Question about "fixing" a potentially non-ideal d'appolito arrangement

[ctrl]

@ctrl is it possible to perfectly control vertical radiation (while keeping horizontal radiation wide) with such designs? Or do we always end up with inconsistencies?

Perfect no, theoretically a good vertical radiation is possible with D'Appolito (DA) concept, as has already been said, it depends on the driver spacing (M-T as well as M-M) and the crossover concept to control the vertical radiation lobes.

In the original DA concept, the distance between the centers of the midrange drivers must not exceed 2*lambda/3 of the crossover frequency when using a third order crossover.
Then you get only one main lobe in vertical direction without side lobes.
Only very few driver combinations can comply with this. In the case of the Nubert nuvero 170 mentioned above, this would be a midrange to tweeter crossover frequency of 1000Hz.

To control the horizontal radiation, it hardly matters whether an M-T-M or M-M-T concept is used - but an LS radiates not only horizontally or vertically, but also in between ("diagonally"), so in theory symmetrical concepts have advantages.

In reality, one tries to use a D'Appolito (DA) or a multiple DA concept (W-M-T-M-W) to take advantage of the disadvantages due to the vertical driver to driver interference to keep the vertical radiation narrow and accepts the comb filter effects of the midrange driver to driver interference for this.

How even or uneven the vertical radiation is in the case of the Nuvero 170 is difficult to estimate. The frequency responses could look pretty wild at larger vertical measurement angles, but at low frequencies it becomes easier to meet the "DA distance rule".

To give a practical example....
My next but one project (should I ever have time for it) uses a DA array to force a very tight vertical dispersion. For this I accept comb filter effects in the vertical radiation. Unfortunately, there are also a few side lobes in the vertical radiation.

In return, an LS with +-90° horizontal and +-30° (-6dB limit) vertical dispersion can be realized. On average, this results in a loudspeaker with very even radiation, even if the details are not perfect at all. The crossover frequency to the tweeter is 3kHz.

 

[abdo123]

Perfect no, theoretically a good vertical radiation is possible with D'Appolito (DA) concept, as has already been said, it depends on the driver spacing (M-T as well as M-M) and the crossover concept to control the vertical radiation lobes.

In the original DA concept, the distance between the centers of the midrange drivers must not exceed 2*lambda/3 of the crossover frequency when using a third order crossover.
Then you get only one main lobe in vertical direction without side lobes.
Only very few driver combinations can comply with this. In the case of the Nubert nuvero 170 mentioned above, this would be a midrange to tweeter crossover frequency of 1000Hz.

To control the horizontal radiation, it hardly matters whether an M-T-M or M-M-T concept is used - but an LS radiates not only horizontally or vertically, but also in between ("diagonally"), so in theory symmetrical concepts have advantages.

In reality, one tries to use a D'Appolito (DA) or a multiple DA concept (W-M-T-M-W) to take advantage of the disadvantages due to the vertical driver to driver interference to keep the vertical radiation narrow and accepts the comb filter effects of the midrange driver to driver interference for this.

How even or uneven the vertical radiation is in the case of the Nuvero 170 is difficult to estimate. The frequency responses could look pretty wild at larger vertical measurement angles, but at low frequencies it becomes easier to meet the "DA distance rule".

To give a practical example....
My next but one project (should I ever have time for it) uses a DA array to force a very tight vertical dispersion. For this I accept comb filter effects in the vertical radiation. Unfortunately, there are also a few side lobes in the vertical radiation.

In return, an LS with +-90° horizontal and +-30° (-6dB limit) vertical dispersion can be realized. On average, this results in a loudspeaker with very even radiation, even if the details are not perfect at all. The crossover frequency to the tweeter is 3kHz.

View attachment 257781View attachment 257783 View attachment 257784 View attachment 257785 View attachment 257786

This design appears incredibly ideal for small rooms. I'm very interested in trying it out if you ever feel like sharing it with other people.

Right now it appears as if there are only three drivers involved? Is there a reason not to persue an WMTMW configuration?

 

[fpitas]

This design appears incredibly ideal for small rooms. I'm very interested in trying it out if you ever feel like sharing it with other people.

Right now it appears as if there are only three drivers involved? Is there a reason not to persue an WMTMW configuration?

Done all the time. It gives vertical directivuty to fairly low frequencies, which can help with room modes.

 

[abdo123]

Done all the time.

I don't think I've ever seen it done this cleanly before, the small iLoud MTM monitor comes close

 

[fpitas]

I don't think I've ever seen it done this cleanly before, the small iLoud MTM monitor comes close

The ones I've seen are rather large, standing about 6 feet tall (180cm?), with 12" or 15" woofers.

 

[abdo123]

The ones I've seen are rather large, standing about 6 feet tall (180cm?), with 12" or 15" woofers.

Like this?

 

[fpitas]

Like this?

Your link needs repair...

 

[ctrl]

Right now it appears as if there are only three drivers involved? Is there a reason not to persue an WMTMW configuration?

