful to include a peak indicator in the form of an LED that will illuminate once the peak level reaches 3dB over 0VU; the same peak level present on a sine wave at 0VU. Since publishing the article, various discussions on adding peak detection to Figure 7 have not led to any satisfactory conclusion. Being presented with the LED output peak detection circuits available elsewhere on the web, for use as add-ons to the circuitry described in the article, that either offer very crude unipolar detection (only sensitive to the AC waveform in one direction), insensitivity to brief peaks, non-linear loading of the line that promises severe distortion, or complete atrocities that just won’t work at all, the author has taken it upon himself to contrive a complete VU meter drive circuit with high quality bipolar peak detection.
VU 表的非官方名称“几乎无用的仪表”并非毫无根据。虽然 VU 表确实能给出很好的总体响度和平均电平指示，但它们完全无法揭示超过峰值电平的任何信息，而这个峰值电平可能比平均电平所暗示的要高得多，这可能是由于瞬态峰值或高度非对称的波形，如人声。因此，在 LED 峰值指示器中包含一个峰值指示器非常有用，一旦峰值电平达到 0VU 以上的 3dB，它就会发光；与 0VU 的正弦波上的相同峰值电平。自发表文章以来，关于在图 7 中添加峰值检测的各种讨论并没有得出任何令人满意的结论。面对网上提供的 LED 输出峰值检测电路，这些电路要么提供非常粗糙的单极检测（仅对单向的交流波形敏感），要么对短暂的峰值不敏感，要么对线路的非线性加载承诺严重失真，或者根本无法工作，作者便自己设计了一个完整的 VU 表驱动电路，具有高质量的 bipolar 峰值检测。


VU meter circuit with bipolar peak LED
Figure 9. VU meter driver circuit with bipolar peak detection
图 9. 双极峰值检测 VU 表驱动电路

Figure 9 exhibits the product of an evening in the workshop languorously fiddling with some common parts on the test bench. The topology is a necessary departure from the current driven bridge rectifier arrangement of Figure 7 as a consequence of the need for a rectified DC voltage relative to ground for the purposes of peak detection. Audio comes in through coupling capacitor C1 and is full-wave precision rectified to a positive DC voltage by U1 and U2, whose gain is adjustable via R5. U2 drives the VU meter through R6 and calibration trimmer R7, with C3 performing a little smoothing and meter damping. NPN transistor Q1 performs peak detection via its base threshold voltage on the rectified output of U2 through current limiting resistor R8, switching on PNP transistor Q2 once the base voltage exceeds 550mV or so. Q2 drives the peak indicator LED L1 through current limiting resistor R12, and also applies positive feedback towards Q1 through C4 and R11, so that the detector switches hard on for long enough to be visible once the peak threshold is met.
图 9 展示了在车间度过的一个晚上，悠闲地摆弄测试台上的某些常用部件的产品。由于需要相对于地面的整流直流电压来进行峰值检测，因此拓扑结构是必要的，与图 7 中所示的当前驱动桥式整流器排列有所不同。音频通过耦合电容器 C1 进入，并由 U1 和 U2 全波精确整流为正直流电压，其增益可通过 R5 调整。U2 通过 R6 和校准微调器 R7 驱动 VU 表，C3 进行一些平滑和表头阻尼。NPN 晶体管 Q1 通过 U2 整流输出的基极阈值电压进行峰值检测，一旦基极电压超过约 550mV，就切换 PNP 晶体管 Q2。Q2 通过限流电阻 R12 驱动峰值指示 LED L1，并通过 C4 和 R11 向 Q1 提供正反馈，使得检测器在达到峰值阈值时足够硬地切换，以便可见。

