[aggielaw]

I've been told "yes" and "no" by a few people whose opinions I respect in the audio profession.

I've heard all parts matter, I've heard only the caps matter (for you can't hear a difference in the quality of resistors and other parts) and that the parts basically don't matter because the difference between, say, Sonicaps and top-of-the-line Mundorf (or whatever other high-end brands are out there) is almost indiscernable even in a high end system.

Your thoughts are greatly appreciated.  I have limited resources to spend on an upcoming project, and don't want to dole out close to $1000 for the very best parts if they aren't going to make much difference.

Although I ask the question in the absolute sense, you may find it helpful to also tell me if you think the parts will matter in my personal system.  

The speakers to have the XO modded are nOrh 9.0's.  I'm feeding them with a Cary 303/300 CDP and McCormack DNA-125 Platinum amp.  Ridge Street Poeima! cabling all around.

Thanks!
Howard

[Marbles]

IMO you should at least go with Sonicaps, Mills resisters and a nice air core inductor in those XO's.

The DM XO is a second order and doesn't have that many parts that you should skimp.

That is if you are going to go to the DM XO in the big 9.0's.

Just upgrading parts from the original XO probably won't get you much as good parts were already used in what I consider a mediocre at best circuit.

[ScottMayo]

I've been told "yes" and "no" by a few people whose opinions I respect in the audio profession.

I've heard all parts matter, I've heard only the caps matter (for you can't hear a difference in the quality of resistors and other parts) and that the parts basically don't matter because the difference between, say, Sonicaps and top-of-the-line Mundorf (or whatever other high-end brands are out there) is almost indiscernable even in a high end system.

Your thoughts are greatly appreciated.  I have limited ...

I can't talk about your particular speaker, but yes, the parts matter. Speaker crossovers do critical work. If you have two speakers, you want them to behave the same way - and they only way to do that is to build with parts that are within tight and identical tolerances. You won't get that from Radio Shack.

I'd go through the exercise of hand testing components to get close to identical values (for things like resistors, it's easy); I'd buy good caps, and I'd probably hand-wind inductors. Trimmers for resistive values might be a good idea, too. Parts age and change.

Mind you, I don't do this. I used to play with crossovers, but these days I'd much rather buy from a builder I trust. But IF I were building my own today, I'd splash out for the tighter tolerances.

[Bill Baker]

Hmmmm, tricky area. The crossover parts matter in ways you probably don't realize. Putting their "quality" aside, you really need to consider other factors such as their electrical characteristics.

 For example, simply replacing an inductor can have many different outcomes. For a speaker that was properly designed from the beginning, even things such as the DCR of an inductor were taken into account as this is a series resistance that the driver sees. For example, if you were to replace an inductor with one of higher DCR from the original, it [could] be like adding in a resistor.

 With capacitors, the size and type of dialectric can even play a toll as they can also add resistance, although not much, it is still there.

 Even internal wiring can effect the sound. Not due to the quality per-say but rather how thick it is and what it is "adding" to the circuit in conjunction with all the other components you choose to replace.

 These things must be taken into consideration. I am not saying that upgrading x-over components will not yield good results rather the changes you hear may be caused but things you didn't realize.

 So yes, crossover parts matter.

[andyr]

I've been told "yes" and "no" by a few people whose opinions I respect in the audio profession.

I've heard all parts matter, I've heard only the caps matter (for you can't hear a difference in the quality of resistors and other parts) and that the parts basically don't matter because the difference between, say, Sonicaps and top-of-the-line Mundorf (or whatever other high-end brands are out there) is almost indiscernable even in a high end system.

Your thoughts are greatly appreciated.  I have limited ...

Hi Howard,

I'm amazed anyone has told you "no" ... although possibly there are speaker manufacturers who use the best possible parts in their crossovers.

As a nOrh owner, you wouldn't haunt the Maggie Asylum ... but if you did, you'd see post after post about people who've replaced the crap stock components Magnepan uses with things like Jensen PIO or Mundorf caps and Goertz or North Creek inductors.  And these transform the speaker!

Regards,

Andy

[jeffreybehr]

Howard, yes, quality of parts in passive crossovers can matter, and increasing that quality can create QUITE an improvement in the sound.  Of course the improvement depends on several things such as the quality of what's there now, the quality of what you put in, and your hearing acuity or 'golden-earededness'.  

