Simulation Analysis - Screws in Waveguide

[JustAnandaDourEyedDude]

The isoparametric quadratic setup ensures that both geometry and sound field are resolved to second-order. Both the tweeter surface and screw surfaces are resolved with 24(!) elements per wavelength. A convergence study is beyond the scope of this analysis, but for a client I would probably dive a little deeper into certain aspects of this. The same mesh can be used, because I apply internal hard walls on the screws and turn that on and off, respectively. Otherwise the meshes will differ a little, so this is the best that can be done.

Thanks for the clarifications. I have not kept up with Comsol (I attended a one-day hands-on workshop on it about 25 years ago; the instructor was one of the developers at the time, flew from Sweden to the US for a week to lead the workshops). So I conclude that there is mesh clustering at the boundary that gives 24 elements per wavelength, and that any curved part of the boundary is handled accuractely due to the isoparametric elements. I guess turning off the internal hard walls in the screws then makes them transparent to the acoustics. Great, builds confidence in the results, thanks!

P.S. Typo: "fascilitate" should be "facilitate".

 

[René - Acculution.com]

Thanks for the clarifications. I have not kept up with Comsol (I attended a one-day hands-on workshop on it about 25 years ago; the instructor was one of the developers at the time, flew from Sweden to the US for a week to lead the workshops). So I conclude that there is mesh clustering at the boundary that gives 24 elements per wavelength, and that any curved part of the boundary is handled accuractely due to the isoparametric elements. I guess turning off the internal hard walls in the screws then makes them transparent to the acoustics. Great, builds confidence in the results, thanks!

Glad to have clarified, and thanks a lot for your interest and kind words.

 

[GWolfman]

Excellent work! Promoted to home page.

Wow, was this a first (i.e., someone other than @amirm getting front page coverage)? At least it is for me (to witness). Congrats @René - Acculution.com

 

[GWolfman]

This is great and all but inconsiquential the moment I fire up an MQA track (guaranteed to hear what the producer and/or artist heard at the studio).


sorry I couldnt help myself

Lolz

 

[Frank Dernie]

- Introduction -

Erin (of Erin's Audio Corner) has expressed an interest in knowing the influence of screws in waveguides; if they protude from the waveguide, can the sound be affected? I have investigated this issue using a numerical simulation software. This is not an exhaustive analysis, but more of a fun weekend project.

- Simulation setup -

Software: COMSOL Multiphysics
Method: Finite Element
Geometry: Realistic 1" tweeter geometry, arbitrary waveguide (5 cm radius, 2 cm depth), 4 screws (0.67 cm diameter, protuding 1.2 mm), in baffle
Mesh: Quadratic Lagrangian elements (6 elements per wavelength @20 kHz)
DOFs: ≈ 262,000
Other: The exact same mesh was used for both cases; with and without screws. Calculation time on laptop around 1 hour per case. This is pure acoustics case, so there is no explicit solid mechanics; instead an acceleration is put on the tweeter surface. Exploited quarter symmetry.

View attachment 127320


- Observations -
The sound pressure level was evaluated with and without the screws, both in a 1 m distance and at 5 cm. The difference is neglible for both distances, and at the most relevant distance of the two (1 m) the difference is less than 0.5 dB. Even in the near field right around a screw the field is largely unaffected, but the overall influence of screws is larger than what I would have initially guessed...

SPL at 5 cm and at 1 m: (Yes, yes, this is not a good design, but it does not matter for the objective at hand, other than that relative difference typically show up more around dips and peaks as is also the case here; move a resonance slightly and you can get large differences.)
View attachment 127322


SPL difference, 1 m ('Screws' minus 'No screws'):
View attachment 127323

Sound pressure difference ('Screws' minus 'No screws') at 20 kHz visualized:
View attachment 127332

- Conclusions -
For this particular case the influence of the screws in fairly minimal, yet it is there. The wavelength even at 20 kHz is 1.7 cm, which is larger than the screws themselves, and so it is to be expected that the sound field will not be affected much. It should be noted that such an analysis should be carried out on a case by case basis, as waveguides (and screws) differ from product to product, and if we get near 1 dB difference it could be relevant. It is a computationally cheap single-physics analysis, and so companies with simulation capabilities should at least do this investigation internally, and provide results to those who request it.

/René

- About me -
BSEE, MSc (Physics), PhD (Microacoustics), FEM and BEM simulations specialist in/for loudspeaker, hearing aid, and consultancy companies. Own company Acculution, blog at acculution.com/blog

Splendid to see this sort of analysis here.
A friend of mine has been mechanical and magnetic circuit modelling for speaker manufacturing clients for over 20 years but it is his business so commercially confidential, both what he is modelling and the techniques. Even the clients absolutely do not want it known that it is not they themselves doing it!