The precision rectifier is a fairly standard arrangement of the kind that is frequently shown as an example of its kind on electronics websites and text books. A slight quirk is that the values of the input resistor of the half-wave rectifier stage, R1, and its feed into the summing amplifier, R4, are selected so as to double the half-wave rectifier's current into the summing stage relative to the non-rectified input, through R3, by making the product of their values double to the square of summing input resistor R3. This is a little different to having to select E24 values to double the current, or pairing up resistors to halve or double their values, either by series or parallel arrangements, and is a useful little trick that makes the circuit a little more ergonomic to build. As R1 and R3 feed into virtual earth points, the input impedance is set by R1 and R3 in parallel to present a fairly innocuous 19.4kΩ to the whatever the metering circuit is connected, which should be unburden-some enough for most unbalanced line circuitry, and no problem at all for the balanced receiver to be described in the next section. The gain of the rectifier is continuously variable from zero to 2.12 which allows a peak input voltage of as low as 260mV to be calibrated to trip the peak detector. C2 ensures that inverting stage U2 is stable at HF while the parasitic capacitance of D1 does the same for U1.
精度整流器是一种相当标准的配置，常在电子网站和教科书中作为其类别的示例。一个轻微的特点是，半波整流级输入电阻 R1 及其馈入求和放大器 R4 的值被选择，以便通过使它们的乘积是求和输入电阻 R3 的平方，将半波整流器的电流加倍进入求和级，相对于未整流的输入，通过 R3。这与选择 E24 值加倍电流，或通过串联或并联配置配对电阻以减半或加倍它们的值略有不同，这是一个有用的技巧，使电路构建更加方便。由于 R1 和 R3 馈入虚拟地，输入阻抗由 R1 和 R3 的并联设置，向连接的计量电路呈现一个相当无害的 19.4kΩ，这对于大多数不平衡线路电路来说应该足够轻松，对于下一节中描述的平衡接收器来说则毫无问题。整流器的增益可以从零连续可调到 2.12，这允许峰值输入电压低至 260mV，以便校准触发峰值检测器。C2 确保反相级 U2 在 HF 时稳定，而 D1 的寄生电容对 U1 做同样的工作。

VU meter drive is quite unremarkable, simply consisting of a variable series resistance in the form of R6 and R7. R6 is included to make the circuit foolproof and prevent the possibility of damage to the meter movement if R7 is shorted out during calibration. If a meter with a current rating deviating from that shown in the schematic is put into service, then R6 and R7 will have to be adjusted to the inverse of the difference. C3 performs damping of the meter movement and shunts away most of the HF nastiness that might couple from the meter movement wiring onto an otherwise distortion free audio connection. Note that a different ground symbol is used to indicate that the ground current will be quite polluted with very high levels of high order harmonics as a result of the full-wave rectification. It is unavoidable with this topology that the ground current will be significantly distorted and therefore the meter ground should be connected much further downstream towards the power supply circuitry, preferably straight to the power supply 0V rail, and not be shared with any audio circuitry. It is otherwise practical to drive VU meters from precision rectifiers as shown, and use a capacitor in the range of 1μF in place of C2 to smooth out the harmonics in the feedback loop of the precision rectifier summing amplifier, and therefore make the ground current clean enough to allow a little more flexibility with layout; doing so here will make the peak detection quite uselessly insensitive to transients.
VU 表驱动相当普通，仅由 R6 和 R7 组成的可变串联电阻。R6 包括在内，使电路成为防呆设计，防止在校准过程中 R7 短路时损坏仪表运动。如果将电流额定值与图中所示不同的仪表投入使用，则 R6 和 R7 必须调整到差异的倒数。C3 对仪表运动进行阻尼，并分流掉可能从仪表运动线路上耦合到其他无失真音频连接的大部分高频恶劣信号。请注意，使用不同的接地符号表示，由于全波整流，接地电流将受到非常高的高阶谐波污染。这种拓扑结构不可避免地导致接地电流严重失真，因此仪表接地应连接到电源电路的下游，最好直接连接到电源 0V 轨，不要与任何音频电路共享。否则，从精度整流器驱动 VU 表是实用的，并使用 1μF 范围内的电容器代替 C2，以平滑精度整流器求和放大器的反馈回路中的谐波，从而使接地电流足够干净，以便在布局上具有更多灵活性；这样做将使峰值检测对瞬态非常不敏感。

