When using 5" Satori MR13p midranges with a ribbon or AMT driver what is the maximum spacing at a 3khz crossover?
I am looking at the Aurum Cantus AST25120 Aero Striction tweeter (6.3 inches or 160mm)
And the RAAL 140-15D ribbon tweeter (7 inches or 180mm)
So looking at a MTM spacing of 6.3-7" or 160mm-180mm
I am looking at the Aurum Cantus AST25120 Aero Striction tweeter (6.3 inches or 160mm)
And the RAAL 140-15D ribbon tweeter (7 inches or 180mm)
So looking at a MTM spacing of 6.3-7" or 160mm-180mm
Last edited:
I don't have a straight answer, and in fact I'm not sure if there is one.
My experience so far (I'm not one of the most experienced around here): I have Beyma TPL-150H, a large AMT tweeter. Waveguide height is 23cm. I was looking into using two 8" in MTM crossing at 2kHz. Simulation models told me I would get nasty vertical lobbing and wouldn't work. A year ago I bought two B&C 8PE21 thinking of TM use vs MTM...but since I could test it for free I built one prototype in MTM and ran it mono, and purchased the materials to wind an inductor to place in front of on of the 8" so both would be active at the lower end but just one active at the xo to the tweeter (to reduce lobbing). Built the MTM prototype and played it without inductor, and to my surprise lobbing wasn't that big of a deal to my ears, in my room. I have not yet tried the inductor, but in April I purchased two additional 8PE21 because I'm definitely going MTM (high efficiency is important to me), and been playing two speakers in MTM. I also tried TM and MT in the same baffle (just disconnecting one M), and preferred MTM.
Someone made a comment explaining AMTs have large surfaces emitting and hence the simplification in center-to-center mathematical modelling doesn't work so well. Maybe that is why simulation software predicted such different results? Or maybe I have tin ears? Dunno, but works for me. I can hear the improved treble from my AMTs vs my B&W 804S, though.
I'm awaiting the arrival of a very revealing 45-based SET to drive the tweeters and test again. I'm kind of hoping I can drive the mids with another one. Obviously mine is an active system.
Another finding along my journey: open baffle was such an improvement using the same drivers as in sealed cabinets!
Bottom line: I would suggest you do the research, understand what to expect, then test to confirm. You might be pleasantly surprised, like me!
My experience so far (I'm not one of the most experienced around here): I have Beyma TPL-150H, a large AMT tweeter. Waveguide height is 23cm. I was looking into using two 8" in MTM crossing at 2kHz. Simulation models told me I would get nasty vertical lobbing and wouldn't work. A year ago I bought two B&C 8PE21 thinking of TM use vs MTM...but since I could test it for free I built one prototype in MTM and ran it mono, and purchased the materials to wind an inductor to place in front of on of the 8" so both would be active at the lower end but just one active at the xo to the tweeter (to reduce lobbing). Built the MTM prototype and played it without inductor, and to my surprise lobbing wasn't that big of a deal to my ears, in my room. I have not yet tried the inductor, but in April I purchased two additional 8PE21 because I'm definitely going MTM (high efficiency is important to me), and been playing two speakers in MTM. I also tried TM and MT in the same baffle (just disconnecting one M), and preferred MTM.
Someone made a comment explaining AMTs have large surfaces emitting and hence the simplification in center-to-center mathematical modelling doesn't work so well. Maybe that is why simulation software predicted such different results? Or maybe I have tin ears? Dunno, but works for me. I can hear the improved treble from my AMTs vs my B&W 804S, though.
I'm awaiting the arrival of a very revealing 45-based SET to drive the tweeters and test again. I'm kind of hoping I can drive the mids with another one. Obviously mine is an active system.
Another finding along my journey: open baffle was such an improvement using the same drivers as in sealed cabinets!
Bottom line: I would suggest you do the research, understand what to expect, then test to confirm. You might be pleasantly surprised, like me!
I agree. There is no practical answer that says an MTM with higher xover and/or wider than theoretically optimum driver c-t-c spacing won't work acceptably. I've built a few MTM rule breakers in this regard with limited vertical dispersion HF drivers and lobing wasn't nearly as much of an issue as initially expected, probably due to the narrow vertical HF. The xover slopes do have alot to do with this and the theoretical rules that apply with phase relationships are still in effect, but you really need to test it in the real world rather than just getting discouraged by how poorly it may simulate.
What will happen from theoretical point can be seen in post What are some good example of baffle design to improve diffraction
Increasing distances between speakers to 20 cm brings start of lobbing to 2 times lower frequency.
I had MTM with distance of mid centers 42 cm: AMT Pro-4 as OB tweeter , they are cut on 1 kHz, but I had not measured vertical directivity.
Increasing distances between speakers to 20 cm brings start of lobbing to 2 times lower frequency.
I had MTM with distance of mid centers 42 cm: AMT Pro-4 as OB tweeter , they are cut on 1 kHz, but I had not measured vertical directivity.
