Been reading Vance Dickason "Loudspeaker design cookbok"
Regarding driver spacing it is recommended to keep drivers as close as possible, and to make sure this distance is less than one wavelenght at the crossover frequency.
So typically for a 6,5 and tweeter 2-way, it is hard to get the drivers center to center much closer than about 5,25 inch, wich means crossover should be less than around 2600hZ.
However i have seen a few commercial, and diy speakers with 6,5" + tweeter with a crossover of 3000 Hz or higher.
Does this mean they are poorly comstructed from start, or is is possible to design a good sounding speaker with a crosseover frequency higher than one wavelength?
Regarding driver spacing it is recommended to keep drivers as close as possible, and to make sure this distance is less than one wavelenght at the crossover frequency.
So typically for a 6,5 and tweeter 2-way, it is hard to get the drivers center to center much closer than about 5,25 inch, wich means crossover should be less than around 2600hZ.
However i have seen a few commercial, and diy speakers with 6,5" + tweeter with a crossover of 3000 Hz or higher.
Does this mean they are poorly comstructed from start, or is is possible to design a good sounding speaker with a crosseover frequency higher than one wavelength?
You need to have a look why these recommendations are given and what happens if distances increase. Figure out the actuel length of a "wavelength" in centimeters. Check the distance of your ears. How many waves of what frequency fit between your ears and what does that mean?
It influences localisation. And influences lobing issues. Keep the drivers oriented in a vertical line, keep them close and cross where they need to be crossed to work in their optimum range and you'll be fine.
It influences localisation. And influences lobing issues. Keep the drivers oriented in a vertical line, keep them close and cross where they need to be crossed to work in their optimum range and you'll be fine.
Everything is a compromise at the end of the day. While Vance is entirely correct, so do the best you can to shrink the spacing without losing your mind, then design your crossover accordingly.
Many >25mm tweeters can take a ~1800 Hz crossover point in domestic applications (unless you thrash your speakers), so not all is lost! Paradoxically, shallower crossovers in the overlap that steepen outside the region offer the greatest power handling because the first couple dB of attenuation matter the most.
Many >25mm tweeters can take a ~1800 Hz crossover point in domestic applications (unless you thrash your speakers), so not all is lost! Paradoxically, shallower crossovers in the overlap that steepen outside the region offer the greatest power handling because the first couple dB of attenuation matter the most.
Ok, a particular speaker i am working on now with 6.5+1in tweeter seems to simulate best with a 4k crossover, so this may be better than try to push the crossover frequency down, as I can not make the response flat with with a 2.5k crossover, and also the tweeter reosnance is at 1800 hz.
Not really out of the inteligibility range as usually identified in pro audio field (usually it is considered 300hz/ 6khz) but you are closer to 4,5k where our brain/ear start to be less sensitive ( take a look at Fletcher Munson curves).
It may work or not... it'll depend of drivers, filter,... Are you sure you won't goes into trouble about mismatch of directivity between drivers ? (It is one of the reason to cross lower).
It may work or not... it'll depend of drivers, filter,... Are you sure you won't goes into trouble about mismatch of directivity between drivers ? (It is one of the reason to cross lower).
Ideally the centre-to-centre distance of 2 drivers being crossed over should be such that the XO frequency is less than a quarter wavelength of that C-C.
The 1 wl DIckason uses is a compromise number chosen because his head is into cone + domes where unless you have a coax, the required close spacing is almost always impossible.
Much of the “art “ of speaker design is the set of compromises you choose.
dave
I think that's the part where I'd lose my mind.
To that end, we have a CNC mill at work and I have weekends access to screwing around with it, aka learning to use the thing better, which may well have me heavily modifying driver flanges to get C-C spacing as small as unreasonably possible. But do as I say and not as I do.
But I agree with your recommendations!
The coaxial drivers solve the problem nicely. But there are other aspects of compromises to observe. For example, a good practice is to avoid XF at the 1-3kHz range, since this is where our perception of sound is greatest and any (unavoidable) deficiency will be probably audible. This is why many consider 3 way speakers a better solution.
Also, it is more important to set the XF at least one octave above the tweeter resonance frequency and combine that with a steep (e.g. 3rd order slope). And if the tweeter doesn't have a ridiculously low resonance frequency, you need to go up in frequency and then your woofer will need to operate at higher frequencies where it will have limited dispersion and also most likely have cone breakup that also affects sound.
I think that these guidelines should be observed so as to avoid silly mistakes, like leaving 10" between the woofer and the tweeter and moving the tweeter away from the vertical axis.
Also, it is more important to set the XF at least one octave above the tweeter resonance frequency and combine that with a steep (e.g. 3rd order slope). And if the tweeter doesn't have a ridiculously low resonance frequency, you need to go up in frequency and then your woofer will need to operate at higher frequencies where it will have limited dispersion and also most likely have cone breakup that also affects sound.
I think that these guidelines should be observed so as to avoid silly mistakes, like leaving 10" between the woofer and the tweeter and moving the tweeter away from the vertical axis.
Hi, i don't want to start a new thread but i really want to know about this, can you please explain this to me in simple words?? wave length, center to centre distance
Last edited:
If you measure the centre of one driver to the other (of any 2 drivers being XOed) you have the centre-to-centre distance. Convert this into a wavelength (f=13500/dc-c) divide by 4 and yo have the equivalent ¼ wavelength didtance. Ideally the XO will be below this.
