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

[anphex]

Hey everyone,

have a look at the attached picture. These are my speakers and I love them. BUT in my opinion they have a flaw in the d'appolito design considering the BMR-Tweeter-BMR arrangement.

The distance of those BMR from each other around the tweeter center-to-center is 23cm. This would imply a maximum crossover frequency of 1450 Hz for d'appolito. The real crossover is around 2000 Hz. I think this is mostly a visual appearance decision.

But here's the thing: technically it's no big deal to swap the lower BMR and the tweeter. You can even swap the cables so the crossovers stay correct.
So my thought was switching BMR-Tweeter-BMR to BMR-BMR-Tweeter. This would result in a distance of 12,5cm and a theoretical maximum crossover frequency of about 2700 Hz which is more or less ideal for dispersion. I know it's not a real d'appolito anymore but I think there are only benefits to gain from this. Also the tweeter would be lower and more on axis with the ears when sitting down.

Is there a flaw in my thoughts?

Edit: Noteworthy - the small Low-Mid-Chassis above and below the BMR have their crossover at 450 Hz and have a distance of 60cm which implies are crossover frequency of 550 Hz, so these are placed pretty ideal and there's nothing to worry it seems.

 

[voodooless]

Is there a flaw in my thoughts?

One of the MBRs is just as far away from the tweeter as the old situation. The other one is much further away. This will not make it better, but worse.

At 2kHz, tweeter to mid spacing can be up-to 17 cm. Looks like that is no problem. In any case, it's only a true D'Appolito has 3rd order crossovers. Not shure what you have here.

 

[abdo123]

@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?


Also i don't think the below statement is correct because you will lose the point source radiation (but not the sensitivity) both of which are why MTM configurations are made to begin with. Might as well buy a speaker with 5 drivers instead of 8 next time.

technically it's no big deal to swap the lower BMR and the tweeter.

 

[Vladimir Filevski]

That is Nubert NuVero 140, with exemplary flat frequency response. DO NOT CHANGE ANYTHING!

 

[fpitas]

@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?

A line array comes pretty close to that ideal, at least in the far field.

 

[anphex]

That is Nubert NuVero 140, with very neutral sound and exemplary flat frequency response. DO NOT CHANGE ANYTHING!

Sorry to burst the bubble, but while Nubert is pretty controlled in FR, they aren't flat. On axis they have pretty strong shelving boost of several dB leading up to 7500 Hz. Also the super low end is boosted too. So they come with a slight controlled loudness curve. And I say that as a fanboy. But it's easily fixable with DSP.

 

[abdo123]

A line array comes pretty close to that ideal, at least in the far field.

yeah but that needs some sort of delay to work so that all the drivers arrive in-phase to the listener.

 

[anphex]

One of the MBRs is just as far away from the tweeter as the old situation. The other one is much further away. This will not make it better, but worse.

At 2kHz, tweeter to mid spacing can be up-to 17 cm. Looks like that is no problem. In any case, it's only a true D'Appolito has 3rd order crossovers. Not shure what you have here.

So it will be worse because the further BMR will cause interferences? And yes, the BMR-Tweeter-BMR crossover is 3rd order18 dB.

 

[fpitas]

yeah but that needs some sort of delay to work so that all the drivers arrive in-phase to the listener.

That's all established science at this point. Lots of DIYers do it successfully.

 

[anphex]

One of the MBRs is just as far away from the tweeter as the old situation. The other one is much further away. This will not make it better, but worse.

At 2kHz, tweeter to mid spacing can be up-to 17 cm. Looks like that is no problem. In any case, it's only a true D'Appolito has 3rd order crossovers. Not shure what you have here.

Did you edit the second line in? I was really confused haha.
Yeah the distance from BMR to Tweeter is about 11,5cm, so it's fine then as long as the crossover is 3rd order?

 

[voodooless]

Yeah the distance from BMR to Tweeter is about 11,5cm, so it's fine then as long as the crossover is 3rd order?

I would say it might be fine regardless of crossover topology. Time has told us that other crossover topologies can work as well, you just can't call them D'Appolito, even if many still do .