Far too many peak indication circuits listed elsewhere on the web only light up the LED for the duration of the peak, which may only exist for 10μ in many cases and will be far to brief to be visible unless the users gaze is permanently fixed on the indicator LED with no natural light present. Some require the peak to charge a capacitor to hold the peak for long enough to be visible, degrading the sensitivity to brief transients that may not exist for long enough to charge the sampling cap. The peak detector circuitry described in Figure 9 uses positive feedback via timing capacitor C4 so that once the base voltage of Q1 has been breeched for even the shortest period of time, the indicator LED will stay on for at least as long as it takes for C4 to charge up through R11 and Q1; about 250ms with the supply voltage of ±15V shown. Clamping diode D4 allows C4 to discharge through the much lower resistance of R12 and the indicator LED so that the circuit is ready for action again within 2ms of either the peak threshold voltage disappearing from the detector's input, or C4 completing its charge cycle; whichever occurs later.
太多峰值指示电路在网络上列出，但它们只会在峰值期间点亮 LED，而在许多情况下，峰值可能只存在 10μ秒，除非用户的目光始终固定在指示 LED 上且没有自然光，否则将无法看到。一些需要峰值充电电容器以保持峰值足够长，以便可见，这会降低对可能不足以充电采样电容的短暂瞬变的灵敏度。图 9 中描述的峰值检测电路通过定时电容器 C4 使用正反馈，因此一旦 Q1 的基极电压被突破，即使是最短的时间，指示 LED 也会保持点亮，直到 C4 通过 R11 和 Q1 充电，大约 250ms，如所示±15V 电源电压。钳位二极管 D4 允许 C4 通过 R12 和指示 LED 的更低电阻放电，从而使电路在峰值阈值电压从检测器输入消失或 C4 完成充电周期后的 2ms 内准备好再次行动； whichever occurs later.

High current gain BC550 and BC560 general purpose transistors are prescribed for Q1 and Q2, but they can be any general purpose small bipolar transistors that the constructor has on hand, the parts specified being those in the closest reach of the author. Diode D3 clamps the input of Q1 to prevent it from possible damage should U2's output assume a negative voltage of more than -5V; rather unlikely and to the author's knowledge only realistic with a severe positive input over-voltage on the line in, or perhaps some curious behaviour when powering up and down, however unlikely; it is still included as it is an inexpensive, easily fitted part that is included for purposes conceivably more in line with peace of mind than practical function. If it can go wrong, it will go wrong, as an old mantra eternally echoes throughout any exercise in engineering. The peak LED shares the same ground as the meter to allude to expediencies of wiring that can be accomplished by using 3 conductors to wire the circuitry to the meter movement and indicator LED that may be some distance away on a panel.
高电流增益 BC550 和 BC560 通用晶体管被指定用于 Q1 和 Q2，但可以使用构建者手头上的任何通用小型双极性晶体管，所指定的部件是作者最易触及的。二极管 D3 夹住 Q1 的输入，以防止 U2 的输出在超过-5V 的负电压下可能造成的损坏；这种情况不太可能，据作者所知，只有在输入线上有严重的正电压过压时才会发生，或者可能在启动和关闭时出现一些奇怪的行为，尽管可能性不大；但它仍然被包括在内，因为它是一个便宜、易于安装的部件，其目的可能是为了安心而不是实际功能。如果可能出错，它就会出错，因为一句古老的格言永恒地回荡在任何工程实践中。峰值 LED 与仪表共享同一接地，以避免使用 3 根导线将电路连接到仪表机构和可能位于面板上较远处的指示 LED 的布线便利性。

To calibrate the circuitry, first set the peak indication threshold via R5 with a sinusoidal tone at the desired maximum level so that the indicator LED is just on the verge of illuminating, then adjust R7 calibrate the meter to read 0VU. Alternatively, the peak threshold can be set just below the point of clipping or overload of the source and R7 adjusted to read 0VU with more normal programme levels, so long as there is not too wide a gulf between these two points. The former method is recommended for general use.
要校准电路，首先通过 R5 设置峰值指示阈值，使用所需的最高电平的正弦波，使指示 LED 刚好处于即将点亮的状态，然后调整 R7 校准仪表读取 0VU。或者，峰值阈值可以设置在削波或过载点稍低的位置，同时调整 R7 以在更正常的节目电平下读取 0VU，只要这两个点之间的差距不要太宽。前一种方法建议用于一般用途。