 

[René - Acculution.com]

Splendid to see this sort of analysis here.
A friend of mine has been mechanical and magnetic circuit modelling for speaker manufacturing clients for over 20 years but it is his business so commercially confidential, both what he is modelling and the techniques. Even the clients absolutely do not want it known that it is not they themselves doing it!

Thanks Frank. Yes, I do the same analyses (Structural mech, anisotropic composites, magnetics) for clients, and it is often not seen who did what ;-) I think this forum is good for at least showcasing 1) what can be done, and 2) what consumers should perhaps 'demand' from the manufactures (at the very least more measurements than what is currently the norm)

 

[StuartC]

Hello All,

There is nothing here.

This is as it says a Saturday afternoon musing with some simulation graphics, a graphic what if.

I am more impressed with @amirm ‘s measurements, with and without a “transparent” grille cloth.

Thanks DT

I don't understand the context of this comment. Are you questioning the value of simulation tools? Just curious.

 

[René - Acculution.com]

Rene,
Thank you for taking the time to do this for me. It's good to see some numbers instead of just "I think". As I go forward, when I encounter speakers with waveguide assemblies I suspect may be a problem I will reach out to you with physical measurements and see if you have the time to model them. I'd like to test this myself at some point as well and maybe we can see how the prediction through simulation lines up with measurements.

I do wonder if the symmetry of the placement will also effect things. For example, maybe the response would have more of an on-axis issue but off-axis the symmetry is OK. And vice versa. ??

Sure Erin, and if you are interested in more insight as to what you can (and cannot) do with simulations in general, you know how to reach me.

I looked at the difference between 1 m distance, 45 deg off to the side along a plane in which a screw resides, compared to 1 m, 45 deg, plane in between screws:

and the SPL difference is:

So a couple of percent difference (provided I haven't made any mistakes, the waveguide geometry is perfect, etc. ect). For larger obstructions the simulations and measurements are bound to match well for this type of acoustics problem, but I don't know if it worth caring about screws. I do think that a company that shows that they have INVESTIGATED these things, be it via sim or meas, could attract costumers that are interested in these technical aspect, so there is a 'business case' in this scientific approach.

 

[René - Acculution.com]

Anyway, cabinet edge diffraction is definitely up next, but I wonder if even more general aspects are of interest? E.g. Phase (minimum vs linear, polarity, allpass, effect on response near crossover frequency, ringing, group delay, phase delay) is a topic where I see a lot of confusion? Eigenmodes (room modes, break-up in loudspeakers, modal analysis)? Non-linear vs Linear (non-linear geometry, n-l materials, n-l contacts, n-l electromagnetics? I touch upon these topics on my own blog, and have planned to make YouTube videos on Engineering Mathematics, when I can find some time, but what would YOU like to hear about?

And if anyone knows how to include LaTeX code, let me know (I tried below, but it does not seem to work)

<!DOCTYPE html>
<html>
<head>
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width">
  <title>MathJax example</title>
  <script type="text/javascript" async
  src="https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.4/MathJax.js?config=TeX-MML-AM_CHTML" async>
</script>
</head>
<body>
<p>
$$k_N^2=\left(\frac{\omega_N}{c}\right)^2=k_x^2+k_y^2+k_z^2$$
</p>
</body>
</html>

 

[René - Acculution.com]

For dome tweeters a basic circular arc seems to work pretty well and Ath has a nice function to be able to model the dome and surround as part of it's mesh generation which was added in later versions. That can stay axisymmetric or be modified around the outside to follow other functions and generate ellipses etc. The other main option uses Marcel's superformula to generate a closed form solution mating an OS throat to a Euler spiral / clothoid termination. Changing the parameters lets you go from pure OS to almost JMLC or anything in between.

I am quite excited by the freestanding guides with rollback as in the BEM simulations I have run they seem to be much less affected by surrounding drivers or cabinets. I modelled the cone of a woofer in a box below the guide as a rigid boundary vs the same guide in free air and the difference was surprising. Top line is with woofer cabinet, bottom is without. A waveguide in a baffle with a woofer modelled as a rigid boundary in the baffle was much worse than either of these.

I might look into it, but I have many other topics waiting ;-) Too little time for it all...

