• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Expensive Crossover Parts – Do They Matter?

Mashcky

Active Member
Forum Donor
Joined
Feb 19, 2019
Messages
121
Likes
144
Location
Burlington, Vermont
I am shopping for speakers crossover parts and need an audiophile mythology check that I haven't been able to get anywhere else on the web. I have a specific question and a more general one.

First the specific question; can I replace a "non-inductive" resistor with one that is not labeled as such for a low-pass crossover network feeding the tweeter? Can I replace this with that? The only difference is that the cheaper part is labeled "non-inductive" and 2W less in power dissipation. My preliminary research has taught me that typical wire wound resistors behave like inductors by creating a magnetic field incidental to the way a resistor is wound that increases with frequency in an AC circuit. This increase in inductance will change the impedance of the resistor as frequency increases. What I want to know is; does this matter at all in an audio application?

The more general question is; are expensive crossover parts worth it? There are pricey heavy-gauge inductors, and expensive capacitors on the market that many believe to make a big difference. If these differences are measurable, are they even audible? Looking for some guidance as someone without an EE background.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
The Vishays are also available as non-inductive (NS). If the resistor is being used as a shunt rather than series, the inductance is unlikely to cause a significant issue.

More generally, a LOT of "premium" crossover parts will actually have inferior performance. Capacitors and inductors from boutiques claiming "hand wound" will not be as tight and consistent as mass produced parts from mainstream suppliers. As well, a competent crossover design will take into account the DCR of inductors and ESR of caps, so if you swap them for components having lower DCR or ESR, you will degrade the performance.

Capacitors especially seem to draw the mystics like turds draw flies.
 

JanRSmit

Member
Joined
Jan 21, 2019
Messages
54
Likes
21
I am shopping for speakers crossover parts and need an audiophile mythology check that I haven't been able to get anywhere else on the web. I have a specific question and a more general one.

First the specific question; can I replace a "non-inductive" resistor with one that is not labeled as such for a low-pass crossover network feeding the tweeter? Can I replace this with that? The only difference is that the cheaper part is labeled "non-inductive" and 2W less in power dissipation. My preliminary research has taught me that typical wire wound resistors behave like inductors by creating a magnetic field incidental to the way a resistor is wound that increases with frequency in an AC circuit. This increase in inductance will change the impedance of the resistor as frequency increases. What I want to know is; does this matter at all in an audio application?

The more general question is; are expensive crossover parts worth it? There are pricey heavy-gauge inductors, and expensive capacitors on the market that many believe to make a big difference. If these differences are measurable, are they even audible? Looking for some guidance as someone without an EE background.
I thing you mean the high-pass filter for the tweeter.
2watt diff in the spec does not matter. The inductance is also not of significance.
However the question is: way change that resistor? What do you want to achieve?
 
OP
Mashcky

Mashcky

Active Member
Forum Donor
Joined
Feb 19, 2019
Messages
121
Likes
144
Location
Burlington, Vermont
The Vishays are also available as non-inductive (NS). If the resistor is being used as a shunt rather than series, the inductance is unlikely to cause a significant issue.
I beginning to think this is more of a DIY Audio forums thread, so my apologies if this is not the best place. I'm making the design below. If I understand, the 60ohm resistor (found at the top of the diagram) is being used as a shunt, but the other ones are not?
ER18DXT-final-crossover.gif

More generally, a LOT of "premium" crossover parts will actually have inferior performance. Capacitors and inductors from boutiques claiming "hand wound" will not be as tight and consistent as mass produced parts from mainstream suppliers. As well, a competent crossover design will take into account the DCR of inductors and ESR of caps, so if you swap them for components having lower DCR or ESR, you will degrade the performance.
By Premium I meant more so spending $3 more dollars on a Mills resistors than a Vishay or Dayton Audio. Most of my cart consists of ESRE inductors and Solen capacitors – all still mass produced parts I assume, so not so much in the boutique market. I supposed there are more midrange components simply with higher tolerances and higher watt ratings.
 
OP
Mashcky

Mashcky

Active Member
Forum Donor
Joined
Feb 19, 2019
Messages
121
Likes
144
Location
Burlington, Vermont
I thing you mean the high-pass filter for the tweeter.
2watt diff in the spec does not matter. The inductance is also not of significance.
However the question is: way change that resistor? What do you want to achieve?
Yes, high-pass. I think my logic was that the lower frequencies were the ones being "passed" on :facepalm:.

