pros and cons.. WMTMW vs MTMWW ?

[Andy G]

these are a build for a friend, his funding.

probably all Peerless drivers, nomex 8", nomex 5" and 921 tweeter.

What are the reason for choosing WMTMW or MTMWW ?

May try to do the mids open back.

[LineSource]

Properly executed, a WMTMW can do a very good job at reproducing a square wave. The height makes a statement in the listening room.

http://www.musicanddesign.com/icta.html
http://www.steinwaylyngdorf.com/Model_D-45.aspx

Properly constructed, a MTMWW can be low enough in height to not dominate the listening room. The square wave transient response will have some compromises. A TMWW using a larger diameter M like the Linkwitz Orion dipole is a more common design because of its lower height.

[pheonix358]

Hi Andy, both can work well. I always prefer to use WMTMW. This is a point source array at all frequencies and allows you to use any xover points. If using MTMWW then you must keep the mid to woofer xover point below about 100Hz or so to help with imaging. Most speakers are not made like this for one simple reason $$$$$$$$$$$!

Terry

[SY]

I'm unconvinced that, given the wavelengths involved, there's any difference in the point source characteristics. BUT... there is a difference in the response notch due to floor bounce.

[pheonix358]

Originally Posted by SY

I'm unconvinced that, given the wavelengths involved, there's any difference in the point source characteristics. BUT... there is a difference in the response notch due to floor bounce.

Hi Sy,

A lot depends on the slopes of the xover. Point source with 1st order xovers always sound and measure better in WMTMW. If you take that as read then it will also be better with higher order but you may not hear a difference. Depends on your ears.

[Hylle]

To reproduce good squarewave response in this WMTMW setup, do the crossover have to be low orders, maybe even 1 order to make it work well? If you want good squarewave response?

How do you test this, can a Clio meassuring program or similar do this, or do you use a tonegenerator and a mic?

[pheonix358]

Originally Posted by Hylle

To reproduce good squarewave response in this WMTMW setup, do the crossover have to be low orders, maybe even 1 order to make it work well? If you want good squarewave response?

How do you test this, can a Clio meassuring program or similar do this, or do you use a tonegenerator and a mic?

I am not being a smart a**e when I say this but I don't listen to square waves. I often listen to square music as my sons always tell me.

For a speaker to faithfully reproduce music you have to count everything as important. 1st order puts way less parts in the audio path, is easy to impliment but, the caveat is, the drivers have to like each other.

Driver selection is the most important part of a design. Get that right and the rest is fun. Get two drivers that don't get on well and you will have a bit** of a time designing the crossover.

Terry

[SY]

1st order puts way less parts in the audio path, is easy to impliment but, the caveat is, the drivers have to like each other.

It also screws up polar pattern and increases distortion in pursuit of a goal which is questionable at best. It's actually quite difficult to implement, at least if you want an actual 1st order crossover, not just an electrical circuit that looks like one- see some of Thiel's speakers, for example. Those crossovers are VERY complex.

[Andy G]

Originally Posted by pheonix358

Hi Sy,

A lot depends on the slopes of the xover. Point source with 1st order xovers always sound and measure better in WMTMW. If you take that as read then it will also be better with higher order but you may not hear a difference. Depends on your ears.

The crossover will almost certainly be a 1st order (electrical) linked series x-o. Hence the choice of drivers.
The 'client' has heard all of the speakers that I have hooked up here and really likes the sound, (but none suit his decore, darn).

[Andy G]

Originally Posted by Hylle

To reproduce good squarewave response in this WMTMW setup, do the crossover have to be low orders, maybe even 1 order to make it work well? If you want good squarewave response?

How do you test this, can a Clio meassuring program or similar do this, or do you use a tonegenerator and a mic?

To reproduce a square wave, you need to have a tweeter that can go really high, and pretty darn good phase coherance through the double x-o points.

(gotta love them fourier series )