 

[Objectivist01]

On this Topic, I had a vintage/old speaker from ALR acoustics Germany on whihc the tweeter waveguide had a recessed edge which had the screws. There was a a foam ring over this screws which in turn was flush with the front baffle. I found that a great approach to hide the screws from the waves. I belive the rings were there for some acoustical reason too.

 

[freddi]

fascinating thread - I have a request for a simulation of something with "more in the way" than a screw head. It may appear a frivolous request but besides my hobby, might be a good illustration of what can be done without building a physical prototype:

https://www.diyaudio.com/forums/software-tools/368826-adept-akabak3-sim-dutch-k12.html

 

[jonfitch]

Well I suppose KEFs solution here is the slightly better one in terms of absolute performance when installed correctly compared to Revel or Buchardt, but I don't get why it's removable at all. They should just have it superglued on out of the factory instead of letting people take it out and play around with it.

 

[Koeitje]

Well I suppose KEFs solution here is the slightly better one in terms of absolute performance when installed correctly compared to Revel or Buchardt, but I don't get why it's removable at all. They should just have it superglued on out of the factory instead of letting people take it out and play around with it.

And what if your tweeter breaks? You cannot replace it? Lets not do this okay.

 

[fluid]

I might look into it, but I have many other topics waiting ;-) Too little time for it all...

Available time is my enemy too, for an overview the user guide https://at-horns.eu/release/Ath-4.7.0-UserGuide.pdf is worth a look. It is a time drain to learn a new tool, waveguided speakers are my current interest and I have sunk a huge amount of time into it. BEM sims have provided a lot of insight to me about general acoustic issues so I'm sure the nice graphics that comsol can produce will help to demonstrate some of these issues and possible solutions to a wider audience. Any time Earl Geddes is impressed it's worth a look

 

[edechamps]

Or buy speakers designed and engineered to be used with grills on, which is what I do.

I like Genelec's approach: each driver has its own sturdy, rigid metal grille that cannot be removed. This way you don't have to choose between sound quality and protection, and you don't have to ask yourself if they were measured with or without the grille, or if they would sound better without it.

@René - Acculution.com This is incredible work by the way. It's one thing to see speaker manufacturers brag about "simulation" and show off with renders. Seeing it done live and in detail by a professional is something else entirely!

 

[Frank Dernie]

Thanks Frank. Yes, I do the same analyses (Structural mech, anisotropic composites, magnetics) for clients, and it is often not seen who did what ;-) I think this forum is good for at least showcasing 1) what can be done, and 2) what consumers should perhaps 'demand' from the manufactures (at the very least more measurements than what is currently the norm)

The modelling I was interested by was calculating enclosure vibration and radiation as a proportion of the whole output.
AFAIK this can only be calculated since something like the Klippel measurements have no way of separating any output from the enclosure from that from the drivers.
I have been told the effect is important but never seen it quantified.

 

[DerRoland]

It's yet sunday morning and another very interesting aspect of loudspeaker design is born on asr, thanks René.

As a person with less science knowledge I asked myself, what influence does a protection metal mesh in a common compression driver like this one have?

(BMS4550)

The manufacture data sheet promise good sounding quality, why is this possible with that kind of audio obstraction?

 

[René - Acculution.com]

I like Genelec's approach: each driver has its own sturdy, rigid metal grille that cannot be removed. This way you don't have to choose between sound quality and protection, and you don't have to ask yourself if they were measured with or without the grille, or if they would sound better without it.

@René - Acculution.com This is incredible work by the way. It's one thing to see speaker manufacturers brag about "simulation" and show off with renders. Seeing it done live and in detail by a professional is something else entirely!

Thanks a lot. This is certainly one of the simpler (simple vs complex) simulations to set up (does not mean that the operator can't still make mistakes); it takes some effort to set up but it is easier than a lot of other simulation set ups (easy vs difficult), in the sense that I could easily find students that could replicate this, whereas a complete speaker setup would be impossible for them to do.

 

[René - Acculution.com]

It's yet sunday morning and another very interesting aspect of loudspeaker design is born on asr, thanks René.

As a person with less science knowledge I asked myself, what influence does a protection metal mesh in a common compression driver like this one have?

View attachment 127615(BMS4550)

The manufacture data sheet promise good sounding quality, why is this possible with that kind of audio obstraction?

Hi. BMS probably know exactly what they are doing, so I would not question that. I imagine that the mesh cannot be removed and so it is designed for having this on always. I imagine that removing it would result in certain peaks being ever so much higher, as the mesh is probably mostly resistive. If I were given the mesh geometry, this would be very easy to investigate.