*A regulated split power supply of ±15V is recommended for use with Figure 9, although it is feasible to reduce this down to ±9V without greatly affecting the operation of the circuit. U1 and U2 will need the usual power supply decoupling capacitor of 100nF across their power supply pins to prevent oscillation. If the power supply voltage is to be reduced then LED current limiting resistor R12 should be reduced accordingly; as shown it admits an LED current of about 12.5mA which is about as much as it is safe to administer to a GaAs indicator LED. If a higher current LED is to be used then it may also be reduced further, but not so much that it pulls more than about 50mA through the collector of Q2. If a low current half-wave split supply is used with a small transformer of less than 3VA, then it may be a good idea to connect the cathode of L1 to the negative supply rail, rather than power ground as shown, and double up the value of R12 to avoid unipolar current woes in the power transformer. C4 can be adjusted to either increase or decrease the persistence time of the indicator to suit the preference of the constructor, from 10nF up to 10μF, although the author believes the value shown is just right, this may be a matter of personal preference. If the meter movement appears either over or under-damped, then C3 can also be adjusted, but the value shown should work perfectly for all but the most fussy or cumbersome meter movements.
建议使用±15V 的稳压分立电源与图 9 配合使用，尽管将其降低到±9V 也不会对电路的运行产生很大影响。U1 和 U2 需要在其电源引脚上使用通常的 100nF 电源去耦电容器，以防止振荡。如果需要降低电源电压，则应相应地降低 LED 限流电阻 R12；如所示，它允许约 12.5mA 的 LED 电流，这是对 GaAs 指示 LED 安全施用的最大电流。如果使用更高电流的 LED，则还可以进一步降低，但不要降低太多，以免使 Q2 的集电极电流超过约 50mA。如果使用低电流半波分立电源和小于 3VA 的小型变压器，则将 L1 的阴极连接到负电源轨，而不是如所示连接到电源地，并将 R12 的值加倍，以避免在电源变压器中产生单极电流问题。C4 可以调整以增加或减少指示器的持续时间，以适应构造者的偏好，从 10nF 到 10μF 不等，尽管作者认为所示值恰到好处，这可能是个人偏好的问题。如果仪表运动看起来是过阻尼或欠阻尼，则还可以调整 C3，但所示值应该对所有但最挑剔或笨拙的仪表运动都适用。


Balanced line receiver  平衡线路接收器
A couple of months after the new article was first published on the site, the author received several e-mails asking how best to use the improved VU meter drive circuit of Figure 7 in conjunction with a balanced line. The answer to the question depends upon what might be connected to the other end of said balanced line and whether or not common mode noise can be tolerated on the VU meters reading. If the stage driving the line has a fully differential output referenced to ground, then it is possible to use the Figure 7 circuit as-is with the input connected to either the hot or cold side of the line. There will be a small penalty on overall common mode rejection if another balanced line input is used on the line, as the slight loading of the metering circuitry will ever so slightly unbalance the common mode impedance of the line (although with modern op-amp based equipment this will be very small) and reduce the effective CMRR. If the line output the meter is monitoring is either impedance balanced and not fully differential or transformer balanced and essentially floating, then all of the above goes out of the window and a balanced line receiver will need to be deployed if anything approaching an accurate reading is to be presented.
几个月后，该新文章首次在网站上发布，作者收到了几封电子邮件询问如何最好地将图 7 中改进的 VU 表驱动电路与平衡线结合使用。问题的答案取决于可能连接到该平衡线另一端的是什么，以及是否可以容忍 VU 表读数上的共模噪声。如果驱动线路的级具有以地线为参考的完全差分输出，那么可以直接使用图 7 电路，输入连接到线路的热端或冷端。如果在线路上使用另一个平衡线输入，整体共模抑制将会有轻微的损失，因为对仪表电路的轻微负载将略微不平衡线路的共模阻抗（尽管在现代基于运算放大器的设备中，这将非常小）并降低有效 CMRR。如果仪表监测的线路输出是阻抗平衡且不是完全差分，或者变压器平衡且基本上浮动的，那么上述所有内容都不适用，如果需要呈现接近准确的读数，则需要部署平衡线接收器。


Balanced line receiver for VU meter circuits
Figure 10. Balanced line receiver for VU meter drivers
图 10. 平衡线路接收器，用于 VU 表驱动器

