The Crossover - Brain of your Loudspeaker System

[Loren42]

Nope. Sorry. Not even close. This is not to say you cannot take good drivers, and a good box and screw it all up by building your own crossover without measurement equipment and a clue as to what you are doing. What is does mean is that designing a good high quality driver is both far more difficult and far more expensive to do than designing a good crossover. I have 34 years experience building crossovers, and only 26 years experience designing drivers and systems, but yeah, I am sure about this.

Paul,

Sorry to disagree, but I think you misunderstood my point.

I won't challenge your experience in the matter. Clearly, 26 years and 34 years of experience is a noble accumulation of knowledge and I respect that.

However, the original discussion context really did not focus on driver design, but crossover design and in specific, the degree of variance that the design and implementation a crossover can have on the total quality of a speaker system.

While it is arguable that one can totally screw up the creation of a driver, the premise that I made was that the driver and cabinet design were of reasonable quality. The exact quote was, "As long as even a fair job is done with driver selection and cabinet design."

If you place lower limits as to how lousy a driver you can use and how horrible a cabinet design you can make, then the next variable in the equation that can exhibit the widest degree of variability on the system Q would be a crossover network (assuming no limits on how lousy or good you make that crossover).

I understand that "fair job" is a subjective term, but my point was that the design and execution of a crossover in the total system Q is of very high importance and is something many people overlook in the process. Would you not agree?

In the DIY world most builders put a lot more effort into driver selection (probably above all other criteria) and placing that above all other factors, neglecting a very important element of the system; the crossover.

It looks like you are a recent recruit to the forum. Welcome to the fold and your experience and knowledge in the field will be a very welcome asset to the community.

 

[Dr. Noisewater]

Dope article.

 

[lsiberian]

Unfortunately you can't really compare the two speakers in such a cut and dry method. As stated already other factors are definitely at play when comparing two speakers of different make, design, and parts. A good speaker is the sum of it's parts. A great cabinet with great drivers and a mediocre x-over will likely have less than outstanding performance. On the other hand putting an excellent x-over with bad/good drivers in a good/bad (role reversal) will result in poor performance again.

It's very tough to say that one part takes prescedence over the other because each component is very crucial to the end performance the product creates. So I would argue that the x-over, cabinet, and drivers are all of equal importance with none making the bigger difference, since if one is faulty it's all faulty.

Excellent post and point

Everything is important. The cabinet is most important IMO assuming one builds a decent crossover or has an EQ to help smooth things out. Even cheap drivers in a great cabinet sound terrific. Take the Recession Buster Kit as an example.

But that's not to say other parts aren't important, but if you want to get the most out of your speakers build them well. You can always use an active crossover to do the work if necessary.

 

[Paul_Apollonio]

Authorship

It was not intended to be mutually exclusive, though I can see how you might believe this to be fact, regardless if your chicken feather soup could or could not be recreated with a poor crossover coupled with good drivers.

I not only refuse to answer this question due to the rhetorical implication, but because I do not intend to continue with what I find to be both a contentious, and most of all, presumptuous disposition.

I could explain myself further, much further in fact, but it would be a disservice to this forum in my eyes, if not downright annoying, and hence my wish to discontinue.

EDIT: I wanted to add that there are a number of times that I do direct questions to experts, and I do not always get a reply. I've even directly asked the author of this article, Gene, questions for which I've never received a reply. There are other examples, even just at this forum, for which I did not receive a reply from one who is widely accepted to be an expert. I assume no reason for the lack of response, nor am I offended. I am not presumptuous.

Dear Jostenmeat,

I am the author of the article about crossovers. Gene, my boss, contributed some section titles and a photograph. He also proofreads it as a peer. If you still have questions, or if anyone does, you stand the best chance of getting those questions answered by emailing me directly. papollonio@yahoo.com

By the way, the phrase is "chicken soup from chicken feathers". (I do not make chicken feathers at all, I make the soup). This phrase was from the late great Laker announcer Chick Hern, in reference to the Clippers coach trying to create a winning Ball team.

I am glad we are both of like mind in not wanting to get into a contest here in the forum. It was wrong of me to mock your phrase "some people".

 

[croseiv]

Unfortunately you can't really compare the two speakers in such a cut and dry method. As stated already other factors are definitely at play when comparing two speakers of different make, design, and parts. A good speaker is the sum of it's parts. A great cabinet with great drivers and a mediocre x-over will likely have less than outstanding performance. On the other hand putting an excellent x-over with bad/good drivers in a good/bad (role reversal) will result in poor performance again.

