Yeah, well.. Honestly! I don't like "smoothing" in the least! It smacks of enabling intellectual laziness and condescension to newbies (I'm touchy about that sort of thing, being an electronics newbie myself!).
Well of course my likes and dislikes +$50 will still.. Well.. I guess you get it..

So what's the deal? Dialect? Newspeak? Humor gone awry?
Really now? Is ELF signal processing terminology a descriptive lingo all its own distinct from AF and RF?

Please bring me up to speed! Preferences and peeves aside, lord knows communication is key!

Thanks!

 

hi theo'
I have always considered smoothing caps as being used to smooth out rectified mains voltages ie: low frequencies
Filter caps are used to remove unwanted electrical noise from a circuit path. ie: higher frequencies
Eric
Don't forget some caps are called reservoir caps aka smoothing.

 

Smoothing is generally British; filtering is more common in North America. I call them "filter caps" but smoothing is arguably more accurately descriptive of the function - they fill in the gaps between the peaks. Filtering is generally used to describe a selective process.

You best get over being incensed at variations in language real quick or you will constantly be moaning about it instead of accomplishing things.

 

"Smoothing" and "Low Pass Filtering" are essentially the same thing. At best it is a distinction without a difference. I agree with ebp, that semantic nit-picking is the province of pedants and sophists.

 

Look at half wave or full wave rectifier circuit. They are using capacitors to smooth out the output of the rectifier. Make it more uniform. Nothing is being filtered out.

Also note that input into the rectifier is a signal of single frequency... This is different from filter because filter input is signal of multiple frequencies and filter filters out undesirable frequencies.

 

Look at half wave or full wave rectifier circuit. They are using capacitors to smooth out the output of the rectifier. Make it more uniform. Nothing is being filtered out.

Also note that input into the rectifier is a signal of single frequency... This is different from filter because filter input is signal of multiple frequencies and filter filters out undesirable frequencies.

It might be worth pointing out that most LC filters in power supplies will also prevent high frequency noise on the power line from making it down stream. The capacitor following the rectifier may in fact be part of a passive filter.

 

Guys! Thanks! Now I've got it! Where my head was at, all focus was on the actuality of an AC to DC converter's ripple filter being a standard low pass C or LC filter. A single reactance or a network of reactances. Same design formula, same topology same everything applicable to filters operating in higher frequency bands. I had to question whether ELF circuitry had its own terms of art or if it was down to power supplies falling on the borderland between electronics and "electrics".
But now I see the distinction! Unlike most other filter circuits, A DC power supply's ripple filter network strives to reduce all AC signal components as much as possible. To approach pure DC is to "smooth" the output! Perfect!

You best get over being incensed at variations in language real quick or you will constantly be moaning about it instead of accomplishing things.

I agree with ebp, that semantic nit-picking is the province of pedants and sophists.

Yeah! Sorry I came across that way! The tude was mostly a failed attempt at humor, mostly.. As for the rest? Well.. As an "adult newbie" I'm occasionally hypersensitive to anything remotely smacking of condescension. Which, of course, wasn't the case

And just so you know. I'll probably continue to use "filter" in this context but it's not judgmental! It's just how I'm used to thinking (or how my mind works).

Thanks!

 

@theodoravain I totally understand how u feel cuz we all have our like u say _pet peeves_ so just for example with me it's ppl using _transient_ incorrectly or automatically thinking _radiation_ means ionizing radiation. Now with @Hypatia's Protege it's ppl referring to quantities with just inflected unit names (like _Voltage_ for EMF and _Amperage_ for Current).

So FWIW I have just single rule of etiquette for that situation! No matter how much it galls you: Never, ever correct ppl on their usage when it's not changing what they're saying from what they mean!

So in your own words:

communication is key!

But now I see the distinction! Unlike most other filter circuits, A DC power supply's ripple filter network strives to reduce all AC signal components as much as possible. To approach pure DC is to "smooth" the output! Perfect!

Theo I say that's vry good observation cuz putting it like that, filtration is really part of rectification cuz pulsating DC still has time varying component and so is just AC with enough offset to prevent zero-crossing! So purists will say _non zero crossing AC_ is oxymoron but I say u know what I mean

Also @theodoravain I know you weren't being rude in OP but I definitely understand how ppl who don't know you thought you were! It's the downside of you having _teacher gene_ (which, BTW, I'm totally envious of) cuz I say an instructor afraid of condescension is really just afraid of their own medicine!

 

The language does tend to change according to context.

In any filter circuit, energy is shifted around from one component to another - it cannot be otherwise. Where the concern is primarily dealing with gain and phase versus frequency, inductors and capacitors aren't usually thought of primarily as energy storage devices.

When you get into the context of power supplies, inductors and capacitors, at least the "main" ones in the power path, are selected based on their ability to store energy. For example, in a switching power supply when doing the power part of the design, one is concerned with energy storage, current magnitudes, ripple currents and voltages, power losses and the like - everything in the time domain. The "filtering" behavior is largely completely ignored. As soon as the power path has been worked out on those sorts of criteria, the thought processes change to the frequency domain and looking at those same components in terms of frequency dependent gain and phase - the "filter" behavior - and the energy storage issues are largely completely ignored. With the change in context it is common to find some change in language referring to exactly the same parts.

Sometimes the process is the opposite - you design an LC low-pass filter to keep "noise" on its own side the the "barrier" but then have to look at what happens with the inductor that has 30 amperes of DC flowing through it if the current is suddenly interrupted and all that stored energy has to go somewhere. Do the job badly, and you can find yourself with a 100 volt spike on something that normally runs at 20 volts - when you started out trying to keep out some 40 volt spikes.

re are two critical energy storage components - the inductor and the output capacitor. The inductor stores energy during part of the switching cycle when the switch is ON and when the output capacitor is supplying 100% of the energy to the load. When the switch turns off, the energy stored in the inductor is delivered to the output capacitor and the load. But even though the primary purpose of the L & C is to store and deliver energy, as soon as you shift from being concerned with energy to closing the feedback loop, the thought processes change to more conventional "filter" concepts of frequency domain gain and phase.

There are some terms that crop up that seem quite weird. "Leakage inductance" is one i think falls into that category. How the heck can there be "leakage" inductance. Leakage current, conductance and even capacitance all sort of make intuitive sense - a current flowing through and unwanted path, the path itself, and capacitance resulting from simple proximity of conductors at different (varying) potentials.

[EDIT] further to my original post: If I'm working on a switched mode power supply that runs from AC mains, I generally call the capacitors after the input rectifier "filter capacitors", but almost always call those across the final DC output just "output caps." Both store and deliver energy, both have influence on frequency spectrum, and I call them different things.

 

Sometimes the process is the opposite - you design an LC low-pass filter to keep "noise" on its own side the the "barrier" but then have to look at what happens with the inductor that has 30 amperes of DC flowing through it if the current is suddenly interrupted and all that stored energy has to go somewhere. Do the job badly, and you can find yourself with a 100 volt spike on something that normally runs at 20 volts - when you started out trying to keep out some 40 volt spikes.

ebp I say that brings up vry good point a lot of ppl just don't appreciate! By nature of the _beast_ reactive filters are transformation networks! So needless to say it gets _double ugly_ when poor design leaves clear field for resonance! Not just at ripple freq but also parasitic resonances supporting transient-pumped ringing cuz of excessive Q! So I say if ppl would just learn 2 think b4 they do world would definitely be a better place!