Figure 10 details a suitable balanced line receiver for the Figure 7 and 9 circuitry that will not only ensure accurate reading of floating or non-differential balanced lines but will also give handsome rejection of common mode noise (although whether common mode noise will be visible on the meter movement is rather debatable for most high quality setups). The balanced receiver shown in Figure 8 is designed specifically for use with level metering circuits such as Figures 7 and 9 and is therefore not intended to be a high quality low noise balanced input. Rather it is optimised to present a high impedance to the line of greater than 200kΩ and to bring a +4dBu professional balanced line level of 1.228V RMS down to a more workable 286mV RMS that the meter drive will want to see. It is possible to put this level into the meter without any prior attenuation, but calibration may be a little more fiddly towards the end of the trimmer potentiometer's range where higher rotational accuracy will be required of the hand on the end of the adjusting screwdriver. The level drop also makes running the receiver off lower supply rails of less than ±15V without the possibility of clipping more feasible. With input and feedback resistor values in the hundreds of kilo ohms so as to keep the loading on the line low enough that additional equipment might be connected to it, the receiver's output is rather noisy at -100.2dBV, made even worse by the fact that the differential gain is -12.7dB to yield a shocking EIN of -87.5dBV, but this is still at least an order of magnitude lower than what will be visible on the meter movement.
图 10 详细说明了适用于图 7 和 9 电路的合适平衡线路接收器，它不仅能确保对浮点或非差分平衡线路的准确读取，还能很好地抑制共模噪声（尽管对于大多数高质量设置，共模噪声是否会在仪表运动中可见是有争议的）。图 8 中所示的平衡接收器专门设计用于与图 7 和 9 的水平仪表电路一起使用，因此并不打算作为高质量低噪声平衡输入。相反，它优化了向线路提供大于 200kΩ的高阻抗，并将+4dBu 专业平衡线路电平 1.228V RMS 降低到更易于工作的 286mV RMS，这是仪表驱动器希望看到的。可以将此电平直接输入仪表，而无需任何预先衰减，但校准可能在调整电位器的范围末端会更加繁琐，此时需要更高的旋转精度。水平下降也使得在没有削波可能性的情况下，从低于±15V 的较低电源轨上运行接收器变得更加可行。由于输入和反馈电阻值在数百千欧姆，以保持线路上的负载足够低，以便可以连接其他设备，因此接收器的输出在-100.2dBV 时相当嘈杂，由于差分增益为-12.7dB，导致惊人的 EIN 为-87.5dBV，但这仍然至少比仪表运动上可见的量低一个数量级。

When feeding the Figure 7 driver circuit from Figure 10, rather than directly from an unbalanced line, it is possible to short out and omit C1 and R3 from Figure 7, as Figure 8 already performs RF rejection and DC decoupling, making these redundant functions in the meter drive circuitry. Figure 7 and 8 may be built from a single TL072 or TL084 op-amp to form a complete meter drive and balanced receiver with a single IC. Stereo versions may also be built with the TL074 and TL084 ICs. It is the opinion of the author that it is well worth the additional complexity of building Figure 8 for a balanced line, as opposed to just optimistically connecting Figure 7 to one side of the line. If the balanced receiver is to drive any circuitry in the audio through path, then Figure 8 is quite unsuitable and a high quality balanced receiver must be used unless very poor noise performance and probably poor linearity is acceptable.
当从图 10 为图 7 驱动电路供电，而不是直接从不平衡线路供电时，可以短路并省略图 7 中的 C1 和 R3，因为图 8 已经执行了射频抑制和直流耦合，使得这些冗余功能在仪表驱动电路中不再需要。图 7 和 8 可以使用单个 TL072 或 TL084 运算放大器构建，形成一个完整的仪表驱动和平衡接收器。也可以使用 TL074 和 TL084 集成电路构建立体声版本。作者认为，为了平衡线路，构建图 8 的额外复杂性是值得的，而不是仅仅乐观地将图 7 连接到线路的一侧。如果平衡接收器要驱动音频通路中的任何电路，那么图 8 非常不合适，必须使用高质量的平衡接收器，除非可以接受非常差的噪声性能和可能较差的线性度。


Split supply for improved circuit
分割供应以提高电路
A complaint intermittently heard when split supply rails are included in a design, is that split supplies are awkward to build, using twice as many components and being unduly complex compared a single supply rail design. It is certainly true that the supply circuit itself is much simpler, maybe using near half as many components, but the additional complexity starts to rack up very quickly in the biasing networks required in the audio path. It is simply not possible to reduce the component count by using a single supply unless the audio path is very simple; a single stage, single channel circuit topology. Revisiting Figures 7 and 8, it becomes clear that an additional 5 components are necessary to make it work.
在设计中包含分供电源轨时，间歇性听到的投诉是分供电源构建起来很麻烦，需要使用两倍数量的组件，并且与单电源轨设计相比过于复杂。确实，电源电路本身要简单得多，可能只需要一半的组件，但音频路径中所需的偏置网络的额外复杂性开始迅速增加。除非音频路径非常简单，即单级单通道电路拓扑，否则根本不可能通过使用单电源来减少组件数量；重新审视图 7 和图 8，可以清楚地看出还需要额外的 5 个组件才能使其工作。


Simple split supply
Figure 11. Simple split supply for Figures 7 and 9
图 11. 图 7 和图 9 的简单分供电源