Tom Danley showed that if you can get this close the drivers are essentially coincident, ie like a coax but without fewer issues (ie the tweeter on a coax would still need some time delay to be coincident). Means no combing and is a huge aid in getting phase coherence.
dave
Tom Danley showed that if you can get this close the drivers are essentially coincident, ie like a coax but without fewer issues (ie the tweeter on a coax would still need some time delay to be coincident). Means no combing and is a huge aid in getting phase coherence.
dave
Say, you have a 2-way system with tweeter and woofer then you need a crossover to separate the signal between those two. So, you will employ a low-pass which lets low frequencies through but for higher frequencies there will first be a gentle decline and then a steep drop. For the high-pass the same is true in reverse. You want your speaker to play equal volume over the whole frequency range though, so you need those declines to overlap in a way that they add up to the same overall volume produced by each individual chassis in their respective prime discipline. Therefore, at a specific frequency both chassis will play equally loud. You just want them to simply add up in volume but that’s not what happening. Instead you have two sources for sound waves with a certain wavelength which generally speaking are in phase. If the listener is equally spaced from them (or at least the difference is a whole-number multiple of the wavelength) then the two sound waves from the two sources will be in phase and everything is going smoothly but if there is a deviation by 1/2 wavelength the waves will be out of phase at the listener position and the two waves will cancel each other out. So, depending on the position or, simplified, the direction you are sitting in front of your speakers you may get widely varying results. That’s called lobing as for example shown here:
The degree to which this is an issue depends on three things: wavelength (thus, the frequency you choose for your crossover), the distance between the two chassis and the relative volume of the two chassis.
If you cross at a low frequency and position the chassis very close together they might only interfere which each other like depicted in the top row in this picture but if you cross at higher frequencies and separate them more it will look more like in the bottom.
Also, at the crossover frequency this issue is especially striking since both chassis are equally loud producing results like in the picture but even at small deviations from that frequency each chassis will significantly contribute to the overall volume producing the same but less pronounced effect.
The recommendations for 1/4 wavelength are only a rule of thumb since this effect will always be present and you merely try to minimize it.
In general, chassis meant to play bass are larger since they have to displace larger amounts of air while chassis meant for heights are smaller to minimize beaming (which essentially is the same as lobing). So, woofers will naturally have a higher spacing than tweeters but that’s not a big deal since the wavelengths will more than compensate for that. Therefore, in a three-way the woofer-mid crossover is less of an issue than the mid-tweeter crossover.
The lobing happens only in the plane the chassis are placed in. So, if one is placed on the left side of the speaker and one is placed on the right side of the speaker but at the same height you will only have lobes in the horizontal plane. On the vertical plane in front of the speaker there will be no lobes since moving on this plane will not change the relative distances to the listener.
Imho it is therefore best to place chassis as close to each other as possible, especially tweeter and mid. Those two should also be placed with a vertical offset so that the lobes point at the floor and at the ceiling. And finally, they should be crossed at a reasonably low crossover frequency. For mid and woofer all of this remains generally true but is much less important.
The degree to which this is an issue depends on three things: wavelength (thus, the frequency you choose for your crossover), the distance between the two chassis and the relative volume of the two chassis.
If you cross at a low frequency and position the chassis very close together they might only interfere which each other like depicted in the top row in this picture but if you cross at higher frequencies and separate them more it will look more like in the bottom.
Also, at the crossover frequency this issue is especially striking since both chassis are equally loud producing results like in the picture but even at small deviations from that frequency each chassis will significantly contribute to the overall volume producing the same but less pronounced effect.
The recommendations for 1/4 wavelength are only a rule of thumb since this effect will always be present and you merely try to minimize it.
In general, chassis meant to play bass are larger since they have to displace larger amounts of air while chassis meant for heights are smaller to minimize beaming (which essentially is the same as lobing). So, woofers will naturally have a higher spacing than tweeters but that’s not a big deal since the wavelengths will more than compensate for that. Therefore, in a three-way the woofer-mid crossover is less of an issue than the mid-tweeter crossover.
The lobing happens only in the plane the chassis are placed in. So, if one is placed on the left side of the speaker and one is placed on the right side of the speaker but at the same height you will only have lobes in the horizontal plane. On the vertical plane in front of the speaker there will be no lobes since moving on this plane will not change the relative distances to the listener.
Imho it is therefore best to place chassis as close to each other as possible, especially tweeter and mid. Those two should also be placed with a vertical offset so that the lobes point at the floor and at the ceiling. And finally, they should be crossed at a reasonably low crossover frequency. For mid and woofer all of this remains generally true but is much less important.
you didn't understand what i have said
i wanted an example of how to do the whole calculation, i mean in detail and step by step. Suppose i want to build a woofer asisted widerange(WAW is more appropriate term than FAST according to some; including you ) and the C2C distance between the drivers is 6inch, then...? How to calculate the proper XO frequency? or if i choose 250hz as XO then how much should be the C2C distance between the drivers? Believe me i do not want to be spoon fed, but i'm helpless, not good in math
I did explain that in the post before you asked.
It is simple. Measure the driver centre-to-centre. Take the speed of sond (13,500 in/sec), divide by the c-c (in in) then divide by 4 to get the ¼ wavelength frequency. That is the highest XO you can use to achieve the ¼ wavelength criteria.
You can of course choose speed of sound in other units as long as the c-c units are the “same”.
dave
It is simple. Measure the driver centre-to-centre. Take the speed of sond (13,500 in/sec), divide by the c-c (in in) then divide by 4 to get the ¼ wavelength frequency. That is the highest XO you can use to achieve the ¼ wavelength criteria.
You can of course choose speed of sound in other units as long as the c-c units are the “same”.
dave
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.