Do High End Capacitors Make A Difference In The Crossover?

[THD+N]

Any thoughts on why such high voltage ratings for the caps?
A 450V cap in a crossover seems like tremendous overkill to me.

Probably because they are off the shelf parts and using something with higher specs than you need so they run conservatively fits into the marketing of the product. Nobody uses those expensive parts without bragging about it.

I don't buy the boutique cap sound difference. Some do and for those who do, carry on! With less money.

 

[mrehorst]

I think it's often more a function of the materials used than a design choice. But those voltages are also totally appropriate for a tube amp, so maybe it's how they design the whole line.

Cap voltage ratings are primarily dependent on the thickness of the film used to make them. Why do they use such thick film when much thinner film could be used and result in a physically smaller cap at a lower voltage rating? Maybe it's easier to handle the thicker film (less likely to get torn?). Maybe the thicker film is cheaper than the thinner film.
Or maybe the cap manufacturer assumes that the caps might be used in a tube amp and require a high voltage rating. The same high voltage rated cap can be used in either an amp or a crossover, so why make two caps, a low voltage one and a high voltage one, when the high voltage rated cap will serve both purposes?

 

[mrehorst]

Probably because they are off the shelf parts and using something with higher specs than you need so they run conservatively fits into the marketing of the product. Nobody uses those expensive parts without bragging about it.

I don't buy the boutique cap sound difference. Some do and for those who do, carry on! With less money.

I'm not really a believer either, but B&W definitely doesn't shout about the crossover parts like they do the driver construction they use. I found one online review that showed a picture of the bass crossover (which doesn't use very high-grade parts) and no mention or photos of the mid/high crossover. I haven't seen anything else that shows either crossover or mentions the parts used.

Maybe they don't shout about the parts used in the mid/high xover because it gives them the option to switch to cheaper parts in the future.

 

[StimpyWan]

I'm not really a believer either, but B&W definitely doesn't shout about the crossover parts like they do the driver construction they use. I found one online review that showed a picture of the bass crossover (which doesn't use very high-grade parts) and no mention or photos of the mid/high crossover. I haven't seen anything else that shows either crossover or mentions the parts used.

Maybe they don't shout about the parts used in the mid/high xover because it gives them the option to switch to cheaper parts in the future.

Here's a nice video posted by The Absolute Sound. It covers a tour showing B&W speaker production. The 5th section, "Driver Assembly & Inner Wiring" shows several crossover boards. Lot's of Mundorf Supreme and EVO series capacitors present.

 

[rothwellaudio]

I think it also matters where the cap is in the circuit.

A cap in series with a tweeter could have an impact on the signal passing through it.

But I don't see a cap to ground making any difference.

That's a fallacious argument. There is no hierarchy of importance for series or parallel components in any filter circuit. All the components matter, or else what are they doing there?

 

[Oasis8]

IMO it depends, as you still need to use what is suitable for the speakers and system as a whole.

So whether using what is deemed as lower end caps or higher end caps they still need to be chosen with care and consideration.

 

[toddalin]

If you make a second order Butterworth low pass filter, the signal passes through the choke on the way to the driver and also through the conjugate capacitor on the way to ground. If the value of the conjugate capacitor is too large for that of the choke/speaker combination, it will create a hump/peak just below the crossover frequency.

So if the presence this conjugate capacitor is making the hump, doesn't it stand to reason that it may also be adding something to the signal that does make it's way to the driver?

Another reason to use good conjugate caps is that the values tend to be MUCH closer to the printed value than electrolytics. And changes here can make one speaker sound bright if the cap is less than the printed value and the other sound dull if the cap is greater than the printed value. I can perceive differences when values exceed ~5%.

 

[knotscott]

I'm no expert on high-end audio, but I've soldered together lots of analog circuits and the very idea that any capacitor has a "sound" is as silly as the suggestion that some wire has a "sound." As long as they're functional and even remotely in spec, any difference in sound will be due to no one wanting to admit they just wasted a bunch of time and money

There was a claim not to be a high end audio expert, but nonetheless offered a very absolute summary of why people hear differences. I personally am not capable of determining what others hear, but I know what I hear. I've made hundreds, if not thousands of small changes over the decades, many of which were audible, some which are not, and many of which can take a long time to identify. Some differences I don't like, some are clearly better. The differences are often just a subtle refinement, but can surely be cumulative to become very easily noticeable. The clearer the system gets, the easier it is to identify subtle improvements. A system with 17 bottlenecks is much tougher to hear small differences with.

Doing nothing guarantees no changes in a system, which is fine if that's the objective, but can really limit potential if you're looking for next level performance without dropping major coin for new components.

 

[RVT1K]

That's a fallacious argument. There is no hierarchy of importance for series or parallel components in any filter circuit. All the components matter, or else what are they doing there?

Because I'm talking about potential "sound" of a cap and not just the slope of a given filter.

And I do mean potential. I've never done any kind of capacitor comparison but I am skeptical of the glaring differences some claim.

 

[rothwellaudio]

Because I'm talking about potential "sound" of a cap and not just the slope of a given filter.

If both series and parallel components have an impact on the slope of a filter, why would only series components have an impact on "sound"?

 

[StimpyWan]

My next recap, for my Martin Logan speakers. I will have to mount the new capacitors off the circuit board, due to replacement cap size...! Interesting to find a stock mix of cap types, which I assume are film for the series caps and NPE for the shunt caps...!

 

[FB night]

I'm no expert on high-end audio, but I've soldered together lots of analog circuits and the very idea that any capacitor has a "sound" is as silly as the suggestion that some wire has a "sound." As long as they're functional and even remotely in spec, any difference in sound will be due to no one wanting to admit they just wasted a bunch of time and money.

You are certainly entitled to your opinion

 

[toddalin]

My next recap, for my Martin Logan speakers. I will have to mount the new capacitors off the circuit board, due to replacement cap size...! Interesting to find a stock mix of cap types, which I assume are film for the series caps and NPE for the shunt caps...!
View attachment 3475491

Yes, most probably so. But this is in the interest of saving $$$ and space. It looks to me like a value of 180 mfd @100 volts and that would be huge and expensive. Look how big the 39 mfd @100 volt film cap is.

 

[StimpyWan]

Yes, most probably so. But this is in the interest of saving $$$ and space. It looks to me like a value of 180 mfd @100 volts and that would be huge and expensive. Look how big the 39 mfd @100 volt film cap is.

I agree. Solen seems to be the only one that makes a decent 180uF film cap. But, it's $52.81 at Parts Connexion. As such, I'll just probably leave the Bennic NPE in place, and add a bypass film cap to it, to improve it's discharge rate. Though, I do have a Mundorf 180 E-Cap that I could try too? Also, I have a Mundorf Supreme EVO Oil film for the 22uF, a 39uF ClarityCap CSA film, and a choice of either Mundorf M-Cap or Jantzen Cross Cap films for the 68uF. Now, I need to get my butt in gear and finish the work...!

 

[Nikko75]

That's a fallacious argument. There is no hierarchy of importance for series or parallel components in any filter circuit. All the components matter, or else what are they doing there?

This is correct.

The parallel components forms an integral part of the network. Without them, a second-order filter becomes first order network. A notch or contour filter no longer functions without the parallel components.

The series component forms the voltage divider portion with the parallel component, while the parallel is the current divider. If a part distorts while it is in the parallel position, it is subtractive and the inverse of that distortion is imposed on the signal to the tweeter, which is still distortion. Pull the parallel component out of the crossover and they will very quickly realise it resides in the signal path.