Figure 11 demonstrates how a practical split supply circuit, that satisfies the requirements of the Figures 7 and 9 can be brought to fruition with little effort and only 8 components that most constructors will already have lying around. A simple half-wave rectifier to the positive and negative rails, via D1 and D2 respectively, derives a positive and negative voltage with respect to ground, which is then smoothed by reservoir capacitors C2 and C3. Additional RC smoothing is effected through R1 and R2, in partnership with C2 and C4 respectively. No regulation is required, and a ripple voltage of less than 100mV on the outputs with a loading from the Figure 7 circuit of up to 10mA was measured. Considering that the TL072 affords upwards of 70dB, typically 100dB, of power supply rail rejection, this is absolutely nothing to worry about. Regulating the supply as it is would prevent it from developing sufficient output voltage with lower voltage AC sources, such as 6.3V valve heater windings, for the Figure 7 circuit to work with due to the drop-out voltage. Additional quiescent current draw, and safety margin over drop-out voltage, inherent to linear regulator arrangements would put further unnecessary limitations on the flexibility of the power supply.
图 11 展示了如何通过少量努力和仅使用大多数构造者已有的 8 个组件，实现一个满足图 7 和 9 要求的实用分供电路。通过 D1 和 D2 分别连接到正负轨，得到相对于地面的正负电压，然后由电容器 C2 和 C3 进行平滑。通过 R1 和 R2 与 C2 和 C4 的配合，实现额外的 RC 平滑。无需调节，输出端测量到的纹波电压小于 100mV，负载来自图 7 电路，最大为 10mA。考虑到 TL072 提供超过 70dB，通常是 100dB 的电源轨抑制，这根本不值得担心。按照这种方式调节电源将防止它从较低的交流电源（如 6.3V 电子管加热线圈）获得足够的输出电压，因为存在压降。线性稳压器配置固有的额外静态电流消耗和压降安全余量将进一步限制电源的灵活性。

The smoothing resistors only need to be quarter-watt types, and the capacitors should be rated for twice the AC input voltage to secure longevity. The AC source can be a small plug-in wall wart, or a small (<5VA) transformer if the constructor is comfortable and competent in AC wiring. If not, then the wall-wart type must be used as it will most certainly be double insulated. The no-load AC voltage should be above 6V RMS, but must not exceed 14V RMS at no load, as this will result in an output voltage possibly high enough to damage the op-amp. Most transformers output voltages are specified at full load, and extra windings are usually added on the secondary to overcome the voltage drop at full load due to winding resistance. With no load present the output voltage may be up to 40% higher than the rated voltage (a rated 12V transformer might put out over 15V!), especially with smaller, less efficient transformers. It is a good idea to measure the no-load voltage of the wall-wart supply or transformer with a multimeter to make sure it does not exceed 14V before bringing it into a potentially frustrating service. If the constructor has free choice of transformer secondary voltage, as opposed to using a pre-existing one already in his possession, the author recommends 9VAC as a safe bet.
平滑电阻只需要四分之一瓦特类型，电容器应额定为交流输入电压的两倍以确保使用寿命。交流电源可以是一个小型插墙式电源适配器，或者如果构造者对交流线路感到舒适且熟练，可以使用小型（<5VA）变压器。如果不适用，则必须使用插墙式电源适配器，因为它肯定具有双重绝缘。空载交流电压应高于 6V 均方根值，但不能超过 14V 均方根值，因为这可能导致输出电压足够高，足以损坏运算放大器。大多数变压器的输出电压是在满载时指定的，通常在次级增加额外的绕组来克服由于绕组电阻引起的满载电压降。在没有负载的情况下，输出电压可能比额定电压高 40%（一个额定 12V 的变压器可能输出超过 15V！），尤其是在较小、效率较低的变压器中。在将电源适配器或变压器带入可能令人沮丧的服务之前，使用万用表测量空载电压是一个好主意，以确保它不超过 14V。如果构造者可以选择变压器次级电压，而不是使用他拥有的现有预置变压器，作者建议 9VAC 是一个安全的选择。

A 6.3V heater winding, found inside most valve amplifiers, as previously alluded to, will also comfortably feed the split supply, making the circuit especially useful to the constructor with a penchant for these devices. The winding must not be connected to the chassis ground by any other route than the ground node in view at the output end of Figure 11, or ground loop trouble may be encountered.
一个 6.3V 加热线圈，如前所述，通常位于大多数电子管放大器内部，也将轻松地为分立电源供电，使电路对喜欢这些设备的构建者特别有用。线圈不得通过图 11 输出端可见的接地节点以外的任何途径连接到机箱接地，否则可能会遇到接地环路问题。

