WWWMTMWWW, does it work?

[Jussi]

WWWMTMWWW. Does it work?

3x8", 7", 5" ribbon, 7" and 3x8". With floor reflection at least bottom 8's create sort of line source on their range, ribbon lines up as well at one point of it's range, MTM mids beam but not up to line source.

No-Go mismash or potentially operating setup?

[Paul W]

How do sims look? Should work, especially if outer Ws are rolled off a little earlier than inner Ws (like 0.5 configuration). 7" seems large for horizontal directivity match with ribbon.

[Jussi]

Originally Posted by Paul W

How do sims look? Should work, especially if outer Ws are rolled off a little earlier than inner Ws (like 0.5 configuration). 7" seems large for horizontal directivity match with ribbon.

Should stay under 3-way config since my DCX doesn't support any more channels.

Ribbon is going to get waveguide so radiation pattern shouldn't be a problem.

My primary thought of this is that woofers are lining up causing cylinder shape radiation pattern in vertical level and MTM can't match that. Sims don't help that much since I can't simulate when and how drivers line up as linesource and where and how they work more like point sources.

[Paul W]

Yup, WG can solve ribbon directivity issue.

IIRC, you use LspCAD? If so, you should be able to sim polar response. Try using a (hypothetical) flat 20-20k frd file to isolate array polars from room interference, driver response, etc.

One way to gradually change vertical directivity of a 2 parallel/3 series WWW-WWW so that it better agrees with MTM at XO is to use a capacitor shunt across outer pairs of drivers. (Line length decreases with increasing frequency.) That configuration can preserve power handling and manage vertical directivity without more active channels or changing combined FR of the array.

[StigErik]

You can get a second DCX if you want to do 4, 5, 6-way. Its not all that expensive, but you'll need more power amps.

I once did a 4-way tower almost like your proposal. It was 18" - 8" - 5" - AMT - 5" - 8" - 18". XO frequencies was 70, 500 and 1700 Hz. Quite ok sound, but absolutely horrible vertical polar pattern!

[Jussi]

Paul,

I've simulated vertical responses with that method. There all seems possible but that's just simulation in free space. Difference is where and how WWW-WWW sections run as linesource and how they integrate to point source mid section. I don't know how significant thing this is and that's why I'm asking. Basically there is a risk that linesource pushes (part of) bass range to listeners face while midrange imaging lays back causing uneven stereoimage. Basically similar thing with ribbon tweeter that beams up to a linesource at one point of it's range causing similar effect. Also there I don't know how significant effect is. At least MTM mids are better with longish ribbon compared to standard MT config since it has some vertical directivity.

StigErik,

I remember your 4-ways. Looked great though. Shame I haven't found a picture or website about them anymore.

70Hz lowest cross doesn't create my situation since there is only single driver close to floor and it's lowpassed at so low frequency. It doesn't beam as linesource.

Allthought I build one MTM prototype using these same drivers and experimented kind of Dunlavy SC-6 - Genelec 1036A fusion. Only the midsection prototype was tested. MTM with integrated mid-tweeter waveguides. Seemd promissing but the whole thing is just too darn big. The idea was to stack up 18" BMS woofers to both ends but looks like I have to ditch this idea because of it's size. Maybe in the future if I have a large separated listening space...

Second DCX is a good idea but I have only 6-channel volume control that doesn't fit with more than 6-channels. Normal preamp would naturally do it (like I've used to have) but it puts the volumecontrol into wrong place in the signal line. CD (DA conversion) - Preamp - (AD conversion) DCX (DA conversion) - poweramps - speakers. While it's now in form CD - DCX (DA conversion) - 6-channel volume - poweramps - speakers. I can run DCX with full level signals for max resolution even at low volumes as well as avoid any unnecessary DA/AD conversions.

[Jussi]

Originally Posted by Paul W

One way to gradually change vertical directivity of a 2 parallel/3 series WWW-WWW so that it better agrees with MTM at XO is to use a capacitor shunt across outer pairs of drivers. (Line length decreases with increasing frequency.) That configuration can preserve power handling and manage vertical directivity without more active channels or changing combined FR of the array.

I think I'll have to simulate on that one. I thought to connect 6 drivers with 3 parallel and those set of 3 to series creating nominally little over 4 ohm minimum impedance.

The driver in mind is Seas L22RN4X/P. I know it hasn't got the greatest motor but here some perspectives for it:

- Suitable cabinet requirements (number 1 issue and even more so if triple driver setup for each cabinet is even considered)
- Long linear excursion (+-7mm)
- Similar products have done well with Zaphaudio distorsion tests, especially at bass, midbass and lowmid areas
- Good looks (yep, that's a criteria) compared to competitors
- Made in Norway

So your idea would change this setup quite a bit. I'd have to connect top and bottom drivers as one pair, middle ones as one pair and closest to MTM as one pair. In parallel they all form around 3 ohm minimum and those added to series creates bit over 9 ohm minimum there. For lower parallel cap values should I use parallel pairs (3 ohm minimum) or series pairs (12 ohm minimum)? Don't remember and don't have the sim at this computer.

I just don't know how such connection breaks up the linesource. Basically simulated vertical response isn't a problem if you just simulate it. In theory all is well and vertical directivity is at pretty high level with very broad frequency range. This all adds well with MTM vertical beaming and ribbons behavior. I'd say 3-way D'appolito is a better to start with compared to 2-way D'Appo with just mids and tweeter there. Using 2nd order slopes it's possible to maintain very stable vertical polar pattern and still gain rather good directivity figures. 1st order slopes don't have much directivity (Dunlavy style) and 4th order slopes don't smooth up the humps that well (since drivers responses don't blend over so wide range). With Pro5i I'm out with 2nd order tweeter highpass as well as W18 mids mean steep lowpassing but 2nd order is a true option with lower crossoverpoint. W18s have excellent lower midrange and only L22 distorsion behavior (trough high Le) limits their use.

One suitable option is to ditch WWWMTMWWW config and stick with WWMTMWW. There lowest woofers are far away enough from floor, there isn't additional linesource behavior and basically whole array should work as simulated. Using 2x70-75 litres cabinets with 24-25Hz tuning L22's also deliver quite adequate SPL and dynamics capasity. It's also easier to keep cabinet smaller and use large enough vents to keep airflow speed at low levels to avoid compression and noises.

By the way where do you consider limits of vent airspeed are reached and how low levels they should be for a high end system?

[sreten]

Originally Posted by Jussi

My primary thought of this is that woofers are lining up causing cylinder
shape radiation pattern in vertical level and MTM can't match that.

Hi,

The woofers will only do that for wavelengths longer that the cabinet width
but shorter than the total driver height, so it depends on how low you
low pass the woofers for this to be a major issue.

rgds, sreten.

[Paul W]

It would certainly be interesting to try WWW-WWW both ways to learn subjective differences in-room. Standard implementation might have odd interaction with MTM....shunt bypass should blend more smoothly with MTM. Which is the better overall in-room experience...I dunno??? Maybe build throw-away test boxes just to experiment? Whole system seems like a good candidate for modular build.

Haven't worked with ported boxes for many years, so not qualified to address air speed question.

[whgeiger]

Originally Posted by Jussi

WWWMTMWWW. Does it work?

3x8", 7", 5" ribbon, 7" and 3x8". With floor reflection at least bottom 8's create sort of line source on their range, ribbon lines up as well at one point of it's range, MTM mids beam but not up to line source.

No-Go mismash or potentially operating setup?

Try this config as well

|W|M|W|
|W|T|W|
|W|M|W|

Regards,

WHG