For caps in series with the signal, I too would use SoniCaps.  Jeff Glowacki seems to have done a VERY good job of designing a fine-sounding cap at low prices.  If your crossover has large shunt caps (ie, in parallel, not in series, with the driver and the price of SoniCaps is getting too high for you, consider the SCR/Axon caps (from Michael Percy, here http://www.percyaudio.com/Catalog.pdf ) that are much less expensive but still better sounding than 'lytics.  I'd replace all resistors with Mills 12W. noninductive wirewounds.  Since total cost is a serious consideration for you, I would not replace inductors unless there are iron-core inductors in the midrange or tweeter circuits.  If all are air-core, I'd leave as is.  Some of my efforts at improving passive crossovers and eq. circuits and using original inductors can be seen here http://community.webshots.com/album/238619181WkZSKm ; these series networks include SoniCap-1s bypassed with teflon-film SoniCap Platinums.


Good luck.  Let us know how it turned out.

Here http://www.soniccraft.com/products/capacitors/film.htm is Sonic Craft.

[Bill Baker]

I would not replace inductors unless there are iron-core inductors in the midrange or tweeter circuits. If all are air-core, I'd leave as is.

 Very true. You will most likely find inductors in series with the low pass filter (mid/bass). As mentioned, get rid of any iron core if they are in there and go with a good air core or foil unit. The shunt capacitor (paralleled from + to -) in this circuit can be assembled with decent caps such as Dayton or Solen units and you can parallel caps to get the desire value. Again as mentioned, don't use lytics but if you must, make sure it is a non-polar design.

 The other consideration would be any filters in the upper end. I do not know what the crossover looks like in these speakers but if there is a Band Reject Filter located in the high pass filter (inductor, capacitor and resistor all in parallel with each other), this is in series with the signal so it is important to use good quality components here also.

 Don't let all these writings scare you. Just going with better quality components will open things up for you if run of the mill components were used originally.

 Enjoy the project and more importantly, enjoy the music!

[Kulamata]

I'd second (fourth? fifth?) the comments above. These comments do not apply to any specific speaker system, but are just what I've found over the years.  My bias is to have acoustic instruments sound live; clear, warm, and without the false clarity of high frequency emphasis.

The better the drivers and the associated system (and your ears' training) the bigger the difference crossover parts make.  I use Michael Percy a lot. North Creek Music, Sonic Craft, Marsh (at capacitors.com)  and the Parts Connexion are all good places to know.  (The same Richard Marsh who wrote the original article on the audibility of dielectric absorption. )

Actually, I've done OK using steel Laminate cored coils for the very large series woofer inductor.  Their advantage is reasonable DC Resistance (and cost) for high values of inductance.  Their potential disadvantage is that, unlike air core coils, they can saturate, so one rated for much  more power than you will use is a good idea.

The Mills resistors work out VERY well for me.  I would not consider anything else now.

I used to use ChateauRoux/Solen/ SCR(?) capacitors, but now find the North Creek and Reliable caps smoother.  (There are many more capacitors out there that are highly regarded; but I am discussing only what I've heard/used.) In general, I would rate capacitors in the following order; metallized polypropylene, multi-tabbed metallized polyprop, polyprop/alum foil, multi-tabbed polyprop(PP)/alum foil, pp/tin foil, multi-tabbed pp/tin foil, and multi-tabbed polystyrene/tin foil.  I have never tried the copper foil and or teflon types, as they are way out of my price range.  Also in general, higher voltage units are usually nicer than lower voltage.  

Many consider only series elements to be important.  I feel that they are only a bit more important than shunt elements.  (Or equal.  Too close to be certain, and I don't really care, as I use the same for both. )

Wiring matters too.  I use the Cardas bare silver, or the Vampire cast copper.  I don't hear a difference between the various "audio" solders.  

It took me a long time to figure out that any listening comparisons without burning in are useless.  Awkwardly, the capacitors seem to need 40 hours or so. (The repeated breakins for comparison are not a lot of fun. )  I don't agree that things suddenly "snap in" after some magic period; rather to my ears, the breakin is exponential, with most in the first few hours, and something like forty hours to be essentially done.  I also learned to NEVER, EVER, use clip leads to try out a circuit change.  (I used the cheap store-boughten clip leads. Maybe with decent alligators and wire it would have been different. )

One other general rule; avoid components with steel leads like the plague.  The explanation that seems to make sense is that the hysteresis of the domains flipping is audible.