The primary reason for changing the resistor is primarily to get one that is in stock because I can't find four 3.5ohm 1% tolerance, 10-12W resistors anywhere. Secondarily, I started shopping around I realized how much prices differ depending on specifications and want to know where a builder can save money.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
I beginning to think this is more of a DIY Audio forums thread, so my apologies if this is not the best place.

You'll get fewer answers here, but they'll be far less voodoo-packed. :D

The inductance of that resistor will not be important in that position. Likewise the slightly lower power rating. It's acting as a minor shunt across an inductor which has much greater inductance, likely to provide some minor frequency response tailoring.

Solen caps are OK. The only one with any criticality is C2, so you'll want to use a polypropylene cap in that position. The others can be polyester or even bipolar electrolytic without any issues. They're only used for impedance compensation.

Don't try to get fancy on the inductors because their DCR is either taken into account (L1) or is swamped by other resistances. The only tweak I'd think about is potting the coils to reduce susceptibility to vibration, and that's pretty much gilding the lily.
 

Eirikur

Senior Member
Joined
Jul 9, 2019
Messages
318
Likes
509
The more general question is; are expensive crossover parts worth it? There are pricey heavy-gauge inductors, and expensive capacitors on the market that many believe to make a big difference. If these differences are measurable, are they even audible? Looking for some guidance as someone without an EE background.
When I need a part I try to get the cheapest part with low (manufacturing) tolerance, e.g. I overspent on the resistors to get Jantzen Superes because they are specced at 1% tolerance and have low inductance. This means I can simply order a pair for both speakers and be pretty sure they are sufficiently matched (all 5 pairs were spot on actually).
Similarly for capacitors: again Jantzen Cross Cap as they also happened to have all the exact values I needed (my alternative would have been Audyn Q4) and they are all rated 5% - again an exact match. Next step up to 3% (like Mundorf Mcap EVO) almost triples the price per part so I ABX-ed it with my wallet and decided 5% is good enough.

A much cheaper option is bipolar elcos but the actual values are a bit of a casino (a spare Mundorf ECAP50 47uF measured 50.2uF!). In the longer run they are also seem unreliable, that is to say, the old ones I replaced were already 25% off their rating!
The new PP caps will stay much closer to their initial performance I expect.
 

syn08

Senior Member
Joined
Aug 9, 2019
Messages
376
Likes
461
Location
Toronto, Canada
As well, a competent crossover design will take into account the DCR of inductors and ESR of caps, so if you swap them for components having lower DCR or ESR, you will degrade the performance.

I can see that for inductors, but for capacitors? Unless a crossover is designed with poor quality bipolar electrolytics (which IMO should be avoided as much as possible, anyway), is there any relevant difference between decent film type capacitors ESR? I wouldn't think so, although I can imagine some boutique capacitors advertising as "avoid resonances by internally amortizing".
 

Speedskater

Major Contributor
Joined
Mar 5, 2016
Messages
1,610
Likes
1,333
Location
Cleveland, Ohio USA
Are you rebuilding an existing crossover or designing a new one?
Loudspeaker designers often take into account the imperfections of the parts that they use. So replacing an imperfect part with a perfect one, in effect revoices the loudspeaker.
 

syn08

Senior Member
Joined
Aug 9, 2019
Messages
376
Likes
461
Location
Toronto, Canada
So replacing an imperfect part with a perfect one, in effect revoices the loudspeaker.

Was this ever, in general, proven true?

I would think that a filter (which is what crossovers are) characteristics depend so strongly (as to make the changes audible) on the parasitic elements of capacitors is a poor design. To add insult to injury the capacitors ESR is barely specified to include tolerances (which would be needed for a competent design).

Not saying it is impossible, but wondering how often this may happen.
 
OP
Mashcky

Mashcky

Active Member
Forum Donor
Joined
Feb 19, 2019
Messages
121
Likes
144
Location
Burlington, Vermont
You'll get fewer answers here, but they'll be far less voodoo-packed. :D
Wonderful then!
The inductance of that resistor will not be important in that position. Likewise the slightly lower power rating. It's acting as a minor shunt across an inductor which has much greater inductance, likely to provide some minor frequency response tailoring.
Sadly, you'll have to explain this to me like I'm five... So R5 is not important because its being used as a shunt, but the other resistors inductance could matter? I'm also curious about very slight changes in impedance from the stated values. For example, Solen.ca has resistors rated 2.4ohm rather than 2.5, and 3.6ohm instead of 3.5, both with 2% tolerances. All else the same, is this a permissible substitution?
Solen caps are OK. The only one with any criticality is C2, so you'll want to use a polypropylene cap in that position. The others can be polyester or even bipolar electrolytic without any issues. They're only used for impedance compensation.
Good to know, this should save me even more. Is there any downside to lower quality capacitors? Do they not last as long?
Don't try to get fancy on the inductors because their DCR is either taken into account (L1) or is swamped by other resistances. The only tweak I'd think about is potting the coils to reduce susceptibility to vibration, and that's pretty much gilding the lily.
When you say "potting the coils" you mean just fixing them to a board so they can't move around?
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
In order:

R5 is specifically being used as a shunt around an inductor. So as long as its inductance is really small compared to the inductor it's shunting, it won't matter. Typical inductances of power resistors rin a few microhenries. The inductor is a thousand times bigger. So the resistor's inductance won't matter. Likewise, resistor values can certain vary slightly- the variation from speaker driver to driver will swamp that. Try to make them close from side to side, though, IOW, don't use a 2.6 on one side and a 2.4 on the other. Generally, if you get resistors out of the same batch from a reputable manufacturer, they'll be close to one another.

Polyesters and polypropylene caps will outlive you, assuming again that you get a mass produced brand like Panasonic or Vishay or Wima or the like (if they're listed at Digikey or Mouser, you're likely safe, they don't do that voodoo).

Potting means essentially encapsulating the coils. It's a bit of an involved process where you dunk them in a liquid resin, then let it cure. Like I said, gilding the lily, so don't sweat it.
 

SIY

Grand Contributor
Technical Expert
Joined
Apr 6, 2018
Messages
10,383
Likes
24,749
Location
Alfred, NY
I can see that for inductors, but for capacitors? Unless a crossover is designed with poor quality bipolar electrolytics (which IMO should be avoided as much as possible, anyway), is there any relevant difference between decent film type capacitors ESR? I wouldn't think so, although I can imagine some boutique capacitors advertising as "avoid resonances by internally amortizing".

The biggest difference between film caps is usually foil vs metallized. And in most cases not significant, but I can think of a few situations where it could be.

I did have a negative experience replacing a bipolar electrolytic with a film cap- this was in the tweeter feed to some Magneplanars I used to own, right after the Jung/Marsh article came out (I was younger and more gullible then). The difference in ESR was enough to make the sound too bright up top. I learned my lesson about ESR and not to trust people like Marsh.
 
OP
Mashcky

Mashcky

Active Member
Forum Donor
Joined
Feb 19, 2019
Messages
121
Likes
144
Location
Burlington, Vermont
In order:
Likewise, resistor values can certain vary slightly- the variation from speaker driver to driver will swamp that. Try to make them close from side to side, though, IOW, don't use a 2.6 on one side and a 2.4 on the other. Generally, if you get resistors out of the same batch from a reputable manufacturer, they'll be close to one another.
Thanks so much! Taking your suggestions I saved a lot of money and more importantly, I'll actually be able to start building this weekend as planned.

My high level take-aways so far are:
  • Very small value changes don't matter for resistors, capacitors, or inductors as long as they are consistent.
  • Depending on the design and placement, it could be more or less crucial to have a high quality component (such as the relative inductance of the R5 resistor shunting a much higher inductance inductor).
  • Boutique caps aren't worth it, but higher tolerances generally are.
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
617
Likes
1,355
Solen caps are OK

I took my 2 dozen Solen fast caps manufactured over a span of decades and using mlssa, tested their reactive impedance. Minus the esr, they were dead on capacitive (r=1.0) 20 Hz to 20 kHz except two that served abusive duty in the rear deck of my Pinto (ya ya) in the early 80s.

