【转】采用EL509/6KG6的直藕单端放大器

jupeter

闲暇时浏览网站，发现一篇帘入的文章，觉得有用，原文转载于此供同好鉴赏。顺带借助翻译工具进行简单直译，不喜毋哂。
The Svetlana EL509/6KG6 is a rugged beam tetrode with a plate dissipation
rating of 35W. It may be classified as a "sweep tube" because it was
originally designed as a horizontal sweep amplifier for TV applications. Its
high sensitivity and efficiency allows for up to 15W output from a single
tube. As with any other beam tetrode or pentode, however, distortion in
single-ended operation can be high. Large amounts of feedback are
generally required to reduce distortion to acceptable levels for audio
reproduction. An interesting mode of operating sweep tubes is known as
"screen drive", or sometimes "enhanced triode mode". In this arrangement
the audio signal is applied to the screen grid (g2 or "screen", for short)
instead of the Control Grid (g1 or "grid" for short). This arrangement can be
seen in the circuit schematic in Figure 1. Screen drive is generally much
more linear than standard tetrode or pentode operation; such that lower
levels of feedback are required. Therefore, this type of circuit lends itself
more easily to Hi-Fi applications.

Svetlana（斯维特拉纳）的 EL509 / 6kg 6是一款强壮的具有35W屏耗的束射管。
它可以被归类为“扫描管”，因为它最初是设计为电视应用的水平扫描放大器。它的
高灵敏度和高效率允许单管输出高达15W。然而，与任何其他射束四极管或五极管一样，在
单端功放用途时可能失真度会比较高。通常需要采用较深的负反馈以将重放音频时的失真度降低到可接受的水平。

这种扫描管还有一种比较有趣的模式，被称为“帘栅驱动”，或有时也称作“增强型三极管模式”。在这种接法中，
音频信号被施加到帘栅极(简称G2或“帘栅”)而不是控制栅(简称G1或“栅极”)。这种接法可以从图1的电路示意图中可以看出。帘栅驱动通常比
标准四极管或五极管接法获得更好的线性，这样就仅仅需要很低的负反馈。因此，这种类型的电路就变身为更适合于高保真应用。

jupeter

本制作文章将描述一个典型的帘栅驱动放大器的设计过程。
基于各种考虑、权衡、限制和最终目标综合，最后修改成适应作者自己的需求或口味。最终的成品
放大器在工作时输出大约10W的低失真功率，仅仅采用最少数量的负反馈。

This application note will document a typical screen driven amp design. It is
based upon various constraints, trade-offs, limitations and goals, which may
be modified by the end user to fit one’s own needs or tastes. The final
amplifier produces about 10W with reasonably low distortion while using a
minimal amount of negative feedback.

jupeter

传上曲线，先睡觉去。明天继续

Background:
Although screen drive has been employed by many designers over the
years, there is currently very little published reference material on the
subject. Screen bias curves (plate curves for a fixed grid bias and various
screen voltages) are generally not published. And if they are, it is usually
only for a zero grid bias. Although the required information can be
interpolated using a tube characteristic "nomograph", the process would be
quite tedious.
Luckily, we are able to take advantage of modern computer simulation tools
to do this work for us. This design effort relied heavily upon a PSPICE model
of the 6KG6/EL509 provided by Duncan Munro on his web site This model
was used to develop a full set of screen bias curves for the desired grid bias.
Actual test data was closely correlated with the model.

fsnzman2016

和当年trs推荐的一样吗？

jupeter

fsnzman2016 发表于 2018-10-14 09:49
和当年trs推荐的一样吗？

图纸可能推荐过。但是这篇文章说的更加详细一些。

背景介绍:
尽管近年来帘栅驱动已经被许多设计者所采用，但其相关资料则很少有公开。例如帘栅偏置曲线(固定栅压下的的屏极-帘栅极曲线)通常不公布。即使有，通常也是仅针对零栅极偏压的。尽管上述所需的信息可以通过使用管特性“诺模图”插值变换得来，其过程将是相当乏味。
所幸的是，我们能够利用现代计算机模拟工具为我们做这项工作。这项设计工作严重依赖于PSPICE模型，邓肯·芒罗在其网站上提供的6kg 6 / EL509中
用于为期望的各个栅压下开发出全套的帘栅极曲线。注意实际测试数据与模型精度有关（仅供参考）。

fsnzman2016

jupeter 发表于 2018-10-14 10:22
图纸可能推荐过。但是这篇文章说的更加详细一些。

背景介绍:

明白，那个帖子也算是胆x轩的第一高楼了

jupeter

设计/讨论:
Svetlana EL509数据表中给出了帘栅极输入曲线，但却是零栅极偏压下的曲线。显示出帘栅极具有惊人的输入灵敏度，然而，这些曲线显示的实际上是有利有弊的。
因为帘栅极的驱动电流随输入信号的大小改变而改变，要求前面最好基于直流耦合的驱动级。在如此高的帘栅极灵敏度下保持直流偏压的稳定将是一个挑战，尤其是对与直藕的全电路。于是，其中一种解决方案是提供负栅极偏压以降低帘栅极灵敏度。同时带来一个好处是让帘栅极取得一个较高的静态偏置电压，来保持稳定的静态屏流。这使得直流耦合变的非常容易，因为前级屏极电压通常就是100伏或略高。

jupeter

The simplest way of providing the required grid bias is by using cathode
bias. This arrangement is self-compensating, so it will further enhance DC
stability. For this design, a grid bias of -33V was chosen. (This was actually
an iterative process based upon the desired plate current and screen voltage
levels.) The screen bias curves for -33V bias are given in Fig. 2. Note that
Fig. 2 also includes screen current curves. Screen current is surprisingly low
(<10mA). The operating point was chosen based upon two component
limitations. Maximum plate dissipation for the EL509 is 35W, and the
maximum DC current allowed in the output transformer is 110mA. So a
plate voltage of about 35W / 110mA = 318V could be used. But for
conservatism, the combined plate and screen current was reduced to
100mA. This determines that the cathode resistor is to be 330 ohms. Since it
will dissipate 33V **2 /330 = 3.3W a wire-wound resistor rated for 10 W, or
more, is recommended.

最简单的提供所需的帘栅极偏置方式就是采用阴极输出器。这种方式具有自动补偿功能，将进一步增强直藕的稳定性。
本设计中，栅极负压选择33V，（这个值的选择是综合所需的屏流大小和所需的帘栅极电压来考虑的）。栅偏压在-33V下的
帘栅极输入曲线如图二所示。注意图二同时也包含了帘栅极电流曲线，帘栅极电流竟然出奇的低。（小于10MA），
工作点的选择基于两个元件的限制，一个是EL509的最大屏耗是35W，一个是输出牛初级允许最大通过电流是110MA。
于是屏压大约可以选择35W / 110mA = 318V，为了保守起见，将帘栅极和屏极的电流之和减小的100MA。可以确定住
阴极电阻是330欧，消耗的功率是33*33/330=3.3W，建议选择标称功耗10W甚至更大的线绕电阻。

qjh

请问27v灯丝电压的pl509可以用吗？

jupeter

qjh 发表于 2018-10-14 11:09
请问27v灯丝电压的pl509可以用吗？

PL509的灯丝电压是40V吧，当然能用，但是屏耗略小，30W，实际使用时可以适当加大阴极电阻，使实际屏耗不超过设计值便可。

jupeter

Referring to the characteristics of Fig. 2, the required screen voltage at
Eb=320V and Ib=100mA is about 154V. Since this voltage is specified with
respect to the cathode, the actual screen potential (measured with respect
to ground) would be 154V plus the 33V of cathode bias, or 187V.
The input/driver stage is based upon the Svetlana 6BM8 triode-pentode.
The triode section is used as a low distortion gain stage. The power pentode
section is used as a high-current, low impedance cathode follower to drive
the screen of the EL509 output stage. All three stages are DC coupled. The
EL509 screen potential of 187V allows for at least 160V on the plate of the
input tube (after considering the bias of the cathode follower stage). This
provides for sufficient plate voltage swing and enough bias to provide
adequate input signal headroom.

参照图二，在电压Eb=320V 和 Ib=100mA 时，帘栅极电压应该是154V，由于此电压是以阴极为参考点，实际的帘栅极电位（对地）就是
154V加上阴极的电位33V，即187V。

输入级兼驱动级采用斯维特拉纳的6BM8三-五极复合管。其中的三极管单元用于低失真增益放大器，功率五级管用于大电流、低阻抗阴极输出器以推动EL509输出级的帘栅极。三级全部采用直流耦合。EL509的帘栅极电位187V要求输入级管子的屏极电压起码160V（已经考虑阴极输出级的偏置电压）。这样的安排使屏极的摆动以及驱动信号具有充分的空间。

jupeter

本来图省事，用机翻，谁知翻出来实在不像“话”，结果还是人工翻译贴切一点。偷懒行不通。

qjh

记性不好了应该是pl504，屏极耗散功率16w，可以用吗

jupeter

qjh 发表于 2018-10-14 12:32
记性不好了应该是pl504，屏极耗散功率16w，可以用吗

可以，还是调整一下阴极电阻的大小，改变屏流至不超屏耗即可。这个电路实际上可以适应很多束射管或五级管。

jupeter

The pentode section is triode-connected by connecting its screen to its plate
(via a 100 ohm resistor). This cathode follower stage must be able to supply
both the DC and peak screen currents. Therefore the quiescent cathode
current is set to about 25mA. Neglecting screen current, the cathode resistor
was selected to be 187V / 25mA = 7.5K. This resistor will dissipate 187V
**2 / 7.5K = 4.7W, so a 10W, or larger, resistor is required.

五级管部分，将帘栅极与屏极通过一个100欧的电阻相连，成为三极管连接。这个阴极输出器须提供静态电流和峰值时的帘栅极电流。因此静态时的阴极电流定为大约25MA。忽略掉帘栅极电流，阴极电阻应该选择187V/25MA=7,5K。这个电阻消耗功率是187V^2/7.5K=4.7W。可以选用10W或更大功率的电阻。