Hello. I have read a number of rules of thumb about mid tweeter driver spacing. I'm wondering what the actual process looks like for y'all
Here is how I have opted to do it for the project I'm working on now. I'm mating the Dayton ND25FW tweeter to the Dayton DS90 mid. I looked at manufacturer supplied graphs, and figured anywhere from 3-4KHz would probably work as a good crossover point. I settled on 3700Hz as a starting point. So I used the kimmosto rule which specifies the wavelength of the crossover point * 1.2 should be your CTC spacing. In this case
3700Hz is 3.66".
3.66*1.2= 4.392
So I made the spacing 4.4", and have 3700Hz as my starting point for crossover work.
How do you guys do it? I know some people just put them as close together as possible. I've heard below 1 wavelength CTC as another rule of thumb. I'm assuming some people use sims like akabak. I wonder if anyone makes vertically adjustable baffles and plays with spacing until they find a sweet spot? I'm just interested in the process here among other DIYers.
Here is how I have opted to do it for the project I'm working on now. I'm mating the Dayton ND25FW tweeter to the Dayton DS90 mid. I looked at manufacturer supplied graphs, and figured anywhere from 3-4KHz would probably work as a good crossover point. I settled on 3700Hz as a starting point. So I used the kimmosto rule which specifies the wavelength of the crossover point * 1.2 should be your CTC spacing. In this case
3700Hz is 3.66".
3.66*1.2= 4.392
So I made the spacing 4.4", and have 3700Hz as my starting point for crossover work.
How do you guys do it? I know some people just put them as close together as possible. I've heard below 1 wavelength CTC as another rule of thumb. I'm assuming some people use sims like akabak. I wonder if anyone makes vertically adjustable baffles and plays with spacing until they find a sweet spot? I'm just interested in the process here among other DIYers.
Attachments
An Ideal spacing, for two drivers to act essentially as one, is within 1/4 wavelength. But unless the tweeter is very wide range, such as full-range fans might use, it is usually not possible to physically get it close enough.
In that case vertical lobing will occur. Here I try to set the spacing and crossover to direct the lobe(s) to work well with the listening position. It can even be advantageous to place the cancellations or nulls to reduce floor or ceiling reflections; 1/4 wavelength isn't the only option.
Some people will consider other approaches, such as coaxial drivers and I personally am keen on multiple entry horns. But there are always compromises and complications with any approaches.
In that case vertical lobing will occur. Here I try to set the spacing and crossover to direct the lobe(s) to work well with the listening position. It can even be advantageous to place the cancellations or nulls to reduce floor or ceiling reflections; 1/4 wavelength isn't the only option.
Some people will consider other approaches, such as coaxial drivers and I personally am keen on multiple entry horns. But there are always compromises and complications with any approaches.
As @Kev06 pointed out, the best spacing is 1/4 wavelength, but this can be stretched to 1/3 wavelength without much problem... but if this spacing can't be achieved, closer is not necessarily better.
The 1.2xWaveLength rule works well if the goal is to get a smooth DI curve for the two octaves around the crossover frequency. A smooth DI curve means that the on-axis curve and the sound power curve have a similar shape. This is a big plus, because a DI curve which smoothly increases (or is flat) from low to high allows a speaker to be placed in a wide variety of locations in the room, i.e. it is not finicky about placement.
There are two downsides:
First, the vertical bandwidth is narrow. We will need the tweeter to be positioned at ear level. If the tweeter is at ear level while sitting, the speaker is not going to sound its best while we are standing. With a mid/tweeter spaced further than 0.25xWaveLength, this is the only way the DI curve can be smooth... The vertical directivity must be narrow.
Second, it looks odd. Sometimes it looks very odd. This can be a significant drawback if we are trying to sell the speaker to a customer, or if we just don't like the looks of it. I have gotten accustomed to the look, but others just can't get past it.
Examples:
2-way passive speaker, 170 mm woofer and 25 mm dome tweeter, 2k crossover, CtC spacing of 203 mm. Sounds great, looks strange...
3-way active speaker, 150 mm mid, same 25 mm dome. Crossover frequency 1.9k, CtC distance 210 mm. It looks odd, and many people have asked me if it was a mistake, or wondered if I was crazy or incompetent. Sounds great, looks strange.
The 1.2xWaveLength rule works well if the goal is to get a smooth DI curve for the two octaves around the crossover frequency. A smooth DI curve means that the on-axis curve and the sound power curve have a similar shape. This is a big plus, because a DI curve which smoothly increases (or is flat) from low to high allows a speaker to be placed in a wide variety of locations in the room, i.e. it is not finicky about placement.
There are two downsides:
First, the vertical bandwidth is narrow. We will need the tweeter to be positioned at ear level. If the tweeter is at ear level while sitting, the speaker is not going to sound its best while we are standing. With a mid/tweeter spaced further than 0.25xWaveLength, this is the only way the DI curve can be smooth... The vertical directivity must be narrow.
Second, it looks odd. Sometimes it looks very odd. This can be a significant drawback if we are trying to sell the speaker to a customer, or if we just don't like the looks of it. I have gotten accustomed to the look, but others just can't get past it.
Examples:
2-way passive speaker, 170 mm woofer and 25 mm dome tweeter, 2k crossover, CtC spacing of 203 mm. Sounds great, looks strange...
3-way active speaker, 150 mm mid, same 25 mm dome. Crossover frequency 1.9k, CtC distance 210 mm. It looks odd, and many people have asked me if it was a mistake, or wondered if I was crazy or incompetent. Sounds great, looks strange.
I learned about the kimmosto method from you. Definitely a weird look, and I share the prejudice coming from starting my hifi enthusiasm with vintage gear, where similar layouts are generally just bad design. The knowledge that it's been done thoughtfully in your implementations ameliorates my apprehension pretty well. I can appreciate that it looks pretty neato when I'm not concerned about the performance. Just takes some getting used to like your said. Form and function in happy harmony 👍
I put them as close together as possible because I don't really know what the crossover point will be until I measure the drivers in the cabinet.
Don't forget that larger CTC spacing only required if you want a relative low XO frequency with the 1.2× (or 1-1.4×) wavelength rule.
If you have a midrange driver that can play higher, your tweeter would be more happy with a higher XO and larger (also weird looking 🙂 ) CTC distances don't required.
Vintage speakers may have larger CTC distances in some cases but they typically don't utilizes low crossover frequencies between mid and tweeter.
If you have a midrange driver that can play higher, your tweeter would be more happy with a higher XO and larger (also weird looking 🙂 ) CTC distances don't required.
Vintage speakers may have larger CTC distances in some cases but they typically don't utilizes low crossover frequencies between mid and tweeter.