They get a bad rap but Ive never been able to measure any real distortion off them, and they are as close to perfect caps in xfer function as you'd want.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,834
Likes
16,496
Location
Monument, CO
I agree with all the rest, natch. I'd like to spin off a few other thoughts based on the schematic below as presented by @Mashcky. N.B. I am not a speaker designer so take with a block of salt; these are observations based on (sometimes painful) previous experience and analyses done years ago when I was piddling with speakers. I made, and helped others make, many, many crossover mods back 20+ years ago when it was all the rage. I measured and conducted numerous DBTs. With very few exceptions none of the grandiose high-dollar mods were audible. Some were measurable in time and/or frequency domain. The biggest changes were with speakers having extreme impedance excursions or just very low impedances to begin with as one might expect.
  • The main crossover frequencies look to be set by L1 and C2. The other circuits are series resonators to shape the response, either to compensate for driver impedances, or cabinet issues like baffle step, or both. Or something else. But, I did not simulate (let alone hand calculate) the response; it may look second-order'ish at some frequencies.
  • Crossovers are low-order things and pretty tolerant of things like parasitics in components, like ESR and DCR., within reason. A small shift in crossover response will occur with changes in components but I'd be surprised they were audible. The times a component swap was audible was IME either a very special case or a mistake. Usually the latter.
  • Damping factor, i.e. amplifier output impedance, matters more with some speakers than others. In this case, note L1's DCR is 0.6 ohms and there is 2.5 ohms (R1) in series with the tweeter. So you have 0.6 ohms in series with that 8-ohm woofer; if the amplifier were perfect and cables superconducting, the effecting damping factor seen by the woofer is 8/0.6 = 13.333. Kind of puts the marketing for huge damping factors into perspective.
  • Often what matters more than absolute component values is matching values between two crossovers so a pair of speakers is well-matched. Of course there is driver variance to contend with...
  • Big non-polar electrolytics get a bad rap in the audio press but by and large have minimal impact used in a speaker crossover. Ceramic capacitors (doubt you'd use them in this application but) I tend to avoid as they tend to change in value with voltage and are vibration-sensitive. Audible? Who knows...
  • Placing inductors at right angles to reduce coupling in your layout can actually make a difference. Small, but measurable, and maybe audible. Maybe. Capacitors don't tend to couple as much IME but I wouldn't abut them; inductor fields will couple if they are mounted side-by-side.
  • Pay attention to current flow (in this and all things). I have seen some very bad layouts that modulated the tweeter with woofer signal due to poor ground layout. Again, I could not say how audible, but it was certainly measurable. In the primordial past I would bring in the signal and ground then split the ground plane on a PCB (or use separate ground bus if point-to-point) so the woofer side signal/grounds flowed together and tweeter signal/grounds together in their own little islands; they met only at the input connector. Frequency is low but power can be high.
FWIWFM - Don
View attachment 31238
 

syn08

Senior Member
Joined
Aug 9, 2019
Messages
376
Likes
461
Location
Toronto, Canada
Big non-polar electrolytics get a bad rap in the audio press but by and large have minimal impact used in a speaker crossover. Ceramic capacitors (doubt you'd use them in this application but) I tend to avoid as they tend to change in value with voltage and are vibration-sensitive. Audible? Who knows...

Polar or not, electrolytics dry out and change value in time. They also have larger tolerances (+/-20%) compared to good film caps. That all that myself can complain about them.

Ceramics, only high K devices behaving as you describe. There are no NP0 ceramics available in the value ranges for crossovers (at least not without a big cost hit) ceramics are out of discussion indeed.
 

pozz

Слава Україні
Forum Donor
Editor
Joined
May 21, 2019
Messages
4,036
Likes
6,825
Bruce Hofer from Audio Precision has a good general presentation on sources of distortion in circuit design. The second half focuses on parts specifically. It's fairly interesting and gives his views on varying manufacturing quality. Might be of interest if you're trying to track down the reasons why the impacts are what they are.

Here's a link to the accompanying article (you'll need to register to download it): https://www.ap.com/download/designing-for-ultra-low-thdn-2/

 

captain paranoia

Active Member
Joined
Feb 9, 2018
Messages
293
Likes
218
I can't find four 3.5ohm 1% tolerance, 10-12W resistors anywhere

I see no reason why you can't parallel up larger value, lower power resistors; 5*18 ohm and a 130 ohm will come pretty close (3.503 ohm), and those are standard E24 series values. You'd only need 3W for each of the 18ohm.
 
OP
Mashcky

Mashcky

Active Member
Forum Donor
Joined
Feb 19, 2019
Messages
121
Likes
144
Location
Burlington, Vermont
I see no reason why you can't parallel up larger value, lower power resistors; 5*18 ohm and a 130 ohm will come pretty close (3.503 ohm), and those are standard E24 series values. You'd only need 3W for each of the 18ohm.
The only reason was cost and a busier crossover board. I thought about two 7 ohm resistors but that doubled the cost. I ultimately found some that were close enough to 3.5 ohm (3.6) and was comfortable enough that it would not make a difference in sound.

As an aside – would tolerance matter less with more resistors in series? My understanding is that tolerance is just margin of error, so the more resistors you have, the closer you get to their expected value even if each individual tolerance is high.
 
Top Bottom