AMZ Power Supply Filter alternative part question

[mozz]

All this is way over complicated if it's use is going to be a pedal power supply filter. Assuming you already have DC, just about any cap and any inductor will work, if you want better add more CLCLCLCLCLCLCLCLCLCLCLCLCLC.

[antonis]

Shouldn't there be a voltage divider near C2, such that Vce is maybe around 0.5V at least, or maybe 1V ?
If Vce is zero, maybe we cannot handle down swings on the ripple

V_CE can't be zero for none other "ideal" BJTs..
TR1 configuration is the simplest (and most ineffective) BJT bias one but here it only serves as capacitance multiplier..
[C2*(h_FE + 1)]

[antonis]

Assuming you already have DC,

What DC..??

Just DC..
More DC than above..??
Ultimate DC..??
Absolute DC..??
THE DC..??
Enchephalogrm of a Pontious(*)..??

P.S.
(*) Greek racist denomination - something like "Belgian" for French..

[bushidov]

Here is a quick and simple power supply filter that I drew up for someone not long ago:

Nice! I'll breadboard it this weekend and fiddle with it.

What DC..??

What Antonis said. I can make assumptions on "my" power supply, but not others.

[antonis]

@Erik: I think you're a perfectionist..
(but let it stay between us..)

[PRR]

> - Feeding a zener with a current source, which gives low ripple without a cap.
> - There's a self bias zener circuit,
> You can add an output transistor for more current.
> It's possible to reg-jig this for lower drop-out
> Notice how the zener and feedback keep the zener current constant
> These are all old school designs.

If you are doing old-school (sea attached), and need more porn, then you want Hans Camenzind's free e-book.

[Rob Strand]

I'm sure this is an ancient idea :




(Or use LDO with an adjust pin or feedback pin and use emitter follower/capacitance multiplier with it)
( Or use a Zener instead of a LDO regulator)

Caveat : Always operate in safe zone as per Rob's formula

It should work.  The active filters are often added after a regulated supply in ultra low noise designs.   The active filter is only removing very small amounts of ripple and noise so there's no drop-out issue.  On the schematic the only difference is the collector connection.  Your schem will have a low drop at the active filter and will fend-off more ripple before dropping out if it were present.


See figure 4.   The idea here is the discrete design doesn't have great PSRR so you add an active filter.   The assumption is the supply rail is already clean.

http://application-notes.digchip.com/006/6-8814.pdf   

IIRC, the chapter on noise in the Art of Electronics book has a noise measuring jig with a multi-stage active supply filter.

[Rob Strand]

All this is way over complicated if it's use is going to be a pedal power supply filter. Assuming you already have DC, just about any cap and any inductor will work, if you want better add more CLCLCLCLCLCLCLCLCLCLCLCLCLC.

Even RC filters work.

I'm not sure exactly what the complete set-up is.  It's only clear we are dealing with mains ripple.  Normally a regulated supply will be enough for a pedal.  If the regulator is dropping out and ripple is getting through then the solution is to stop that happening.  Either raise the input voltage or drop the output voltage.    Another way is a bigger supply filter cap on the input.  It reduces the input ripple and raises the average voltage, often enough to stop regulators dropping out.

[Rob Strand]

For filters we get more rejection by increasing the filter order or by lowering the cut-off.

For the old-school current source designs the idea is the current does not change and it's largely independent of frequency upto fairly high frequencies.

Here's a sim showing ball-park performance you can get from those old-school designs.   You can see why Bob Pease still used it in the video. The designs I've shown here push for lower drop-out where are Bob Pease's design would go for lower tempco.  The difference is the top part of the current source uses a transistor (also choice of zener voltages come into tempco considerations.)



In the sim I step up the input voltage around the point where it drops out.  You can see the performance drop below that point.  However, once the circuit kicks in fully the performance flattens off.  You can also see how the "ring of two" out performs the basic current source.

FWIW, the LM78xx regulators are probably about the same as design 1, at least at low frequencies

[bushidov]

@Erik: I think you're a perfectionist..

Well... you know... how am I going to disagree with that?

From a practicality standpoint, assuming my wall wart power supply is delivering 9V, but is noisy, likely a 60 Hz noise, and I implement these filters, how much ripple is still "acceptable"?

For example, I take a couple of hundred of millivolt ripple and reduce it to say a couple of tens of microvolts of ripple. Is that acceptable for a guitar pedal, such as one with high gain?

[Vivek]

That depends upon

A) Your idea on what amount of hum is acceptable

B) The ripple rejection ratio of the pedal you have