is the crossover the heart of all loudspeaker sound?

Hi o.j.

I believe that most if not all commercial crossovers can be improved upon though ultimately the reasons why super speakers sound better is down to cabinet construction and driver quality as well as the crossover.

For commercial reasons of price and availability the components in most crossovers are not very esoteric. If you wanted to replace the components with same value high quality ones this would certainly lead to a different sound, and in most cases a better one.

This should be approached with a great deal of care, especially if changing the design of the crossover, as it is not a task that is trivial to get right! Small changes in tweeter attenuation (for example) may measure only slightly differently yet the subjective sound will be noticeably changed - and possibly not to your liking.

Just my 2 cents...

Richard.

the cabinet is the main cost in speaker design. Different matrials are used instead of wood (carbon fibers, ceramic/mineral filled polymeric materials, aluminum, etc). It is all about vibration control.

Naim That Tune

with some experience of speaker building I would concur with what you say.


There have been (heated at times) discussions on this topic: my position is that all crossovers are a necessary evil.Without wishing to open another can of worms here, designers(as always) have to choose a balance of compromises, for example, the simplicity of a crossover in a two-way versus the power handling, reduced driver demands etc of a three- way with a more complex crossover.

The ultimate sound obtained ultimately depends on a whole host of subtle, complex and interdependent design choices.

I think that some might be surprised at the relatively small proportion of the overall price of commercial speakers that is allocated to the drive units, making it an attractive proposition to build one`s own, given the requisite time and woodworking skills

laurie S

quote:

---

unless of course you have a pair of Seventh Veil Nonsuch which don't use any crossover at all.

---

Tom


I have built a pair of speakers which use a single "full-range" driver.

although lacking extension in bass and treble,there was a directness and coherence about them that i found went a long way to compensate....so I fully appreciate the benefits of having no crossover

laurie S

Not sure:but is not the difference of an near wall design to an off wall design made in the crossover? often the housings are looking nearby the same and there are used same tweeters and woofers in on and off wall designs. (not talking about kan1 housing with the frame in the front or thiels housing )
O.J.

quote:

---

unless of course you have a pair of Seventh Veil Nonsuch which don't use any crossover at all.

---

IMHO when used with the sub-woofer this speaker would probably benefit from an active high pass filter. Obviously getting an active filter working well can be a whole other can of worms...

Even the Quad ESL63 and successors is only full-range in its centre panels, but it's probably the closest thing you'll hear to a single full range drive unit. I'm a big fan of the idea and the experience, not to mention the price.

Paul

quote:

---

Originally posted by Paul Ranson:
[QUOTE]IMHO when used with the sub-woofer this speaker would probably benefit from an active high pass filter. Obviously getting an active filter working well can be a whole other can of worms...

---

Good point, Paul. Before Seventh Veil's own filter was available I used an AR EC21 active crossover. This had a high-pass section. However, I found that the system sounded more open when I didn't use it. One tends to design for one's own taste in speakers. You may prefer it with the high-pass.

Returning to OJ's original question ... No, you won't make a bad sounding cheap speaker into a supersounding one just by changing the passive crossover, although you can make improvements.

Of course, all other things being equal (which they rarely are) the best crossover is NO crossover. However, there are dramatic differences in sound between the various full-range, no-crossover speakers out there, so there's more to it than just crossovers.

Enclosure design and construction makes a huge difference. This is not just for reasons of vibration control (although that's very important). The enclosure governs the bass performance of the speaker and also makes a big difference to internal standing waves and resonances. Diffraction characteristics also vary with the size and contour of the front baffle and, as Paul says, the crossover works in combination with this to vary the baffle step response.

The timbre and dispersion patterns of the drivers used are also significant, particularly as you go from one driver to another (another benefit of not using a crossover).

Quality of the drive-units used is, of course, paramount and even the internal cabling can have a significant effect.

I hope that I've clarified more than I've added to the confusion but when you look at the vast array of different designs out there you see that speaker design can be a complex business.

Steve Margolis
defy convention - make music

[This message was edited by 7V on Wed 28 April 2004 at 11:16.]

[This message was edited by 7V on Wed 28 April 2004 at 11:17.]

Hy james!Excuse my bad english but did i understand right that Bsc is done in the crossover?(as i supposed).
So it is (theoretical) possible to design on or of wall speakers only by changing frequqencie crossover specification.I know it
sounds more simple than it will be in reality,but maybe crossover is the logical starting point by building a better loudspeaker.An example:i have this vecteur premiere speaker with its five shilling tweeter
and i think there is a special design in crossover (woofer is not elctrical limited by crossover?)indeed a precise and correct sounding loudspeaker,but also
with aspecial formed housing front and therefore i cannot compare the sound to the shoebox form speakers.
O.J.

