[goldlizsts]

Hi all... learning about speaker crossovers.  Here's my question for this time:  for a capacitor need of say 80 uf, is it better to have one (1) single 80uf cap or, for that matter, 4 X 20uf + a 0.1uf, or even 30+20+20+0.1uf kind of combination?  Which is better?  Are there any merits for the multiple caps vs a single higher value one?  I was told that for a 4 X 20uf lineup, they would charge up faster, hence a faster response time, etc.  Any expert advice is appreciated.

[brj]

See this thread in the lab circle: Single vs. multiple capacitors

[Ethan Winer]

GL,

> I was told that for a 4 X 20uf lineup, they would charge up faster, hence a faster response time <

You were told wrong. Especially for a crossover (as opposed to bypass capacitors on a circuit board), all that matters is the total capacitance. However, all capacitors are not the same, and some types are better than others for different applications. But that's a different issue from what you asked.

--Ethan

[art]

There are some disadvantages to putting film caps in parallel, if they touching each other. Probably not an issue for a crossover, and too technical to go into details on this sort of forum.

In this case, size, price, availability may be the deciding factors.

Pat

[goldlizsts]

GL,

> I was told that for a 4 X 20uf lineup, they would charge up faster, hence a faster response time <

You were told wrong. Especially for a crossover (as opposed to bypass capacitors on a circuit board), all that matters is the total capacitance. However, all capacitors are not the same, and some types are better than others for different applications. But that's a different issue from what you asked.

--Ethan

Thanks.  I am getting conflicting advices, from both sides of the fence.  I tend to agree the quality of the caps matters.  I should have added a 0.1uf also to the lineup.  Would that change the issue?   Mark

[goldlizsts]

GL,

> I was told that for a 4 X 20uf lineup, they would charge up faster, hence a faster response time <

You were told wrong. Especially for a crossover (as opposed to bypass capacitors on a circuit board), all that matters is the total capacitance. However, all capacitors are not the same, and some types are better than others for different applications. But that's a different issue from what you asked.

--Ethan

Thanks.  I am getting conflicting advices, from both sides of the fence.  I tend to agree the quality of the caps matters.  I should have added a 0.1uf also to the lineup.  Would that change the issue?

[goldlizsts]

GL,

> I was told that for a 4 X 20uf lineup, they would charge up faster, hence a faster response time <

You were told wrong. Especially for a crossover (as opposed to bypass capacitors on a circuit board), all that matters is the total capacitance. However, all capacitors are not the same, and some types are better than others for different applications. But that's a different issue from what you asked.

--Ethan

Thanks.  I am getting conflicting advices, from both sides of the fence.  I tend to agree the quality of the caps matters.  I should have added a 0.1uf also to the lineup.  Would that change the issue?   Shek

[BrunoB]

Hi all... learning about speaker crossovers.  Here's my question for this time:  for a capacitor need of say 80 uf, is it better to have one (1) single 80uf cap or, for that matter, 4 X 20uf + a 0.1uf, or even 30+20+20+0.1uf kind of combination?  Which is better?  Are there any merits for the multiple caps vs a single higher value one?  I was told that for a 4 X 20uf lineup, they would charge up faster, hence a faster response time, etc.  Any expert advice is appreciated.

A link about cascading : http://www.northcreekmusic.com/Bypassing.html

[audioengr]

I believe the answer is: it depends on the capacitors.

Some caps, like Multicaps are already paralleled inside.  Others, such as oil caps may have very thin leads and high ESR as a result.  In this case, it may be advantageous to parallel smaller values to get the ESR down, but the benefit probably depends on the efficiency of the drivers in the design.  Bypassing with smaller caps has never worked for me in the past.  However, I hate to an infomercial, but the Chis VenHaus Teflon V-caps really do work as bypass caps, so I've changed my mind on this.

[brj]

Bypassing with smaller caps has never worked for me in the past.  However, I hate to an infomercial, but the Chis VenHaus Teflon V-caps really do work as bypass caps, so I've changed my mind on this.

Steve, can you list what other types of caps you've tried that definitely did not seem effective as bypass caps?

Thanks!

[GHM]

Here's a site I found sometime back. I've tried several caps on this list.The guy doing the sound  comparisons seems pretty accurate too me.
http://home.zonnet.nl/geenius/Cap.html
http://home.hetnet.nl/~brian.gee/Cap.html

[ctviggen]

I find it interesting that bypass capacitors really work, as "theoretically" a bypass capacitor simply adds to the total capacitance of the node.  It's not clear to me why certain frequencies would go to particular capacitors (apparently, smaller valued caps - higher frequencies?).  I'll have to think about this more.

