a few more links
ORION challenge to DIY note the 120Hz and 1440Hz crossover frequencies....voice is covered by one driver
PLUTO-2.1 omni-directional loudspeaker note the 1kHz crossover frequency
ORION challenge to DIY note the 120Hz and 1440Hz crossover frequencies....voice is covered by one driver
PLUTO-2.1 omni-directional loudspeaker note the 1kHz crossover frequency
isnt 1440hz right in the middle of vocal frequencies? Not that im suggesting to avoid it at all costs, im not, but isnt it just a 2 way?
So here we are. Are there magical crossover frequencies and slopes 🙂
Passive vs Active ?
Please join the discussing and let us get to the bottom of this 🙂
This is just the beggining, we may also need to consider box size, shape lining etc
Passive vs Active ?
Please join the discussing and let us get to the bottom of this 🙂
This is just the beggining, we may also need to consider box size, shape lining etc
To answer OnAudio,
The only magic crossover frequency is the one that works best with the drivers it is used on. Please consider our choice of crossover frequencies is a result of our driver selection. Only when you can design your own drivers will you be free from this constraint.
Passive AND active.
Bottomless pit. We can't get there.
Now you are talking about the easy stuff. Boxes.
The only magic crossover frequency is the one that works best with the drivers it is used on. Please consider our choice of crossover frequencies is a result of our driver selection. Only when you can design your own drivers will you be free from this constraint.
Passive AND active.
Bottomless pit. We can't get there.
Now you are talking about the easy stuff. Boxes.
OnAudio;
How does it capture human voice in one driver if we know human voice extends from 100Hz to 10KHZ (male/female)
Why would you like to cut in the mid of voice (and most instruments) band, if we know of all the problems that ceossovers can introduce. I read that brosure about KEF Muno, haven't heared it, must be good as it is KEF, a lot of money have been put into it; still they talk more about design and guy who designed it, but no word about why they decided to cross at 2.3KHz (though, they tell us why fancy design and two woofers on the back)...
How does it capture human voice in one driver if we know human voice extends from 100Hz to 10KHZ (male/female)
Why would you like to cut in the mid of voice (and most instruments) band, if we know of all the problems that ceossovers can introduce. I read that brosure about KEF Muno, haven't heared it, must be good as it is KEF, a lot of money have been put into it; still they talk more about design and guy who designed it, but no word about why they decided to cross at 2.3KHz (though, they tell us why fancy design and two woofers on the back)...
Crossover points in the dipole speakers like the Orion and the NaO (125 and 2.2k) are determined by consideration of polar response and max SPL capability.
John k, ok, I am not familiar with dipole designs, of course I can read about it but, in short, can you say what is advantage or disadvantage of dipole over bass reflex, or the other way around...
True. This thread title is: "Why 2nd order is best or not" so there is somekind of guess that 2nd order has magic. Does it?
The best option is "no crossover", if there is no other side (ill) effects.
Next best option is 1st order, if there is no other side effects.
Next best is 2nd order. Why 2nd order is more popular? May be because we can still find many 2nd order designs that work, not like 1st order design that requires expensive custom order drivers only used by expensive speaker manufacturers.
Why the slope order is so important? Because we commonly use the filter to separate sound reproduction in the critical frequency so that we want both sound reproduced by tweeter and woofer to blend as natural as possible, without big issue as mostly created by the effect of phase.
If it is a 2-way, or your drivers are like those mostly found in the market, yes, you have no option but to cross it in critical frequency.
How about high order such as LR4 filter crossing in at critical frequency, does it have a chance? I don't want to say it. But that's not my option.
But I have wondered in the past, why 2nd order (electrical) can be so musical (not like 3rd order electrical). I speculated that there is an "uncontrolled" effect of woofer's voice coil in series with external coil without the shunt capacitor in between.
But as tvrgeek implied, everything depends on your ability to judge your drivers capability. Sometimes you need good ears to know what your drivers really like. Then you use your technical knowledge to put those drivers to operate at their best ability.
Maybe it's the case that some designers found out long ago that the first order series and the second order parallel networks have the advantage that the 'looking back' impedance afforded by the inductor being in shunt with the tweeter acts as a dampener for the treble unit;around the xover freq. and below.
I guess this does depend on the drivers being used .. like somebody mentioned above. An LR4 isn't exactly a "brickwall". I'm using an old pioneer 12" woofer (out of an HPM100) crossed w/LR4 @ 1320hz to a heil amt. I can not tell, by ear, where each driver takes over. I still have content out of the woofer up to 2500hz.
Both drivers overlap from 1k to 2.5k (checked). Might be that the directivity of the woofer-heil at the xo is "just right" .. don't know .. but it sounds perfectly fine to me. Crossing higher or lower just sounds "wrong" .. w/instruments and voices.