It's very tough to say that one part takes prescedence over the other because each component is very crucial to the end performance the product creates. So I would argue that the x-over, cabinet, and drivers are all of equal importance with none making the bigger difference, since if one is faulty it's all faulty.

Okay, say the speakers, and cabinet are obviously of good quality, but the crossover seems to be the weakest link. Would any audible benefit come from building a better more robust crossover (I think the answer is yes)?

 

[R-Carpenter]

Okay, say the speakers, and cabinet are obviously of good quality, but the crossover seems to be the weakest link. Would any audible benefit come from building a better more robust crossover (I think the answer is yes)?

Robust is not necessarily correct word. A crossover which would give you flattest frequency response or I should say a USER DEFINED response is better, yes, it would turn things to your liking. A better crossover parts by themselves are most likely is just a waste of money.
On the subject, boutique crossover parts and its advantages and differences have been debated for a longest time and started a few flame wars. My opinion is that they are not going to hurt anything but bank account.

 

[Paul_Apollonio]

Relative importance of drivers vs crossover

Paul,

Sorry to disagree, but I think you misunderstood my point.

I won't challenge your experience in the matter. Clearly, 26 years and 34 years of experience is a noble accumulation of knowledge and I respect that.

However, the original discussion context really did not focus on driver design, but crossover design and in specific, the degree of variance that the design and implementation a crossover can have on the total quality of a speaker system.

While it is arguable that one can totally screw up the creation of a driver, the premise that I made was that the driver and cabinet design were of reasonable quality. The exact quote was, "As long as even a fair job is done with driver selection and cabinet design."

If you place lower limits as to how lousy a driver you can use and how horrible a cabinet design you can make, then the next variable in the equation that can exhibit the widest degree of variability on the system Q would be a crossover network (assuming no limits on how lousy or good you make that crossover).

I understand that "fair job" is a subjective term, but my point was that the design and execution of a crossover in the total system Q is of very high importance and is something many people overlook in the process. Would you not agree?

In the DIY world most builders put a lot more effort into driver selection (probably above all other criteria) and placing that above all other factors, neglecting a very important element of the system; the crossover.

It looks like you are a recent recruit to the forum. Welcome to the fold and your experience and knowledge in the field will be a very welcome asset to the community.

Dear Loren42,

First of all, thanks for the kind words. I appreciate the welcome! Actually my first experience on the blog was back when I wrote the articles on power handling (3 parts). I understand your point better now, but I still take issue with one sentence in your post I was in disagreement with.
"As long as even a fair job is done with driver selection and cabinet design, the crossover is the single most important element of the system."
Whether you do a good or bad job in selecting the drivers, they are STILL the most important element of the system (not the crossover); even if the biggest improvement one can make to a system without changing the drivers or box is to fix their crossover. Drivers will predominate the sound you hear over many octaves, while the crossover will at most destroy one to two octaves (albeit the octaves which are likely to be the most important). If your point was a lot of audiophiles pick very good drivers, and then completely underestimate the importance of the crossover in the determination of the final sound, we agree 100% on this. I believe this is from lack of information and experience, and that is the primary reason I chose to pick this subject to write on. I intend to focus on this important point of yours in the next article written on this subject, and illustrate with a real world example.

There seems to be some confusion as to authorship on the crossover article. Perhaps I need to ask Gene to get top billing next time.

 

[croseiv]

Robust is not necessarily correct word. A crossover which would give you flattest frequency response or I should say a USER DEFINED response is better, yes, it would turn things to your liking. A better crossover parts by themselves are most likely is just a waste of money.
On the subject, boutique crossover parts and its advantages and differences have been debated for a longest time and started a few flame wars. My opinion is that they are not going to hurt anything but bank account.

I was not refering to botique parts. I merely meant a better than what is already in there design. But that can be interpreted as "different than" what is in there too. As you say though, being able have a user defined response seems to be the real ticket.

 

[TLS Guy]

Good speakers are the sum of their parts.

The easiest part is driver alignment to the box. Making a good enclosure takes skill.

However no speaker will sound good if it does not have a nice flat midband response, on and off axis. This achieved by the summed slopes of acoustic and electrical roll offs

Selecting drivers that will work well together is key to a good design. Thinking that you can make any combination of drivers work well with high order crossovers is not true, but a common mistake and misconception.

Crossovers can not be cookie cut. There is no substitute for experience in crossover design. The crossover is where the designer learns to get the voicing and sound stage he wants.

Components have to be of adequate quality, but exotics are not neessary.

