Mid-dome directivity mismatch? (Vituixcad simulation)

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all,

For a 3-way project I would like to try a mid-dome. I have used the Dayton RS52AN-8 to simulate the response in Vituixcad.
But I run into a problem with the directivity, where I would like your insights on.

To describe troubleshoot the directivity, I will post the procedure I used to come to the results. Imaged from the results can be found below.

First:
1. I used the measurements from Dayton Audio RS52AN-8 | HiFiCompass
2. In Vituixcad --> used SPL-trace to get the .frd file.
3. Simulate half-space response from Enclosure tool in Vituixcad. I used two possible designs. 2 SB17NAC-8 in parallel (2x 6") or a Dayton RS225-8 (8").

4. Simulate off-axis response for the woofers via the Diffraction tool in Vituixcad. A guide can be found here: YouTube
5. As a tweeter I used the Vifa NE19VTS-04, also SPl-traced from the measurements at Hifi-compass. Vifa NE19VTS-04 | HiFiCompass
6. Simulate the speaker response in Vituixcad with EQ and linear-phase XO's. XO's are at 667 hz and 4000 hz.

The result can be seen in the pictures below. I marked the directivity problems in red boxes.
The results is practically the same when I used the infinite-baffle measurements for the mid-dome and tweeter.

Now I wonder, is this problem the result of my procedure or caused by using measurements in a wrong way.

Maybe I don't understand how mid-domes should be used :). Anyway, I hope you experienced guys are willing to teach me.
 

Attachments

  • middome simulatie directivity probleem.png
    middome simulatie directivity probleem.png
    199.2 KB · Views: 334
Hi, you show only horizontal off-axis responses. Check also verticals, they will show more directivity in xo area thanks to lobing. Some compensation there.

4-5" cone midrange will have more directivity around 2kHz. My favourite mid-size! Baffle shape and edge contouring will have some effect too.

Zaph doesn't show measured off-axis responses, but I expect them to follow your VCAD sim http://www.zaphaudio.com/ZDT3.5.html
 
Hi, you show only horizontal off-axis responses. Check also verticals, they will show more directivity in xo area thanks to lobing. Some compensation there.

4-5" cone midrange will have more directivity around 2kHz. My favourite mid-size! Baffle shape and edge contouring will have some effect too.

Zaph doesn't show measured off-axis responses, but I expect them to follow your VCAD sim http://www.zaphaudio.com/ZDT3.5.html
The link does'nt work for me, unfortunately.

I want to elaborate on the question: What is the ideal usecase for a mid-dome?
Because the mid-dome is very wide in it's passband, let's say 600-6000hz. And so it radiates the same SPL off-axis as front axis.

But a woofer already radiates less off-axis because of it's size.
That's my explanation as to why the simulation does'nt show a good match, i.e. it widens and narrows at the crossover frequencies.

This I don't understand because I believe there are excellent measuring speakers using mid-domes that don't show this behaviour. It makes me think the problem is caused by how I come to this simulated response or how Vituixcad handles domes
I hope you can help me with my thought process.
 
Member
Joined 2014
Paid Member
Sim your mid-dome at 800,900, and 1000hz.
As domes are displacement-limited they are more comfortable crossed closer to 3x fs rather than 2x.
As it happens, this will help match directivity a bit.


You can also calculate the lowest cross for your driver at the max spl you want.
 
Arnand, please show your baffle and speakers in it as image (screenshot from VCAD)

Yes directivity around W_M xo looks strange, but we don't see angles, is it up to 90?

Zaph ZDT3.5 link works for me in Chrome and MS Edge
Here you go.

The sim is up to 60 degrees yes. It is limited because I use the infinite-baffle measurements SPL-traced from the HifiCompass measurements.
 

Attachments

  • diffraction 2x 6 inch.png
    diffraction 2x 6 inch.png
    60.9 KB · Views: 272
  • diffraction 1x 8 inch.png
    diffraction 1x 8 inch.png
    58 KB · Views: 273
Sim your mid-dome at 800,900, and 1000hz.
As domes are displacement-limited they are more comfortable crossed closer to 3x fs rather than 2x.
As it happens, this will help match directivity a bit.


You can also calculate the lowest cross for your driver at the max spl you want.
This only worsens the directivity causing an even bigger dip because of the off-axis response from the woofers.
 
I did VCAD sims for the double woofers with ideal flat response, and VCAd calculating off-axis at 0, 30, 60 and 90 deg

Looks like using a 2" dome is not a good idea. The volt 3" dome in waveguide should be a fraction better.

A simulation shows response around 3-600Hz much better than indoor measurements!
 

Attachments

  • arnand vs flat onaxis.jpg
    arnand vs flat onaxis.jpg
    289.7 KB · Views: 229
  • arnand flat 0 30 60 90 hor.jpg
    arnand flat 0 30 60 90 hor.jpg
    251 KB · Views: 228
Here is my measured response of a proto with double 8" SS woofers, a 4" Audax HM100 mid and ribbon tweeter with waveguide (Fountek Neo3.5H)

Horizontal response smooth as can be! Wiggles come from recflections indoors- a simulation would look much better!
 

Attachments

  • 3wdsp  spl 0-90deg 12ms 16.jpg
    3wdsp spl 0-90deg 12ms 16.jpg
    168.5 KB · Views: 73
  • 3whifi 0 30 60¤ 6ms 124.png
    3whifi 0 30 60¤ 6ms 124.png
    82.1 KB · Views: 80
Last edited:
I did VCAD sims for the double woofers with ideal flat response, and VCAd calculating off-axis at 0, 30, 60 and 90 deg

Looks like using a 2" dome is not a good idea. The volt 3" dome in waveguide should be a fraction better.

A simulation shows response around 3-600Hz much better than indoor measurements!
Then I wonder. What would be a good usecase for this Dayton mid-dome then?


Which mid-domes would work in a band from 700-4500 hz? I mean that in a sense the directivity does match. That Volt dome is crazy expensive :eek:
 
That is because baffle area/width gets smaller (virtually) when offset angle gets higher. We start to see reduction above freq when baffle reinforcement makes a peak in response. This is minor vs. beaming of a cone driver when it's other side gets in "shadow". This shadowing is not happening in domes, but force vector effect happens because it's not a compressing/expanding ball - it moves only in 2D!

A box will compensate for that a little, but simulations typically don't have a virtual box (at least user can't set sidewall depth). Also simulations use flat piston model for radiator.

https://www.klippel.de/fileadmin/_migrated/content_uploads/KLIPPEL_Sound_Radiation_Poster_01.pdf
 
Last edited:
Boden - sorry English is my second language. Yours too?

"Offset angle" means that loudspeaker is (measured) horizontally offaxis, in context of this thread about horizontal off-axis measurements/response.

Klippel pdf gives background data for my text in message, but sorry I can't open up mathematic formulas of it.
 
I can not say i ever have seen in my own measurements for a 6.5” any of such narrowing of dispersion at so low a freq as 600 hz

Many cad programs seems to be so advanced with so many features and options that only few seems to understand them, i would like to see much simplier cad programs with few options and that simulates only ideal drivers on a flat baffle, then learn from the results and apply this knowledge in the real world.
Simulating the wrong things without knowing can make you draw very bad conclusions, and without knowing
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.