[undertow]

Okay I have experimented with "capacitor bypassing" in the past.

Never really great or (necessary) results from ever doing this in speaker crossovers unless it was forced in order to perfectly match a certain value because a single size available cap would not achieve it.

We are talking about full range signals, not a tweeter or something in a passive crossover.

My question is in general being this bypass cap is going to effect certain "Higher Frequencies" what octave range is really the cutoff that these bypass caps kick in and take over? Or is it just effecting the entire range equally?

For example here is a basic octave chart below so do they mostly take over everything from 4 kHz and up? 6 kHz and up?

[Tyson]

When you use a bypass cap on a full range signal, it's not really helpful to think of it in terms of it's affect on different specific frequencies.  It's more like a window - better caps give you a clearer window onto the music.  I'm assuming your using this in an amp, preamp, or DAC?

[undertow]

No I am just curious, not using on anything right now. Essentially the higher frequencies would flow thru the path of least resistance or less capacitance faster in this case as the smaller cap will release the signal a bit faster making a certain range smoother, and faster pushing the signal.

[Speedskater]

No I am just curious, not using on anything right now. Essentially the higher frequencies would follow the path of least resistance or less capacitance in this case as the smaller cap will release the signal a bit faster making a certain range smoother and faster pushing the signal.

I don't think that the capacitors know any of that stuff.
Any signal will follow all available paths, inversely proportional to the impedance of that path at the signal's frequency.
By the way, capacitors  (for the most part) don't act as resistors to the signal path.

[undertow]

I understand that. I used it as and example to express the behavior of the frequencies.

[BobM]

I was always of the opinion that a small bypass on a crossover tweeter circuit assisted the "airiness" of the music out of the tweeter. But a small bypass on a coupling cap in a DAC/preamp/etc assisted in the overall transparency and openness of the overall sound.

So less about actual octave range when used in a circuit but more of that when used in a crossover. Of course, I'm probably completely wrong.

[Speedskater]

I understand that. I used it as and example to express the behavior of the frequencies.

But it's still a incorrect example.

[undertow]

Well saying "Path of Least resistance" I was not really referring to the literal resistance only effecting one bandwidth over another. Obviously all the frequencies are still flowing equally not being filtered out or anything.

[Speedskater]

OK, we'll change it to:
Essentially the higher frequencies would follow the path of least impedance

In this case with film capacitors, the impedance will be capacitive reactance.

Now we have all read the statement:
"Current follows the path of least resistance"
But it's incorrect! Current follows all available paths. It's true that most of the current will take the path of least resistance but there will be some current in each of the other paths.
The same logic applies for signal flow.

So in our circuit, we have a big capacitor and a small capacitor in parallel. In about a 100 to 1 ratio.  The capacitive reactance has the same ratio as the capacitance but inverted.  So at any given frequency the big capacitor will have only 1% of the impedance of the small capacitor. That is about 99% of the signal will flow through the big capacitor.

But if the big capacitor is an electrolytic and the small capacitor is a film. Then we might have to measure the electrolytic  at the frequency in question to determine the ratio.

[Davey]

Undertow,

Subjectivists need to find some technical explanation for what they're hearing.  When they can't you get silly analogies like "clearer windows into the music."
Unfortunately, this is one of those topics that's been horribly skewered by audiophiles.  Even the label "bypassing" is completely inaccurate.

I suggest to step back and consider this topic anew.  SS, as usual, is right on the mark with his technical explanation.  That's an excellent place to start.

Dave.

[Tyson]

Undertow,

Subjectivists need to find some technical explanation for what they're hearing.  When they can't you get silly analogies like "clearer windows into the music."
Unfortunately, this is one of those topics that's been horribly skewered by audiophiles.  Even the label "bypassing" is completely inaccurate.

I suggest to step back and consider this topic anew.  SS, as usual, is right on the mark with his technical explanation.  That's an excellent place to start.

Dave.

And objectivists say silly stuff like "if it measures the same then it sounds the same". 

The rest of your post was good, why start it off by denigrating the majority of people that enjoy this hobby?

[Davey]

And objectivists say silly stuff like "if it measures the same then it sounds the same". 
The rest of your post was good, why start it off by denigrating the majority of people that enjoy this hobby?