One word of warning, crossovers can often be downgraded by substituting inductors with heavier gauge wire, and therefor lower resistance. The specific resistance of the inductor selected is often vital to the performance of the crossover.

 

[croseiv]

Good speakers are the sum of their parts.

The easiest part is driver alignment to the box. Making a good enclosure takes skill.

However no speaker will sound good if it does not have a nice flat midband response, on and off axis. This achieved by the summed slopes of acoustic and electrical roll offs

Selecting drivers that will work well together is key to a good design. Thinking that you can make any combination of drivers work well with high order crossovers is not true, but a common mistake and misconception.

Crossovers can not be cookie cut. There is no substitute for experience in crossover design. The crossover is where the designer learns to get the voicing and sound stage he wants.

Components have to be of adequate quality, but exotics are not neessary.

One word of warning, crossovers can often be downgraded by substituting inductors with heavier gauge wire, and therefor lower resistance. The specific resistance of the inductor selected is often vital to the performance of the crossover.

Thanks for your input TLS Guy. I have seriously been considering "upgrading" or redesigning my crossovers in my Polk Audio RT7s as an "experiment" for fun to see if I can glean any improvment in thir sound quality. Overall FR is 35Hz-26KHz with the crossover frequency at 3.2 Khz. Do you have any recommendations?

 

[TLS Guy]

Thanks for your input TLS Guy. I have seriously been considering "upgrading" or redesigning my crossovers in my Polk Audio RT7s as an "experiment" for fun to see if I can glean any improvment in thir sound quality. Overall FR is 35Hz-26KHz with the crossover frequency at 3.2 Khz. Do you have any recommendations?

I have no recommendations, as I have no TS parameters or reliable acoustic data for the drivers involved. All of that information is essential before any design can begin. As always, you best bet is a design from scratch.

 

[croseiv]

I have no recommendations, as I have no TS parameters or reliable acoustic data for the drivers involved. All of that information is essential before any design can begin. As always, you best bet is a design from scratch.

Undoubtedly those would be under the lock and key of the original manufacturer... I guess I was thinking it would be much simpler, like just having a look at the original crossover and going from there. I have learned alot from this thread though. For example, when you say T/S parameters you are reffering the Thiele/Small paramaters. I was over at wikipedia having a look at those. Fairly complex and challenging to measure.

 

[TLS Guy]

Undoubtedly those would be under the lock and key of the original manufacturer... I guess I was thinking it would be much simpler, like just having a look at the original crossover and going from there. I have learned alot from this thread though. For example, when you say T/S parameters you are reffering the Thiele/Small paramaters. I was over at wikipedia having a look at those. Fairly complex and challenging to measure.

Not with this!

 

[croseiv]

Not with this!

Holy Moly! Awesome. I'll be ordering one for sure (next week)...Sweet. Thanks TLS Guy.

 

[Loren42]

Not with this!

Do you know how well that tool works by any chance?

Any personal experience?

TNX

 

[croseiv]

Do you know how well that tool works by any chance?

Any personal experience?

TNX

The reviews at PE seem to imply it's quite good. I think it should be more than adequate for what I'm trying to do.

 

[avaserfi]

I have one. With proper use it is an accurate device.

 

[Alex2507]

Not with this!

That's a great find with the scale included. I saw this before but thought the sale expired or maybe I just lost track of it. Now I could have yet another 'thing' that I don't know how to use.

 

[R-Carpenter]

Do you know how well that tool works by any chance?

Any personal experience?

TNX

It is very accurate. I've sold mine only because I have Sound Easy set up.

 

[R-Carpenter]

Undoubtedly those would be under the lock and key of the original manufacturer... I guess I was thinking it would be much simpler, like just having a look at the original crossover and going from there. I have learned alot from this thread though. For example, when you say T/S parameters you are reffering the Thiele/Small paramaters. I was over at wikipedia having a look at those. Fairly complex and challenging to measure.

In order to redesign the crossover you will need T/S, FR, Impedance, distortion, and preferably Cumulative spectral decay. After the initial crossover design and trial you need to verify performance and take a multitude of acoustic measurements. Then readjust the crossover and measure again. If you are not planing on doing all this I'd suggest to stay away from changing the crossover. The manufacturer is already ball parked it to the variety of uses and you are very unlikely to improve things unless you take your time and perform all the research involved.
The driver data can be measured very easily so no mater how big is the lock, the speakers are out there and if someone wants to, it can be done.
Is there a specific characteristic in the sound of your speakers that you don't like?
Oh BTW, Gluing anything or sticking crap to the driver's diaphragm for Delta mass testing is not very advisable nor accurate. A test box is recommended.