I am sure there are differences in the way the speaker cabinet interior is designed and damped for wall positioning vs free space positioning.
In the same way surely the "correct" approach to crossover implementation will be affected by enclosure type, (transmission line, ported or infinite baffle etc).
I would assume with all these factors in play it would be difficult to magically affect the performance of a particular speaker by crossover redesign.
Understand I am not talking about subtle improvements in inductor values or component quality, but a radical attempt to change a bad speaker into a superior one .
There is an old saying "You can't make a silk purse out of a sow's ear" which I would think applies to speakers as it does with most things.

regards
GEOFF

7V!Talking about broadband speakers i know only
the big housings with lowther speakers or similar.Think in this case the special housing
form works as a kind of "mechanical"crossover.
Do not know small (shoeboxsize) loudspeakers
without crossover .also tannoy coaxial or kef Q series has a crossover.Only thing i know that a piezo tweeter can work without crossover O.J.

Geoff!you said you are sure,there are mehanical
differences inside the housing like damping etc..
On the neat hompage they say you can use their
Speakers on and of wall.
Naim speakers are designed definitly for on wall use. (arriva too?)
Can anybody bring light on this damping difference inside on /off wall designs?
O.J.

Yes, the Lowther's are generally horn designs where the cabinet size and geometry determines the low frequency roll-off.

With sealed box and ported designs - I believe that 'ported' is a dirty word on this forum and Seventh Veil doesn't use them either - the internal volume of the cabinet, together with the bass driver used, will determine the shape (Q) and extension of the bass response. It's this that determines primarily whether a speaker is best mounted against a wall or free-standing. These different placement options will then result in different diffraction characteristics which may be reflected in the crossover design.

Steve Margolis
defy convention - make music

V7!bass means to me: deep frequencies,at a given Quantity.As i suppose that the direction
of deep frequencies is less important if size of wave gets bigger than the distance left to right ear (the reason why we cannot hear direction of bass )the quantity of bass is the
on/off wall design factor.When housing of speaker is given i see two ponts:a smaller woofer chassis (reduces quantity of mooved air
at same amplitude)
A reduced amp input by crossoverreduces amplitude at same frequence.)
First way seems not to be effective with a chassis in a speaker that is cut frequenciewise by its sheer size.so is the second way the right one? i think bass ports
are costructions in the direction of a mini,mini,mini Lowther horn.
O.J.

"I believe that 'ported' is a dirty word on this forum and Seventh Veil doesn't use them either - the internal volume of the cabinet, together with the bass driver used, will determine the shape (Q) and extension of the bass response."

And the size of the opening and length of the port. A ported speaker behaves as a helmholz resonator, so all of those physics apply.

Its amazing how much effort goes into cabinet design. Naim's speakers are revolutionary in this area: Look at the NBL, the SL2 (separate frame for mouting the tweeter), and the whole separate box concept.

- GregB

Insert Witty Signature Line Here

I know absolutley NOTHING about this subject, but......

would it be possible to make a programmable crossover?
i.e. plug a laptop or PC into a firewire or USB port and tailor the sound to ones likeing?

Much like re-mapping a chip on a car or bike?

Seems then everyone would have the sound they wanted.
Crossover mappings could be uploaded and then downloaded by others on the internet.

Richard

"would it be possible to make a programmable crossover?"

Or just get an equalizer. Crossovers take into account the behavior adn limits of the drivers, so probably best not to mess with the crossover as a tone control. EQs can mess with the phase though, so there is a price to pay.

- GregB

Insert Witty Signature Line Here

"the SL2 (separate frame for mouting the tweeter)"

Ulrich Hohn, who used to post here, modified his SBLs in that way long before naim adopted the technique. Vuk devised the "ball/nutter" which now appears in a refined way in the fraim.



cheers

Nigel

[This message was edited by Nigel Cavendish on Wed 28 April 2004 at 14:57.]

quote:

---

Originally posted by Greg Beatty:
"would it be possible to make a programmable crossover?"

Or just get an equalizer. Crossovers take into account the behavior adn limits of the drivers, so probably best not to mess with the crossover as a tone control. EQs can mess with the phase though, so there is a price to pay.

- GregB

Insert Witty Signature Line Here

---

Greg !correct what you say,so why not bring the measured behaviour and limit data of an chassis in a processor that has the behaviour of acrossover without its technical lacks,as there are nonlinearity and phaseshifting faults.Processors of today are big and fast enough to do this.(the point is to
include those "behaviour datas".similar to car Esp, also not based on Fix datas ,processing always the wrong behaviour of car.
And we talk about "dynamic" loudpeakers
O.J.

quote:

---

would it be possible to make a programmable crossover?