[Steve]

I would not put a whole lot of faith in capacitor comparison articles based on listening; especially when one simply installs the cap in a component.  

[audioengr]

Bypassing with smaller caps has never worked for me in the past.  However, I hate to an infomercial, but the Chis VenHaus Teflon V-caps really do work as bypass caps, so I've changed my mind on this.

Steve, can you list what other types of caps you've tried that definitely did not seem effective as bypass caps?

Thanks!

I have tried them all.  I dont want to start flame wars by naming names....

[Ethan Winer]

Bob,

> I find it interesting that bypass capacitors really work, as "theoretically" a bypass capacitor simply adds to the total capacitance of the node. <

In theory maybe, with perfect capacitors. But all real-world capacitors have a combination of capacitance, inductance, and resistance.

The kind of capacitors that are good for audio are "wound" using long strips of metal separated by plastic. The length of the winding increases the cap's inductance, which tends to block the highest frequencies. This is likely irrelevant in a crossover that operates at audible frequencies, but power supplies need to avoid high frequency oscillation. So there, the large electrolytic caps (also wound) often have a disk ceramic cap in parallel.

Disk ceramic caps are not so good for audio, nor do they come in large values, but they're not wound so they have very little series inductance. This is why a very large electrolytic cap is often put in parallel with a very small disk ceramic cap.

--Ethan

[ctviggen]

Thanks, Ethan.  That makes sense.  The problem with having an EE degree is that we learn everything theoretically, but our actual, hands-on knowledge is low.  I gained experience as an engineer, but it was on the digital side -- very few engineers these days have anything to do with analog.

[jeffreybehr]

I would not put a whole lot of faith in capacitor comparison articles based on listening; especially when one simply installs the cap in a component.  

Oh yea.  We certainly wouldn't want to test AUDIO components by listening to them!

OK, sorry, sarcasm attack over.

Pray tell, Steve--why not?  That seems to me like saying we wouldn't want to determine which IC cable sounds best in our systems by inserting them into our systems and listening carefully.

Perhaps there were flaws in his testing methodology?  If your objection is based on the fact that some of us make conclusions FAR too quickly and NOT carefully, I agree, but did that person commit those sins?

[Steve]

<<

Oh yea.  We certainly wouldn't want to test AUDIO components by listening to them!">>

   No offense taken Jeff. In different components, some caps will sound better than others. As a simple example, a component sounding bright with a "perfect" cap installed May sound better if a "darker" sounding cap is installed to even out the tonal balance. But that "darker" sounding cap isn't better, it is just compensating for a bright sounding component. I think it would be much better to have a tonally balanced component with a tonally balanced capacitor. Below will explain more fully my point.

<<"Pray tell, Steve--why not?  That seems to me like saying we wouldn't want to determine which IC cable sounds best in our systems by inserting them into our systems and listening carefully.">>

Agreed. Same with ICs. If tonally bright components were connected by tonally "dark" sounding ICs, the components Might sound ok. But I would rather have tonally balanced components with tonally balanced ICs.

Again these are just simple examples.

<<Perhaps there were flaws in his testing methodology?>>

I think that is right. He was only testing the caps in one, I believe, component. How it will sound in a different component will depend on the next component's sonic signature.
 
For example, the author gave a pretty high score to one cap, but when I tested it (awhile back) to a straight wire (which is tonally balanced), the cap was waaaaaay off, very bright sounding. That told me the authors component was way off, very dark sounding. I think that skews the results of all the capacitors tested.

<<"If your objection is based on the fact that some of us make conclusions FAR too quickly and NOT carefully, I agree, but did that person commit those sins.">>

Right, I think sometimes we all judge far too quickly and not carefully enough. I know I have been guilty of that.

As we find component designs that sound superior to other designs, the need for sonic correction is lessened. Then I think our results with testing capacitors, or other parts, are worth more.

Unfortunately, most don't have the facilities or expertise to test parts, so they have to go by others reviews. I am just saying to be careful when reading posts, articles etc.

[jeffreybehr]

Steve, I think we're quite well in sync here.  I assumed he had a relatively balanced system to start with.  I too disagree heartily with selecting components with substantial colorations just to offset other significant colorations.  I once called it something like selecting ICs as tone controls.

I try very hard to test carefully.  Not being a GEA*, I have to work hard at determining some differences.  My philosophy is that if I can hear the addition (not substitution) to my system, I don't want it!


*  That note in my signiture is more true than I like!