In a 3 to 4-way system, consider 1-3kHz for low x-over point and 4-6 kHz for high x-over point, the tweeter’s crossover point should be at low or high frequency point? I see many hi-end speaker manufacturers at present usually use their tweeters at “low” frequency point. Does this a trend for nowaday (hi-end) speaker? Compare to those hi-end builds from 70s to 90s, they tend to be “high” x-over point for the tweeter. Please help to clarify my suspicion. Further discussion about advantage and disadvantage would be really appreciated. Thank you
Last edited:
When you create a three way system instead of a two way, you have the benefit of being able to make the cross higher. This makes the job of the tweeter easier because it needn't be overwhelmed by having to handle lower frequencies.
To do this well requires a midrange driver that can go higher. A smaller unit will go higher before encountering the issues of breakup, and other changes that make it unusable at higher frequencies.
A two way system usually requires a larger driver to handle bass, and this can limit the upper cross for that woofer.
In addition to this I think it is accepted that errors in a cross that is in a sensitive region of the spectrum will be more of a problem. As a result it has been a trend lately to cross low, and perhaps not a bad one.
To do this well requires a midrange driver that can go higher. A smaller unit will go higher before encountering the issues of breakup, and other changes that make it unusable at higher frequencies.
A two way system usually requires a larger driver to handle bass, and this can limit the upper cross for that woofer.
In addition to this I think it is accepted that errors in a cross that is in a sensitive region of the spectrum will be more of a problem. As a result it has been a trend lately to cross low, and perhaps not a bad one.
Crossover points are highly dependent on the characteristics of the drivers used, the type of speaker --2, 2.5, 3, 4 way-- and in some cases the box design.
Yes crossover points have crept down over the last 15 or so years. Quite often when I get inside a 1970s speaker I will discover a very simple "a coil and a cap" crossover at about 3 or 4khz. But newer stuff seems to want to bring in the tweeter much earlier, sometimes as low as 1000hz.
You will also discover that speakers have become noticeably less efficient in recent years. Back in the 70s and 80s 92 and 93 db per watt was not uncommon. Now we're down in the mid 80s, averaging 86 and 88 db.
This is, at least in part, due to another trend towards far more complex crossovers than in the good old days. We've gone from this...
for the Klipsch Heresy ... to this...
for SVS speakers and beyond.
One of the reasons for the higher crossovers in the past was that tweeters generally don't like bass. They tend to overheat and die on you. Recent changes including ferro-fluid for cooling and more robust voice coils have allowed them to lower this point considerably.
With today's 5.25 and 6.5 inch woofers, crossovers at 2000hz for a two way or 500 and 2500 for a three way generally make sense. Larger woofers will need lower crossover points getting down, in some cases to less than 300hz.
In three way systems, the woofer is often taken down early and the tweeter brought in late so that the midrange speaker gets some "airtime". How much depends on the driver's characteristics and the designer's personal choices.
Does that help?
Yes crossover points have crept down over the last 15 or so years. Quite often when I get inside a 1970s speaker I will discover a very simple "a coil and a cap" crossover at about 3 or 4khz. But newer stuff seems to want to bring in the tweeter much earlier, sometimes as low as 1000hz.
You will also discover that speakers have become noticeably less efficient in recent years. Back in the 70s and 80s 92 and 93 db per watt was not uncommon. Now we're down in the mid 80s, averaging 86 and 88 db.
This is, at least in part, due to another trend towards far more complex crossovers than in the good old days. We've gone from this...
An externally hosted image should be here but it was not working when we last tested it.
for the Klipsch Heresy ... to this...
for SVS speakers and beyond.
One of the reasons for the higher crossovers in the past was that tweeters generally don't like bass. They tend to overheat and die on you. Recent changes including ferro-fluid for cooling and more robust voice coils have allowed them to lower this point considerably.
With today's 5.25 and 6.5 inch woofers, crossovers at 2000hz for a two way or 500 and 2500 for a three way generally make sense. Larger woofers will need lower crossover points getting down, in some cases to less than 300hz.
In three way systems, the woofer is often taken down early and the tweeter brought in late so that the midrange speaker gets some "airtime". How much depends on the driver's characteristics and the designer's personal choices.
Does that help?
Interesting. Just going off memory, some '70s tweeters had 1/2" voice coils. Today that's typically 1", or four times the circumference.
Twice the circumference, 4x the radiating area.
The reason for introducing more crossover points is this: clean SPL, particularly at the frequency extremes.
Let's set a fixed bandwidth: 30Hz-20kHz. That covers most music just fine.
A 3" driver can manage that on its own, but it has limitations: the >8kHz region will be bumpy and have narrow directivity due to the relatively large radiating area. The LF output will be extremely limited because of the relatively small radiating area. So, bumpy treble and limited bass.
Moving up, a 6" 2-way is probably the most common. This fixes most of the problems of a single driver. The treble range is smoother, and there's more LF output. If you go coaxial and do a decent crossover, the midrange ought to be seemless, too. A general improvement.
While a couple of 6" cones might provide enough bass for most people, we want to go further. There's also the matter of uneven directivity from stretching the divers to cover relatively wide bandwidths, and the tweeter might still be struggling towards the bottom of its range.
Cue the 3-way.
A 12" per side will provide a lot more bass than the above speakers. Add in a 4-5" midrange and the same tweeter as above, and you're on to a winner. You'll have some pretty serious SPL capability across the desired frequency range.