---

Yes, preferably digital of course. Behringer make some very cheap (and probably nasty?) units that do this. I would think these are a very good place to start when developing a speaker crossover, get it working active with the programmable crossover and then develop a passive or dedicated active unit from that base line (ho ho).

Paul

quote:

---

Originally posted by Paul Ranson:

quote:

---

would it be possible to make a programmable crossover?

---

Yes, preferably digital of course. http://www.behringer.com/DCX2496/index.cfm?lang=ENG make some very cheap (and probably nasty?) units that do this. I would think these are a very good place to start when developing a speaker crossover, get it working active with the programmable crossover and then develop a passive or dedicated active unit from that base line (ho ho).

Paul

---

exactly this do all professional crossover designers.

James

You seem to be describing a somewhat "chicken and egg" situation, though I accept it is probably the best way for the vast majority of speaker builders. The likes of B&W I gather resort to methods such as using lasers to monitor cone behaiviour and no doubt use anechoic chanbers as well, though they probably still finally optimise their crossovers by the method you describe.

For the rest which comes first? Having settled on the type of enclosure do you say it is going to be that volume because that's the size I fancy? Do you say I like a particular make of speaker so I will use the manufacturers parmeters to calulate the volume of the box? Which way round?

How do you avoid making it too challenging for the crossover design to make it all work by choosing an unfortunate combination of driver types and box shape and volume just by bad luck. Which comes first?

regards
GEOFF

quote:

---

Originally posted by Geoff P:
You seem to be describing a somewhat "chicken and egg" situation, though I accept it is probably the best way for the vast majority of speaker builders. The likes of B&W I gather resort to methods such as using lasers to monitor cone behaiviour and no doubt use anechoic chanbers as well, though they probably still finally optimise their crossovers by the method you describe.

For the rest which comes first? Having settled on the type of enclosure do you say it is going to be that volume because that's the size I fancy? Do you say I like a particular make of speaker so I will use the manufacturers parmeters to calulate the volume of the box? Which way round?

How do you avoid making it too challenging for the crossover design to make it all work by choosing an unfortunate combination of driver types and box shape and volume just by bad luck. Which comes first?

---

Of course, I'm not a typical manufacturer as I don't use crossovers. However, if it helps ...

I use the drive-unit manufacturer's parameters to determine a range of possible cabinet volumes. I then decide on the volume, depending on the bass response that I'm aiming for and how many of the full-range drivers I'm using - if I'm planning to use less drivers I decrease the volume to reduce cone excursion and increase power handling.

Having designed a cabinet with the chosen internal volume, I then build it, listen to it and measure it. Modern impulse and impedance measurements effectively remove the room from the measurements and give a good idea of any resonances that can be treated, or not, by the use of damping materials. I don't believe that anechoic chambers would add anything useful to the process (although I'd love to have that laser gear for examining cone behaviour).

Having listened and measured I then go back and resize the enclosure, if necessary. In practice, it has rarely been necessary.

I would imagine that this is how most speaker manufacturers would fit an enclosure to their bass drivers - whether they use crossovers or not. I hope that it's helpful.

Steve Margolis
defy convention - make music

James & Steve

Thanks for you excellent replies.

regards
GEOFF

Paul!ianswered your behringer hint in a fast
but not really serious way.I am sure it is possible to bring all given datas of chassis
into the Behringer by writing it down into
a laptop.but the dynamic behaviour of a chassis
can only be measured when it works(with an amp).there were tries to measure this behaviour
with magnets on chassis (T and A.)others did it
with reflecting laser light.this feedback has to be brought back to amp and preamp.so a correct processor network needs something to get dynamical feedback.Esp in cars gets this information by sensors during you drive the car hard.Behringer gets this information maybe
through induktivity from the voicecoil of chassis (changing impedance).but this says nothing about the acoustical behaviour of membrane(out of different materials,and therefore different diffraction and distortion.)I mean voicoil moves or should move
always like a piston.Membrane does it in quite another way.So the goal should be :measure membran movement and bring it with nearby light
velocity back to a fast processor.imo theoretical no problem because voltage does always nearby light velocity.Membran moves a lot slower out of the reasons mass,gravitation,mechanical acceleration.
This timewise step infront and a superfast processor should handle behaviour of chassis.
all confentional measurements with microphones
are not that effective because they are always
done with tone velocity in air: only 333meters
during one second.
O.J.

What i think about loudspeaker designing by
Listening:there are so many factors like source
amp, music and last not least the non linearity
of our hearing at different levels that i would
prefer to build a loudspeaker which is technical perfect correct and later on bring its concept to optimum.as i remember it is said:form (or in this caseptimizing) follows function.I wonder if there are non technical correct shoebox designs on the market that sound good (Neutral) under several circumstances.several level,several kind of music. do not think so.
O.J