Ideally the sources should be within 1/4 wavelength at the crossover frequency. At 3 kHz the wavelength is 11.4 cm, which means a maximum of 2.9 cm. Note that 'distance between sources' is not just the distance between the tweeter and one of the midranges, as there are two midranges these play simultaneously. The centre-to-centre distance of the midranges is a better measure of 'distance between sources'.
Having the midranges at 2.9 cm centre-to-centre, will not happen in an MTM speaker. A reasonable compromise is one wavelength (11.4 cm), which is feasible for an MT speaker but still impractical for an MTM. Decreasing the crossover frequency increases wavelength (at the crossover frequency) and allows for a larger spacing. The RAAL 140-15D has a recommended crossover frequency of 1.6 kHz which allows for almost twice as large a spacing. For the Aurum Cantus AST25120 it is even lower, at 1 kHz.
Other than that (1): large ribbons have a dispersion that narrows with increasing frequency. Above a few kHz it can become really narrow. If a constant vertical directivity pattern is what you are after, skip the large ribbons and skip MTMs. I mention it because you ask about vertical dispersion.
Other than that (2): crossing a 5" midrange at 3 kHz is not ideal for horizontal dispersion either. Either pick a smaller midrange driver or cross over at a lower frequency.
Having the midranges at 2.9 cm centre-to-centre, will not happen in an MTM speaker. A reasonable compromise is one wavelength (11.4 cm), which is feasible for an MT speaker but still impractical for an MTM. Decreasing the crossover frequency increases wavelength (at the crossover frequency) and allows for a larger spacing. The RAAL 140-15D has a recommended crossover frequency of 1.6 kHz which allows for almost twice as large a spacing. For the Aurum Cantus AST25120 it is even lower, at 1 kHz.
Other than that (1): large ribbons have a dispersion that narrows with increasing frequency. Above a few kHz it can become really narrow. If a constant vertical directivity pattern is what you are after, skip the large ribbons and skip MTMs. I mention it because you ask about vertical dispersion.
Other than that (2): crossing a 5" midrange at 3 kHz is not ideal for horizontal dispersion either. Either pick a smaller midrange driver or cross over at a lower frequency.
Last edited:
Maybe a crossover strategy rethink would be in order for an MTM incorporating a long(er) AMT/ribbon. It might be wise to roll off the lower mid at say 1.2-1.6khz (in this case) and let the upper mid continue to 3khz on it's own.
I forgot to mention about my crossovers and slopes: Nevielle-Thiele 2nd order. These are 2nd order around the xo point but then drop very steeply. Maybe that has to do with what I'm hearing?
Right now my xo is 1.8kHz (been moving t up starting from 1.5kHz), c-t-c between M-T is 22.5cm and c-t-c between both M-M is 45cm...modelling will show how "poor" this choice was. Doesn't sound poor though.
My testing: while playing music, move the head up (stand up) and down (kneel). Within the tweeter dispersion window I can't hear much difference. Vertical dispersion is 30 degrees for my tweeters, so that's a relatively narrow window but it was taken into consideration in the design and placement.
I do have measurement equipment and measure quite a bit, but at the end of the day I confirm by listening.
It's important to note my system is active with digital xo using Acourate and a multichannel DAC, so trying alternatives is easy. Will your system be active? Even if not, you are probably best off by getting a MiniDSP to play with xo types and points and once you settle on what you like you can build the passive network for it.
One of the designers I respect billfitzmaruice has said that Joe D'Appolito doesn't adhere to his 1 wavelength rule anymore. I will quote him "Even Joe D'Appolito doesn't adhere to his own 1 wavelength rule any more. One reason is that on the horizontal axis, the axis that matters, you can't hear combing or lobing that takes place on the vertical plane. The other is that vertical pattern control is enhanced by wider spacing, so what's 'lost' in one area is 'gained' in another with wider spacing, within reason of course. I certainly wouldn't go beyond 2 wavelengths, and at shorter listening distances even that would be too much."
So a maximum of two wavelengths since it all combines with enough distance and I am not listening nearfield. So what would the max distance be for two 5" woofers at 1, 1.5, and 2 wavelengths be? I am using a pre-existing crossover per the guide on one of his speakers.
So a maximum of two wavelengths since it all combines with enough distance and I am not listening nearfield. So what would the max distance be for two 5" woofers at 1, 1.5, and 2 wavelengths be? I am using a pre-existing crossover per the guide on one of his speakers.
I was thinking about this when reading your post the other day, too. At what distance do you listen mostly? That will probably help define spacing and xo frequency. Also, it would be good to consider that Joe D'Appolito may not be lumping in a configuration such as what you're proposing here. You're going to have to mock this up and measure at some point.
Loudspeakers are a bag of compromises and driver spacing is one of them. To determine what value you think is best, it helps to know how spacing affects loudspeaker behaviour.
Calculate the wavelength according to this information.
Wavelength - Wikipedia