软件翻译
Mr. Linsley Hood is quite correct when he states that the operation of transistor output stages
in class AB can cause increased distortion, because of the change in the slope of the transfer
characteristic around the crossover point. However, I fear that he is wrong in supposing that a
low source impedance overcomes the problem.
Linsley hood 先生的观点是正确的，他说 ab 类晶体管输出阶段的操作会导致失真增加，因为在交叉点附近的传输特性斜率发生了变化。 然而，我担心他假设低电源阻抗克服了这个问题是错误的。
Fig. 1 shows a test circuit which I constructed to measure the transfer characteristic of the
output stage under various bias conditions and the results are shown in Fig. 2 for 200mA, 20mA and 0mA. Note the prominent change in slope at 200mA bias. In the test circuit the
transistors are operated in the common emitter mode to enable the changes in the slope of
the transfer characteristic to be seen more easily, but this does not alter the validity of the
results since the effect of putting the load into the emitter circuit is only to provide local
negative feedback. Under the same conditions a push-pull emitter follower using an output
stage with the transfer characteristic of Fig. 2(b) will produce less distortion than a similar out-
put stage with the transfer characteristic of Fig. 2(c).
图1示出了我构造的测试电路，该测试电路用于测量输出级在各种偏置条件下的传输特性，结果如图所示。 200毫安，20毫安和0毫安。 注意斜率在200ma 的显著变化。 在测试电路中，晶体管以共发射极模式工作，以便更容易地看到传输特性斜率的变化，但这并不改变测试结果的有效性，因为将负载放入发射极电路的效果只是提供局部负反馈。 在相同的条件下，使用具有图2(b)传输特性的输出级的推挽式发射极跟随器比具有图2(c)传输特性的类似输出级产生的失真小。
To check this I constructed Mr. Linsley Hood's amplifier and measured the distortion at
200mA and 20mA bias current with a Marconi TF2330 wave analyser and TF2100/1M1 low
distortion oscillator. The results are shown in Fig. 3 and show clearly the improvement in
distortion at intermediate output levels produced by the lower bias current. However, in spite
of the excellent results obtained I would not advise constructors of this amplifier to use a bias current as low as 20mA as it tends to be rather unstable. A bias of 50mA would be about the
optimum and at this level there would still be a "hump" in the distortion curve but it would be
smaller than at 200mA bias and removed to a lower power level. I would also consider the
use of a temperature compensating diode or transistor in the bias network strongly advisable,
to minimize thermal variations.
为了检查这一点，我构建了 linsley hood 先生的放大器，并用一个Marconi tf2330波形分析仪和 tf2100 / 1m1低失真振荡器测量了200ma 和20ma 偏置电流的失真。 结果如图所示。 3，并明确显示改善失真在中间输出水平产生的较低偏置电流。 然而，尽管取得了优异的结果，我不建议这个放大器的建设者使用低至20ma 的偏置电流，因为它往往是相当不稳定的。 偏置50ma 将是最佳的，在这个水平仍然会有一个“驼峰”的失真曲线，但它将小于200ma 的偏置和得到一个较低的功率水平。 我还认为在偏置网络中使用温度补偿二极管或晶体管是非常明智的，以尽量减少热变化。
发表于 2020-5-19 05:33
楼主