I've never said that.....but it is, essentially, correct......with some caveats.
And I don't believe they(you)'re in the majority.

Anyways, this is The Lab Circle......which, as I've suggested previously many times, should be a subjectivist-free-zone.  But, nobody....wait for it....listens. 

Cheers,

Dave.

[undertow]

No big deal. I have heard good and bad using smaller caps to bypass regardless of any proven measurement.

I guess bypass is not the right technical term as there are basically bypass caps which are officially in all circuits for a whole different purpose than we are talking about here.

Maybe we can coin the phrase "Audiophile Paralleling" caps.  I can agree it may or may not be a fruitful practice, but does something to cause a technical shift in the way we hear it on the other end.

Mystery closed

[Davey]

Actually, "shunt" is a much more accurate term.  The term "bypass" capacitor dates to many years ago and (generally) describes a capacitor "bypassing" a biasing resistor in amplifier circuits to increase gain.  You would see them mostly across cathode resistors in vacuum tube circuits, but in many other places as well.  I repaired many old radios/TV's and replaced quite a few "bypass" capacitors. 

Dave.

[undertow]

I don't know if "Bypass" is close to an obsolete term or not as its in many other uses from what I can tell...
To summarize, the bypass capacitor is used to dampen the AC component of your DC circuits. By installing bypass capacitors, your DC circuit will not be as susceptable to ripple currents and voltages.
But I do agree "Shunt capacitors" sounds like a better term.

http://www.seattlerobotics.org/encoder/jun97/basics.html

http://www.electro-labs.com/bypass-capacitors-why-and-how-to-use-them/

[Davey]

I didn't say "bypass" was an obsolete term.......just that it dates back a long way......and it's misapplied to the application you've queried about here.  It's still a perfectly valid configuration/technique and capacitors are used in this way in many electronic circuits.  As an example, "local bypassing" is very common around componentry in much digital and analog circuits.  (The two links you posted are applicable to that usage.)

Regarding the audiophile usage of the configuration:  The (primary) underlying assumption here is that the lower ESR of the "bypass" capacitor relative to the primary capacitor will have an audible effect on the result in the passband even though any measurable difference will be well above the audio band.  However, some audiophiles attribute subjective differences to the dielectric materials used, or some sort of synergistic combination of poly and electrolytic, or any number of other magical combinations to arrive at bypass nirvana.

If you're actually "effecting an octave range" then I don't think you're describing the typical "bypass" usage that most audiophiles are thinking about.  SS and Tyson have alluded to this already.  Hope that makes sense.

The whole topic is a can of worms. 

Cheers,

Dave.

[barrows]

"The whole topic is a can of worms."

Yes, and I think that is why we the OP posted the question, right?  I like simple circuits for my DAC output stages (generally) so this means coupling caps are necessary.  Hence, I am interested in the discussion and not so interested in putting down audiophile beliefs.
I will use the term bypassing, because it is commonly used for this, whether "correct" or not.

speedskater posted that with a 100-1  (say 1uF bypassed with .01uF) ratio only 1% of the signal will pass through the bypass cap.  If this is true in the real world, it would suggest to me that adding a bypass is meaningless, right, or does that 1% really matter?

[Davey]

"The whole topic is a can of worms."

speedskater posted that with a 100-1  (say 1uF bypassed with .01uF) ratio only 1% of the signal will pass through the bypass cap.  If this is true in the real world, it would suggest to me that adding a bypass is meaningless, right, or does that 1% really matter?

Is that a rhetorical question?  I sense this is the bait-and-switch. 

Regardless, if you feel this "1%" yields a subjective improvement in your system, who am I to argue with you?

I suppose this all assumes the program material of interest has some content in the 1% range, yes?  If it doesn't then, by definition, the bypass would be meaningless, yes?

Dave.

[barrows]

No trickery meant on my part Dave.  It is just a straight forward honest question as stated.  Assuming speedskater is correct, does that 1% matter, is it likely to be even audible and if it is, is it likely to be an actual improvement, or just different.

[*Scotty*]

I one were to look at the impulse or square-wave response of the circuit with the bypass in place the leading edge might look cleaner.
Scotty