都说 LM317噪声大，我来发个用于唱放低噪声的。

微尘尘

如图，原理图来了

suzuki352

全部复杂多了，名机里有用78XX，79XX系列的，不一样好声卵跌！

newayjia

suzuki352 发表于 2014-2-16 15:07
全部复杂多了，名机里有用78XX，79XX系列的，不一样好声卵跌！

金玉良言

bazys

这个电路不复杂。

kimm

看起来很复杂。

微尘尘

国外人的实验，英文不好，发原文上来。
“
a LM317 regulator has a noise of several hundred nV/rtHz. Even with a 50dB PSRR this is still to high for an overall 0.3nV/rtHz design target. Therefore, the power supply noise has to be filtered out. A standard emitter follower power buffer (a.k.a. capacitance multiplier) does a pretty good job, however there's one thing that it can't filter out: the emitter follower Rbb. To avoid that contribution, we need some sort of local feedback loop, to obtain a significant lower equivalent Rbb. This is easily achieved by using a Sziklay configuration for the emitter follower. By including the power device in a current feedback loop, it's Rbb is effectively cancelled. It's now the low power device (which can be a low noise device) that dictates the equivalent Rbb of the emiteer follower. The power buffer schematic build around this concept (and with the parasitic elements included) is shown below. It is only the Q2 Rbb (a low noise device) that contributes to the power supply noise. Divided by the head amp PSRR, this contribution is negligible.”

zly1980

MARK!