So I have been repeatedly told, I wanted to cross over a tweeter as low is it can comfortably go.... But I don't understand that. If I had a tweeter that could cross comfortably down to 1200hz, but my Pro Audio midrange was very smooth up to 3000hz or even more, why wouldn't I want my Pro mid to handle the 1200 to 3000 region. This should relieve the tweeters from having to play so low. I would think this would allow the tweeters to sound better and less strained. I only listen from 7.5 feet away from the speakers.
It is obvious I am ignorant, to be thinking backwards like this. Please educate me. Thanks
It is obvious I am ignorant, to be thinking backwards like this. Please educate me. Thanks
The tweeter sound quality is always worse than a paper cone midrange IMO.
In 2004/5 I had a neighbor that have a very transparenr pocket Sonus Faber Concertino Domus with a beatiful ''lute'' enclosure, the xover was very small and the cut freq was 3.8kHz.
It was not easy to drive due 88dB and 4 ohms
too bad.
In 2004/5 I had a neighbor that have a very transparenr pocket Sonus Faber Concertino Domus with a beatiful ''lute'' enclosure, the xover was very small and the cut freq was 3.8kHz.
It was not easy to drive due 88dB and 4 ohms
too bad.One of the top reasons have to do with off-axis response. Think of a hypothetical example of a 8" mid-woofer crossing over at 2.5kHz. At that frequency, the 8" has long gone into beaming, with a very narrow angle of output. Yet as soon as the 1" tweeter kicks in the dispersion becomes perfect. This becomes very audible - your head has to be fixated at the sweet spot and moving just an inch would ruin the FR. This example is not so hypothetical as it is exactly what the active monitors sitting on my computer desk do, and in their defense, they cannot be faulted because they are sold as nearfield monitors.
Lowering the crossover frequency means the mid-woofer will have better off-axis response at that frequency, which reduces the abruptness of the transition.
In a 2-way design, the engineer is always struggling between the bass and the midrange. To have more bass the mid-woofer is bigger (like my 8" example), but it will beam (lose off-axis response) more rapidly. The solution is to use a tweeter which can play lower so that the crossover can be lower.
Lowering the crossover frequency means the mid-woofer will have better off-axis response at that frequency, which reduces the abruptness of the transition.
In a 2-way design, the engineer is always struggling between the bass and the midrange. To have more bass the mid-woofer is bigger (like my 8" example), but it will beam (lose off-axis response) more rapidly. The solution is to use a tweeter which can play lower so that the crossover can be lower.
Rayma is right that this is the _general_ rule and certainly how I do it for two-ways.
This isn't because of the tweeter so much as the mid-woofers though. Midwoofers they are mated with generally have cone break-up and other anomalies at the top of the range, not to mention dispersion matching (like Cyberstudio mentions) issues. This all tends to (again, not written in stone) push designers to a lower x-over point.
However, then some designers like Joseph Audio seem to like doing this even with 3-way designs. They do this in part with very steep crossovers.
On the other hand, I don't know how you could do this with a Thiel or Vandersteen- like design with 1st order crossovers.
Best,
E
This isn't because of the tweeter so much as the mid-woofers though. Midwoofers they are mated with generally have cone break-up and other anomalies at the top of the range, not to mention dispersion matching (like Cyberstudio mentions) issues. This all tends to (again, not written in stone) push designers to a lower x-over point.
However, then some designers like Joseph Audio seem to like doing this even with 3-way designs. They do this in part with very steep crossovers.
On the other hand, I don't know how you could do this with a Thiel or Vandersteen- like design with 1st order crossovers.
Best,
E
Hmm ... I am trying to determine if I can use my Heil AMTs in my new 3 way build, and cross them at 3000hz to this midrange driver:
Ciare NDK6-1.5 - Ciare NDK6-1.5 is a 6.5 inch lightweight neodymium midrange speaker for all midrange speaker systems- Ciare Speakers - Ciare NDK6-1.5 6.5 inch lighweight neodymium midrange speaker. Ciare NDK6-1.5 6.5 inch midrange speakers
This will be all active, with full DSP.