IMO, there's very little benefit to putting a crossover in the upper kHz range, with the exception of coaxial compression drivers.
FWIW, you can follow the same more LF SPL = more crossover points trend in PA systems:
1-way - 8" driver with a whizzer cone. Not great.
2-way - 12" 2-way. Some LF capability, and a good compression driver will get down to 1kHz for a good directivity plot
3-way - add 18" subs to the 12" 2-way.
4-way - add 15" "kicks", horn-load the 12". You end up with something like Nexo Alpha.
Chris
The reason for introducing more crossover points is this: clean SPL, particularly at the frequency extremes.
Let's set a fixed bandwidth: 30Hz-20kHz. That covers most music just fine.
A 3" driver can manage that on its own, but it has limitations: the >8kHz region will be bumpy and have narrow directivity due to the relatively large radiating area. The LF output will be extremely limited because of the relatively small radiating area. So, bumpy treble and limited bass.
Moving up, a 6" 2-way is probably the most common. This fixes most of the problems of a single driver. The treble range is smoother, and there's more LF output. If you go coaxial and do a decent crossover, the midrange ought to be seemless, too. A general improvement.
While a couple of 6" cones might provide enough bass for most people, we want to go further. There's also the matter of uneven directivity from stretching the divers to cover relatively wide bandwidths, and the tweeter might still be struggling towards the bottom of its range.
Cue the 3-way.
A 12" per side will provide a lot more bass than the above speakers. Add in a 4-5" midrange and the same tweeter as above, and you're on to a winner. You'll have some pretty serious SPL capability across the desired frequency range.
IMO, there's very little benefit to putting a crossover in the upper kHz range, with the exception of coaxial compression drivers.
FWIW, you can follow the same more LF SPL = more crossover points trend in PA systems:
1-way - 8" driver with a whizzer cone. Not great.
2-way - 12" 2-way. Some LF capability, and a good compression driver will get down to 1kHz for a good directivity plot
3-way - add 18" subs to the 12" 2-way.
4-way - add 15" "kicks", horn-load the 12". You end up with something like Nexo Alpha.
Chris
Actually, from the other point of view, can I say that is it a trend for today to use the low range in a very low crossover point? Consider the 10” woofers in 3-4 ways, 70s-90s hi-end speaker usually crossed their woofer at 500-1,500 Hz. But for today, it is moved to around 80-300 Hz — very low point. Despite most 10” woofers, in the past, can reach to kHz in a smooth response, and I believe, with today’s technology, the 10” woofers can probably reach to even more few kHz. Why crossing them at very low point? Is it something called “integrated subwoofer system” or something like this? Is it a new trend?
There are wide variations in the design of 10" Woofers. Some are designed for low bass with heavy diaphragms of 50 grams or more and voice coils of 50 mm or more in diameter. These tend to exhibit rough response above a few hundred Hz. There is another group with voice coils around 40 mm in diameter and diaphragms around 35 grams that sacrifice low end for a controlled upper roll-off. A perfect example of the later is the Seas A26RE4, a modern version of the woofers Seas made for Dynaco in the 60s and 70s. In the same basket they manufacture the L26RFX/P, an example of a pure woofer. The design option still exists either way. Orangutan manufactures a very successful 2 way with a derivative of the A26.
Directivity, max SPL, cone breakup and avoiding cabinet resonances all factor into deciding crossover frequencies. With the addition of waveguides for dome tweeters the directivity can be better matched in the crossover region and the reduced directivity at low frequencies for the tweeter boosts the low end output also enabling lower crossovers. Boxsim free software by Visaton has a nice example design when you start it up with a one inch dome with a waveguide. This software plots directivity, max SPL and other things that let you see what changing the crossover point will give you.
Hi Prescott,
If we let aside purely technical concerns about drivers and take a look at how our brain/ ear react it makes sense to locate xover away from the range where we are the most sensitive ( or at places where it is less critical).
The second graphic in Troels 'design criteria' show what he call 'ear's sensitivity' which is a inverse representation of Fletcher and Munson curve at probably around 85dbspl.
From around 250hz to 6,5khz you can see we are at maximum sensitivity range (with a small notch around 1khz and a peak at 4khz).
It makes sense to try to stay away from this area to locate xover point or if we have to cross in the range between 1k and 1,5khz.
80hz is considered to be the point where we aren't able to localize sound. It is a point where we can expect the room to have taken the lead about the behavior of bass too. Last thing about it is this is where most first harmonics and fondemental of most instrumental'ssound start so dolby decided to use this point as a good compromise for their certification.
As a result of all this this is commonly used in satellite+sub approach but not only.
If we let aside purely technical concerns about drivers and take a look at how our brain/ ear react it makes sense to locate xover away from the range where we are the most sensitive ( or at places where it is less critical).
The second graphic in Troels 'design criteria' show what he call 'ear's sensitivity' which is a inverse representation of Fletcher and Munson curve at probably around 85dbspl.
From around 250hz to 6,5khz you can see we are at maximum sensitivity range (with a small notch around 1khz and a peak at 4khz).
It makes sense to try to stay away from this area to locate xover point or if we have to cross in the range between 1k and 1,5khz.
80hz is considered to be the point where we aren't able to localize sound. It is a point where we can expect the room to have taken the lead about the behavior of bass too. Last thing about it is this is where most first harmonics and fondemental of most instrumental'ssound start so dolby decided to use this point as a good compromise for their certification.
As a result of all this this is commonly used in satellite+sub approach but not only.