LED back-lights are now quite commonly used for illuminating VU meters made today, being a good deal less efficient than their rapidly disappearing incandescent counterparts, especially if a series connection is made between the back-lights of two or more meters. If LED back-lights are present, they may also be powered by the supply in Figure 11, in series with a suitable current limiting resistor. The back-light must be powered across both of the the ± output rails, as opposed to between one rail and the circuit ground, to make certain that no DC current can flow across the transformer winding, which it will if one half-wave rectifier section is loaded with a higher current than another. If a net DC current flows through the winding of any mains transformer, it may suffer, these parts being ill-disposed to tolerate unipolar DC flowing through their windings.
LED 背光现在很常见地用于照亮现代的 VU 表，其效率远低于快速消失的白炽灯泡，尤其是在两个或更多仪表的背光之间串联时。如果存在 LED 背光，它们也可以由图 11 中的电源供电，串联一个合适的限流电阻。背光必须跨接在两个±输出轨上，而不是在一轨和电路地之间，以确保没有直流电流能够流过变压器绕组，如果半波整流器部分负载的电流比另一个高，就会发生这种情况。如果任何主变压器绕组中流过净直流电流，它可能会受损，因为这些部件不适合承受通过其绕组的单极直流电。

Calculating the LED current limiting resistor, using a a typical 6.3V AC winding, freely found inside the viscera of most valve equipment, resulting in some ±8V available at the power supply outputs with no load. Take, then for instance two meters, with white back-lights of a forward voltage of 3.4V each, with a rated current of 20mA, giving a total of 7.2V when connected in series are to be used with a forward current of 12mA. The author is not well taken through experience to trust the advertised forward current, and prefers to adopt a safer refuge below 70% of this parameter. Having subtracted the 7.2V drop of the LEDs from the combined voltage of 16V between each supply rail to yield a remaining 8.8V to be dissipated at a current of 12mA, a quick R=V/I calculation reveals a series resistance of 733Ω is required in series with the two LEDs to complete the line up. However, the two smoothing resistors must also be subtracted from this value as they are also in series with the LED string. Subtracting the total series resistance of R1 and R2, 200Ω, from the previous value gives 533Ω. This, after subtraction of a further 50Ω to compensate for the additional loading of the VU drive circuit itself, winding resistance, and various other factors that conspire to reduce overall output voltage in unregulated supplies as loading increases, produces 483Ω. Thus, a 470Ω (E12 value) series resistance should be put into use.
计算 LED 限流电阻，使用典型的 6.3V 交流绕组，通常在大多数电子管设备内部自由找到，导致在无负载时电源输出端有±8V 可用。然后，例如使用两个电压为 3.4V 的白色背光表，额定电流为 20mA，串联连接时总电压为 7.2V，应使用正向电流为 12mA。作者并不完全依赖经验来信任广告中的正向电流，而更倾向于采用低于 70%的更安全参数。从每个电源轨之间的 16V 总电压中减去 LED 的 7.2V 压降，得到剩余 8.8V 在 12mA 电流下消耗，快速计算 R=V/I，发现需要串联 733Ω电阻与两个 LED 一起完成线路。然而，两个平滑电阻也必须从该值中减去，因为它们也串联在 LED 字符串中。从 R1 和 R2 的总串联电阻 200Ω中减去前一个值，得到 533Ω。在减去额外的 50Ω以补偿 VU 驱动电路本身的额外负载、绕组电阻以及各种因素，这些因素在负载增加时共同作用以降低未调节电源的总输出电压，得到 483Ω。因此，应使用 470Ω（E12 值）的串联电阻。


