Capacitance Multiplier in Preamp PSU

[Matt C]

I'm wondering what people's thoughts are on using a capacitance multiplier circuit on the output of a LM317-based PSU for something like a mic preamp (or other line level equipment.  In my case I'm trying it with a Hamptone-style JFET preamp). 

It seems like a relatively simple way to maximize PSU filtering, but I'd imagine I'm more likely to run into noise and hum issues because of improper layout/wiring/grounding, rather than inadequate filtering.  Is the LM317 good enough that the capacitance multiplier just won't make any noticeable difference?

 

[JohnRoberts]

I'm wondering what people's thoughts are on using a capacitance multiplier circuit on the output of a LM317-based PSU for something like a mic preamp (or other line level equipment.  In my case I'm trying it with a Hamptone-style JFET preamp). 

It seems like a relatively simple way to maximize PSU filtering, but I'd imagine I'm more likely to run into noise and hum issues because of improper layout/wiring/grounding, rather than inadequate filtering.  Is the LM317 good enough that the capacitance multiplier just won't make any noticeable difference?

A capacitance multiplier makes little sense on the output of a voltage regulator. I used what was essentially a capacitance multiplier as the unregulated pre-supply for a big console. The soft regulator (capacitance multiplier)  delivered roughly +/- 18V that got regulated down to +/-15V on the strips.
======

The output of an active IC regulator should already be low impedance, and capacitors mainly supplant the HF source impedance.****

The 317 "should" be OK, and layout always matters. The output regulation will only be as clean as the regulator's ground reference.  Note: I have no idea what a "Hamptone" is so take my comments with a grain of salt..

JR

*** I did some experiments with the old 3 terminal regulators (78xx and 79xx) back in the day. They were based on circa 741 op amp technology so had poor HF gain bandwidth product so their HF output impedance measurably suffered. I found on my bench tests that a roughly 1,000 uF electrolytic cap's falling impedance with frequency supplanted the old 3-terminal regulators rising impedance neatly (drive a modest AC current into the regulator output and look at the DC voltage). The inductance of the big electrolytic prevented the cap from destabilizing the regulators at extreme HF. The newer 317 should already have superior output impedance compared to old school 3-terminal regulators. 

 

[gridcurrent]

high frequency "hair" is easily observed on the output of the 78/79 regulators.
not so sure about the 317 family.
in any case, hanging parts on the output of a regulator is less effective than local decoupling.
SSL used 22 ohms and 10 uF in well regarded designs
The "'value" brand that was based in Fort Lauderdale used 10 ohms with fairly large electrolytics.