Lot's of good info/reading in threads like this but sometimes in between the tech/opinions/Hoyle a guy like me starts to second guess a configuration that sounds good ... but shouldn't. 🙂
Least not forget, second order is a lot cheaper than third if you are in the business of building speakers. This is the big advantage DIY has. We can decide if we want that next $15 coil or not.
For many years I used 2nd order 12dB/octave crossovers, in the Q=0.5 Linkwitz-Riley LR12 variety implemented by a Pioneer D23 crossover. 2nd order gives fast transition, which is useful in many ways, with less complexity than higher orders. The LR12 shares property with all Linkwitz-Riley crossovers that the two drivers are always in phase with each other.* That means the greatest "lobe" of maximum response projects perpendicular from the plane of the two drivers, not toward the floor, another plus. (*Actually, for 2nd order, LR12, the two sides are always 180 degrees out-of-phase so you need to invert one. For LR24 or LR48 no inversion is required.)
Making all these plays, however, typically I only accounted for the additional rolloff of the drivers by fudging the crossover frequencies slightly, or sometimes even moved one of the sides to 1st order (effectively 2-3rd order acoustic response in some cases).
Meanwhile, many complain that you need to preserve pulse waveforms perfectly. You can't do that with anything other than acoustic first order. Well many people do first order, or "no-crossover" (cap for tweeter only) designs and think they are getting acoustic first order. And they usually aren't getting anything like that anyway. If you rely on speaker rolloff, you will typically get something faster than first order rolloff.
Now, however, with digital crossovers like DCX2496 I typically use Linkwitz Riley 24 or 48dB per octave (4th or 8th order). There's often benefit from making transition as fast as possible, here it means my electrostatic panels can be played as low as possible, but not so low that they buzz on deep bass, etc. So now with an electrostatic system, my panels play from 80-18Khz, cross over to subwoofer and supertweeters. That wide range from the panels means I have very good pulse reproduction in the midrange, despite using LR crossovers. I have also tried Butterworth or Bessel, sometimes like those better when using cutoffs asymmetrically.
Making all these plays, however, typically I only accounted for the additional rolloff of the drivers by fudging the crossover frequencies slightly, or sometimes even moved one of the sides to 1st order (effectively 2-3rd order acoustic response in some cases).
Meanwhile, many complain that you need to preserve pulse waveforms perfectly. You can't do that with anything other than acoustic first order. Well many people do first order, or "no-crossover" (cap for tweeter only) designs and think they are getting acoustic first order. And they usually aren't getting anything like that anyway. If you rely on speaker rolloff, you will typically get something faster than first order rolloff.
Now, however, with digital crossovers like DCX2496 I typically use Linkwitz Riley 24 or 48dB per octave (4th or 8th order). There's often benefit from making transition as fast as possible, here it means my electrostatic panels can be played as low as possible, but not so low that they buzz on deep bass, etc. So now with an electrostatic system, my panels play from 80-18Khz, cross over to subwoofer and supertweeters. That wide range from the panels means I have very good pulse reproduction in the midrange, despite using LR crossovers. I have also tried Butterworth or Bessel, sometimes like those better when using cutoffs asymmetrically.
Last edited:
This has been studied and it can be done in other ways, such as 2nd order.
If there is an underlying reason this is the case, it may be worth working on those issues. There is also a benefit to transitioning slowly to smooth the blending of driver patterns so this is normally a consideration or tradeoff.
I agree with you there, Allen. Too quick a transition can sound odd. Some of that is the polar pattern but I also think some of it is harmonics. Not always, but often enough.
Exactly. At low frequency there is the argument with regard to room modes of different woofer formats: dipole, cardioid, monopole. Then there is the issue of power response; dipole and cardioids radiate less power for the same on axis response as a monopole. Then the issue of room pressurization... In the midrange and higher there is the issue of uniformity of the polar pattern and constant power response, and reflections from side and back walls.
In sweeping generality, dipoles tend to have a more open sound where as monopole are more focused. As far as what works best in a given environment and what an individual listener preferrs, "it depends".
To narrow that down a bit, at least in the midrange I would say that the "magic crossover frequency" is where the directivity of the mid matches that of the tweeter. The trick is to find a tweeter that can play low enough to match directivity, loud enough to reproduce one's music of choice (a bluegrass fan will have different expectations than someone who judges a speaker's dynamic capability on something like Shostakovich 10), and still plays high enough.
Not just dipoles!
IMO, if crossover points aren't chosen by consideration of directivity/polar response (i.e. to sum either flat, or with some holes, but no mushroom clouds of midrange energy) then the speaker isn't well-designed.
- Status
- Not open for further replies.