Ciare NDK6-1.5 - Ciare NDK6-1.5 is a 6.5 inch lightweight neodymium midrange speaker for all midrange speaker systems- Ciare Speakers - Ciare NDK6-1.5 6.5 inch lighweight neodymium midrange speaker. Ciare NDK6-1.5 6.5 inch midrange speakers
This will be all active, with full DSP.
You mean if the woofer is large. If you have a small two way, then higher is OK. It's a compromise at the end. I still think the midrange is best left for the woofer, not the tweeter.
I think we have to qualify "3-way" as the "classic" 3-way, with something like a 3" or 4" max midrange.
I have seen 3 way designs which are the ubiquitous 1"+5.25" design with a woofer slapped on only to fill in the bottom midbass octaves (50-200Hz). And I have seen discussions here looking for 8" or even bigger midrange drivers. Those needs to be designed like classic 2-ways as far as the high-mid crossover point is concerned.
Another reason why there cannot be a hard-and-fast rule is when you want to deliberately control dispersion, as in horn-loading the tweeter. Then your crossover point is arrived at in conjunction with the dispersion characteristics of the horn and the midrange or mid-woofer near the crossover region.
I have seen 3 way designs which are the ubiquitous 1"+5.25" design with a woofer slapped on only to fill in the bottom midbass octaves (50-200Hz). And I have seen discussions here looking for 8" or even bigger midrange drivers. Those needs to be designed like classic 2-ways as far as the high-mid crossover point is concerned.
Another reason why there cannot be a hard-and-fast rule is when you want to deliberately control dispersion, as in horn-loading the tweeter. Then your crossover point is arrived at in conjunction with the dispersion characteristics of the horn and the midrange or mid-woofer near the crossover region.
Last edited:
Cyberstudio has the best reply so far. It's true that crossing over to a tweeter as low as possible when the mid/woofer is beaming is a good idea. Also, you do not want abrupt transitions in the response pattern, aka the directivity.
There is another important point that has not been mentioned - lobing.
When two sources (e.g. two drivers, the mid and the tweeter) are reproducing the same frequency signal and are separated in space (that is to say they are non-coincident, meaning their acoustic centers are separated by some distance) an interference pattern will arise. When the separation distance starts to become about half a wavelength of the sound being produced by both drivers, peaks and nulls form in the response. The higher in frequency, the worse this effect. This is often called "lobing" and the response patterns looks like this:
What's less well known IMO is that as you move from lobe to lobe in the pattern the polarity of the acoustic radiation reverses from positive to negative. The null in between is where positive and negative cancel. That can't be good, eh? All of this can lead to problems, acoustically speaking, in the crossover region where both drivers are both contributing to the total output.
By minimizing the separation between the two sources and by keeping the crossover frequency as low as possible, lobing effects are minimized.
You can read more about lobing here:
Speaker lobing calculator - Polar response - Audio Judgement
.
There is another important point that has not been mentioned - lobing.
When two sources (e.g. two drivers, the mid and the tweeter) are reproducing the same frequency signal and are separated in space (that is to say they are non-coincident, meaning their acoustic centers are separated by some distance) an interference pattern will arise. When the separation distance starts to become about half a wavelength of the sound being produced by both drivers, peaks and nulls form in the response. The higher in frequency, the worse this effect. This is often called "lobing" and the response patterns looks like this:
What's less well known IMO is that as you move from lobe to lobe in the pattern the polarity of the acoustic radiation reverses from positive to negative. The null in between is where positive and negative cancel. That can't be good, eh? All of this can lead to problems, acoustically speaking, in the crossover region where both drivers are both contributing to the total output.
By minimizing the separation between the two sources and by keeping the crossover frequency as low as possible, lobing effects are minimized.
You can read more about lobing here:
Speaker lobing calculator - Polar response - Audio Judgement
.
Well, the standard answer is no one thing shall be the only criterion, I suppose. For instance if the tweeter is pushed so low that either its thermal limit or excursion limit is exceeded, or if it starts noticeably distorting, going too low does no good.
Lobing is decided by 2 factors: the distance between your tweeter and your midrange, and the crossover frequency. The latter is a certain wavelength. (v = f lambda) The rule of thumb is that the distance between the tweeter and the midrange shall not exceed the wavelength of the crossover frequency.
