Low-pass filters in PSUs?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Ethan

Administrator
Admin
Moderator
Joined
Jun 3, 2004
Messages
1,602
Location
DC
[Warning: Ethan's been trying to study again...DANGER!]
Is the rationale of using low-pass filters in PSU's to make them quieter? Bare with me on this one... How does the low pass filter work in a DC circuit??? Where's the inductive or capacitive reactance here?

Is the random thermal noise, and other high frequency junk, enough "AC" to make the filter work? I know that's phrased strangely. What I mean is will it have an effect? How would you design the filter, with what f-3 point would be appropriate? I would think as low as possible would be OK (even down to 1Hz) since you're concerned with having pure (...as possible) DC, right?

This seems too easy and good to be true. The only downside I would think would be higher output Z from the added load?

[Insert Umbrella Icon here--I feel a storm abrewing :shock: ]
 
A power supply filter is just a lowpass with its corner frequency as close to "zero" as is practicable. You do not want alternating currents of any frequency on the supply, just DC. The output of a rectifier is pulsed DC, a filter cap smooths this out to a sort of sawtooth waveform, and subsequent lowpass filtering gets rid of the "ripple" altogether, leaving only DC or something deemed close enough :wink:
 
Google found this pretty extensive, yet not TOO geeky tutorial:

http://technology.niagarac.on.ca/courses/etec1120/Files/Unit12.pdf

Bri
 
[quote author="Ethan"][Warning: Ethan's been trying to study again...DANGER!]
Is the rationale of using low-pass filters in PSU's to make them quieter? Bare with me on this one... How does the low pass filter work in a DC circuit???
[/quote]

You can look at a DC voltage as an AC voltage with a frequency of 0. So any lowpass filter will pass DC and any highpass filter will block DC.

[quote author="Ethan"]
Is the random thermal noise, and other high frequency junk, enough "AC" to make the filter work? I know that's phrased strangely. What I mean is will it have an effect? How would you design the filter, with what f-3 point would be appropriate? I would think as low as possible would be OK (even down to 1Hz) since you're concerned with having pure (...as possible) DC, right?
[/quote]
Yes, right.
The only problem is that to get a the f-3 point lower you need bigger capacitors to ground or bigger inductors in series. So the PSU will get more expensive.
This will get worse when the load impedance gets lower( = the amount of current you pull from the psu gets higher) this is why you need big caps in big psu's.
 
[quote author="Ethan"]Bare with me on this one... [/quote]

I'd rather not.

-I don't mind bearing with you, but I don;t know you well enough to get naked together... -just yet!!! :shock:

:wink:

Keith
 
Thanks everyone for your help, but I think only Mendelt understood my poorly worded questions (or so I thought at first). My question stems from reading about passive filters--it triggered a memory of having read about LM317 regulator noise. I forget the website that showed test results, but in passing, I remember it mentioned a low-pass filter after the regulator.

I COMPLETELY overlooked that a simple PSU with rectifier followed by capacitors with nominal resistance, IS a blatant lowpass filter (reading NYD's post now makes sense why his answer seemed like he didn't get my question--o contraire!)

Now, I feel like I just asked one of those self-evident questions, Doh! :grin:
I should have specified "lowpass filter after the regulator", in which case the same idea would apply, but you're also dealing with how the regulator reacts to different loads. I remember it said something about high freq. oscillations at very low Z. I wish I could find that website again.

Thanks Mucho!
 
"I wish I could find that website again. "

You might be thinking about this one, which I included in a post several months ago iirc:

http://www.tnt-audio.com/clinica/regulators2_impedance1_e.html
 

Latest posts

Back
Top