Troubleshooting meter damping
故障排除仪表阻尼
Some constructors who have built the circuits have got in touch to report issues with meter movement damping where the meter exhibits gross overshoot and ringing with normal programme material. In all cases, this has been caused by using low quality meter movements of the kind typically found in the sub $5 price range on direct-from-China retailers on eBay, Aliexpress, or Alibaba. While the author recommends using a higher quality meter movement that may cost only a little more, such as those available from Flash Star, it is still possible to use these parts by increasing the value of the meter damping capacitor, shown as 22µF throughout of the course of the article, most often by a factor of 10. Doing so does of course make the meter a little more sluggish, much like those seen on lower quality equipment from the 1970s, but is a good deal better than having a meter than shoots off the scale in both directions given only the gentlest tickling. In one case, a capacitance of 440µF was required to critically damp the movement; a sure sign that the moving mass was at the bottom of the manufacturers list of priorities.
一些构建电路的构造者已经联系报告了仪表运动阻尼问题，其中仪表在正常节目材料下表现出严重的超调和振荡。在所有情况下，这都归因于使用了在 eBay、Aliexpress 或 Alibaba 等中国零售商的$5 以下价格范围内通常可以找到的低质量仪表运动。虽然作者建议使用更高质量的仪表运动，这可能只会稍微贵一点，例如 Flash Star 提供的那些，但仍然可以通过增加仪表阻尼电容的值来使用这些部件，文章中显示为 22µF，通常增加 10 倍。这样做当然会使仪表稍微慢一些，就像 1970 年代低质量设备上看到的那样，但比只有轻微的触碰就会在两个方向上飞出刻度表的仪表要好得多。在一种情况下，需要 440µF 的电容来临界阻尼运动；这是制造商将移动质量放在优先级最低的明确迹象。

It goes without saying that if the time to construct a high quality driver circuit such as those that have been detailed above is going to be taken, then it is well worth spending what may only be the price of a cup of tea in any high street café extra in obtaining a better quality meter movement. As is always the case any audio circuit can only be as good as the transducer that it is connected to, and metering is no exception.
不言而喻，如果花费时间去构建一个高质量的驱动电路，如上述详细描述的那样，那么在获得更好的仪表运动时，花费可能仅相当于任何街头咖啡馆一杯茶的额外费用是值得的。正如通常情况一样，任何音频电路只能与其连接的传感器一样好，仪表测量也不例外。


Conclusion  结论
This article has ended up reading significantly longer than the original, most of it due to additional information that readers who got in contact felt was necessary for them to be made aware of to complete their builds and make suitable power supplies and line interfacing circuitry for said builds. Hopefully the information in this article of of will be of better use to the future constructor as a consequence of this endless updating.
这篇文章的篇幅比原文长得多，其中大部分是由于与读者取得联系的人认为有必要提供的信息，以便他们完成构建并制作合适的电源和为这些构建提供线路接口电路。希望这篇文章中的信息对未来的构建者来说更有用，这是由于这种无休止的更新。

Many of the circuits listed on the web have numerous flaws including distorted input current, distorted ground current without provision for separate grounding, poor sensitivity, inaccurate and unresponsive half-wave operation and many other deficiencies. While this article contains a total of 5 different driving circuits for VU meters, the author considers only Figures 7, 8 and 9 to be worth building in comparison to the other examples. The greatly increased low-end sensitivity of the improved circuitry described here is readily apparent in use and is quite a pleasant surprise to anyone who is used to the less sensitive contemporaries out there. If the meter scale shows an incorrect reading in the compressed lower range of meters designed to compensate for sensitivity limitations, then the movement can be aligned further to the left. This is most often done by adjusting a screw below the glass window anti-clockwise.
网上列出的许多电路存在许多缺陷，包括扭曲的输入电流、没有提供单独接地的扭曲接地电流、灵敏度差、半波操作不准确和不灵敏，以及其他许多不足之处。虽然这篇文章总共列出了 5 种不同的 VU 表驱动电路，但作者认为只有图 7、8 和 9 与其他示例相比值得构建。这里描述的改进电路的低频灵敏度大大提高，在使用中很容易看出，对习惯于不那么灵敏的同类产品的人来说是个令人愉快的惊喜。如果仪表刻度在为补偿灵敏度限制而设计的压缩低量程仪表中显示不正确的读数，则可以将指针进一步向左调整。这通常是通过逆时针调整玻璃窗下面的螺丝来完成的。

Sometimes using a passive meter such as the ANSI type will ultimately be simpler to implement than any active circuitry, so long as the precautions highlighted are taken. It all boils down to cost, space, and simplicity. However, considering the cost of the meter in comparison to the driving circuitry, which is simple enough to take up very little room and can easily be powered from the range of supply rails present in most audio equipment, it is the sentiment of the author that going the extra mile with some of the active circuitry shown in this article is well worth it.
有时使用被动仪表，如 ANSI 类型，最终可能比任何主动电路都简单易行，只要采取所强调的预防措施。这归结为成本、空间和简单性。然而，考虑到仪表的成本与驱动电路相比，驱动电路足够简单，可以占用很少的空间，并且可以轻松地从大多数音频设备现有的电源轨供电，作者认为在本文章中展示的一些主动电路多走一些路是值得的。

denggboo

传图真费事，打包上传