Lobing is decided by 2 factors: the distance between your tweeter and your midrange, and the crossover frequency. The latter is a certain wavelength. (v = f lambda) The rule of thumb is that the distance between the tweeter and the midrange shall not exceed the wavelength of the crossover frequency.
Last edited:
I guess that is my point. There are a lot of criteria when choosing a xover frequency. Just crossing as low as possible does not seem like a good way to go. Crossing low may give a good polar plot (such as reducing lobing) but it compromises on other aspects.
For example, one chooses different types of driver materials, be it paper, aluminum, ceramic to have a certain sound. Paper is warm sounding, whereas ceramic is little bit more detail. For some reason, nobody is asking how is a tweeter sound like? When you crossing the tweeter low, it has more freq. content in the midrange which can have an effect on the sound. Personally when I design my speakers I tend to cross as high as I can get away with so that the woofer will play most of the midrange.
For example, one chooses different types of driver materials, be it paper, aluminum, ceramic to have a certain sound. Paper is warm sounding, whereas ceramic is little bit more detail. For some reason, nobody is asking how is a tweeter sound like? When you crossing the tweeter low, it has more freq. content in the midrange which can have an effect on the sound. Personally when I design my speakers I tend to cross as high as I can get away with so that the woofer will play most of the midrange.
I have never heard any loudspeaker I liked with dome tweeters going as low as 1500Hz , one of the biggest mistakes in audio imho. If your woofer can't play higher because of breakup use another one, if lobing becomes an issue, use higher order preferably FIR filters, if dispersion characteristics are an issue put your midrange on open baffle or use waveguides for a better mathed dispersion, do anything but don't use a 1" dome as low as 1500Hz. The driver distance vs crossover frequency is just one tiny part of the equation.
Sorry but I thought a higher order of a crossover does not mitigate lobing (of course it could change the angle of the lobe, but it does not reduce the number or severity thereof). At the crossover point the tweeter and the midrange are contributing equally to the output and that is regardless of the order of the crossover, so it must lobe at exactly the crossover frequency. (True for IIR, but the exact same reasoning means it is true for FIR too.)
Now you got me thinking and I am thinking out loud here. We all know that a properly designed higher-order crossover sounds better than a properly designed first-order crossover. (I will grant beforehand this can be a over-generalization with counterexamples abound.) But is that because of lobing? (The crossover "region" has a smaller bandwidth with a higher-order XO.) But rather because of the shallower slope reveals more of the weaknesses of the drivers? IMHO I tend to believe it is the latter.
Now you got me thinking and I am thinking out loud here. We all know that a properly designed higher-order crossover sounds better than a properly designed first-order crossover. (I will grant beforehand this can be a over-generalization with counterexamples abound.) But is that because of lobing? (The crossover "region" has a smaller bandwidth with a higher-order XO.) But rather because of the shallower slope reveals more of the weaknesses of the drivers? IMHO I tend to believe it is the latter.
Last edited:
I'm talking about 60db/oct to 300db/oct FIR filtering here, not 12db or even 24db filters. But nomatter wich way you go, just do not use 1" dome tweeters at 1500Hz, it will sound fatigueing sooner or later.
Steeper filters can (not allways) sound better because of lobing issues, because of driver breakup, because of tweeters running out of steam, because of dispersion differences between the drivers.
Lobing is just a matter of overlapping frequencies originated from sources at some distance apart in either dimension in space. The less overlap, the less lobing. But this is static, just the use of very steep crossovers is helpfull but not the cure for all. You still have the problem that harmonics are originated from another source then the fundementals. But that's much less of an isuue to your ears then severe lobing at the crossover region.
Steeper filters can (not allways) sound better because of lobing issues, because of driver breakup, because of tweeters running out of steam, because of dispersion differences between the drivers.
Lobing is just a matter of overlapping frequencies originated from sources at some distance apart in either dimension in space. The less overlap, the less lobing. But this is static, just the use of very steep crossovers is helpfull but not the cure for all. You still have the problem that harmonics are originated from another source then the fundementals. But that's much less of an isuue to your ears then severe lobing at the crossover region.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.