But the proof of the pudding is in the eating, dig up some off-axis measurements. That should tell you if everything is fine or not.

 

[fpitas]

Did you edit the second line in? I was really confused haha.
Yeah the distance from BMR to Tweeter is about 11,5cm, so it's fine then as long as the crossover is 3rd order?

Yes, you can see for yourself; a quadrature-aligned 3rd order Butterworth filter is the first crossover that D'Appolito used.

Tolvan Data

Later he went to LR4.

 

[anphex]

As for measurements - no, I measured them once on my listening position on axis and then on an when-sprawling-legs-up-on-couch-head-level. I expected a dip around said frequencies but everything looked fine. It was about 15° degrees difference at 3 metres. So I guess I keep looking for demons in heaven.

 

[fpitas]

As for measurements - no, I measured them once on my listening position on axis and then on an when-sprawling-legs-up-on-couch-head-level. I expected a dip around said frequencies but everything looked fine. It was about 15° degrees difference at 3 metres. So I guess I keep looking for demons in heaven.

Yes; if it's not broken, why fix it?

 

[Vladimir Filevski]

Sorry to burst the bubble, but while Nubert is pretty controlled in FR, they aren't flat. On axis they have pretty strong shelving boost of several dB leading up to 7500 Hz. Also the super low end is boosted too. So they come with a slight controlled loudness curve. And I say that as a fanboy. But it's easily fixable with DSP.

Frequency responses measured by Audio magazine and AudioVision magazine are very flat:
https://www.nubert.de/media/3f/82/b3/1657108669/test-2015-09--nuvero-140--set--audiovision.pdf
https://www.nubert.de/media/71/5e/75/1657108668/test-2019-12--nuvero-140--audio.pdf
Changing drivers position require modifying crossover, which is not possible without deep knowledge of crossover design and appropriate measuring equipment. Even then, results may be disappointing.

 

[voodooless]

Frequency responses measured by Audio magazine and AudioVision magazine are very flat:
https://www.nubert.de/media/3f/82/b3/1657108669/test-2015-09--nuvero-140--set--audiovision.pdf
https://www.nubert.de/media/71/5e/75/1657108668/test-2019-12--nuvero-140--audio.pdf
Changing drivers position require modifying crossover, which is not possible without deep knowledge of crossover design and appropriate measuring equipment. Even then, results may be disappointing.

This is of more interest:

 

[Vladimir Filevski]

Thanks for bringing this out. On-axis frequency response is very flat, but horizontal off-axis has dip at about 1000 Hz (which is strange!). Swapping position of BMR and tweeter will not fix it!

 

[anphex]

Thanks for bringing this out. On-axis frequency response is very flat, but horizontal off-axis has dip at about 1000 Hz (which is strange!). Swapping position of BMR and tweeter will not fix it!

You can even see the peak I mentioned above lol. It's not severe, but it's there. I'll post my measurements later that I made with a calibrated measurement mic and a super long M4 Sweep in REW.

Also note that I have the 170, not the 140 which is a pretty different 4-way-system instead of 3,5 way.

 

[thewas]

Thanks for bringing this out. On-axis frequency response is very flat, but horizontal off-axis has dip at about 1000 Hz (which is strange!). Swapping position of BMR and tweeter will not fix it!

Please keep in mind that above plot belongs to the (pseudo-)D'Appolito centre loudspeaker so it would rather correspond to the vertical directivity of the floorstander.

 

[Vladimir Filevski]

Oh, that is why it looks to me so strange. My knowledge of German language is below zero, so I didn't figure out it is the off-axis response of the centre loudspeaker. Yes, it would be the same as vertical response of the floorstander, so it is OK. There are no horizontal off-axis measurements of that floorstander test from the AudioVision magazine, as I can see.
I have noticed that most of the Nubert loudspeakers have very flat frequency response, and 30 degrees horizontal and 10 deg. vertical off-axis responses from Audio or Stereoplay magazine measurements looks very good also. Nothing to improve here.