It is a top speaker for a bass module (crossover frequency 150-350Hz) and six driver per top speaker are used - and some of the usual "tweaks" are involved (e.g. 2.5-way concept).

I also simulated different WMTMW variants, the extra effort for this (would have allowed crossover frequency around 80Hz) and the mediocre results did not convince me.

Done all the time.

It wasn't quite that simple for me
A wide, controlled horizontal radiation as early as possible (<<1kHz) and up to 10kHz (not quite fulfilled) and a uniform narrow vertical radiation, preferably without severe side lobes*** (almost fulfilled) without waveguide or horn was a challenge for me.

***The side lobes above 9kHz could almost be seen as positive, the slight side lobes around 2.7kHz should not be a problem, but are not nice because they are in the area of the ear canal resonance.

the small iLoud MTM monitor comes close

Like this?

The example illustrates well, the possible problems with mono and multiple DA concepts.

The comb filter effects are not a problem as long as they do not affect the DI/ERDI too much.
Severe vertical side lobes are problematic because they can contribute to abrupt tonal changes (humps in the frequency response of sound power and early refection).


It would be extremely interesting to see a loudspeaker like the Nuvero 170 measured using NFS. To see how well or not so well the concept has turned out in terms of vertical radiation.

 

[abdo123]

It is a top speaker for a bass module (crossover frequency 150-350Hz) and six driver per top speaker are used - and some of the usual "tweaks" are involved (e.g. 2.5-way concept).

the crossover filters do not really show six drivers (nor does the sensitivity ).

Nevertheless this is the ideal speaker in a domestic setting for me so I'm very excited to see how this goes. (adding cardioid slots would be the icing on the cake)

 

[anphex]

You know what guys, I'll whip out a ruler, my measurement setup and make some measurements in front of the NuVero 170 tweeter on different vertical angles because I need the proof. Considering that it us easier to get accurate in room measurements for higher frequencies than for lower ones, I think the results would be pretty valid.
I can't make such a nice smooth graph though but an Excel sheet haha.

1 meter distance should be fine, right? Or should I be further away since it's a pretty tall(170cm) speaker?

 

[thewas]

I think if you don't care much about the bafflestep (correction) you could even measure closer to make the gate longer.

 

[anphex]

Sorry for stupid, but "gate"?

 

[fpitas]

Sorry for stupid, but "gate"?

Time gating for time windowed measurements

 

[thewas]

You place a gate/window in time domain usually before the before reflection comes so you get anechoic measurements.

 

[Vladimir Filevski]

... make some measurements in front of the NuVero 170 tweeter on different vertical angles...

1 meter distance should be fine, right? Or should I be further away since it's a pretty tall(170cm) speaker?

Horizontal angles would be more fruitful.
2 meters away, especially for vertical angles.

 

[anphex]

Thank you all. I'll research how to do this in REW this weekend.

@Vladimir Filevski Since there are no chassis "in line" with this axis there will be no problem. I'd even go so far and say that the NV170 is the widest dispersing speaker there is.
Still, it's an idea. I'll do those measurements too.

 

[restorer-john]

What about the BMR enclosures sitting behind the drivers? They will be seriously adhered from the inside of the speaker.

 

[anphex]

What about the BMR enclosures sitting behind the drivers? They will be seriously adhered from the inside of the speaker.

Nah it's just a damping pad and four wood screws into the super massive enclosure front "sail" (it's slightly curved).
I already removed them a few times, so this wouldn't be a problem.

 

[anphex]

Hey guys, I promised to do some measurements and here they are!

I attached a picture of the 4ms gated results. I chose to align the measurement mic head in 1000mm distance and then to the height of the chassis so it's easier to relate, maybe.
Here are the distances and angels:
Tweeter - upper BMR: 12cm and about 7°
Tweeter - upper midwoofer: 26cm and about 15°
Tweeter - upper subwoofer: 46cm and about 25°

I also attached the raw measurement file and a picture of the setup so you can tell me some mistakes I may have made. But I think the results are pretty valid.
Edit: File too large for ASR it seems, so here's a link to my private secure cloud: https://anphex.one/index.php/s/X58G4Ecib3meyix

As for the dispersion: Yeah, horizontally this speaker is golden, but vertically turn out we really have a problem at the crossover I kept talking about.
I will have to debate with myself wether this is an acceptable tradeoff for the low end extension and horizontal dispersion. But considering it's a 4-way speaker specifically designed for wide dispersion it is a bummer.

Makes me think whether my idea of swapping the BMR the way I suggested would actually bring an improvement.

I am going to keep this setup like this for a while so I can quick new measurements with your suggestions.

Edit2: Well maybe I'll just go ahead and do a little swapping and measure right away as long as the setup is there? It will probably take only a few minutes.

 

[anphex]

Swap done.

Result: Pretty much the same! Though now without D'Appolito setup in the BMR. Seems like the Tweeter and BMR don't add up well off axis, but I think this is a problem in all non-coaxial speakers. The only ones who do this perfectly are probably the 3-way-Genelecs, and those are coaxial.