Waveguide model

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

Hello everyone. I recently started to use ABEC 3 to model a mid range horn and adjust its directivity pattern. I started by experimenting with the examples and reading the published papers/ help section. I have also familiarized myself with Akabak. However, I still do not know if I am going in the right direction. I would be very grateful if you could give me some feedback for my plan.

My plan is to measure the Frequency Response, Impedance and directivity of the driver practically. I wish to use this information in ABEC3 with 'Measured Radiator'. I am doing this so that I will not have to model the driver in ABEC. I can then model a suitable horn. I am finding it difficult to do it though. I am only using the BEM and LE modules. The horn has been modelled in BEM and I have used 'Pressure_box'. I am getting the error:"Driving group 1001 cannot be linked".

I wish to know if I am going in the right direction. Would it be better if I model the driver in Akabak, then use it with the horn in ABEC 3? Any help would be greatly appreciated.
 
Last edited:
Help Section info

I found this information from the 'help' tab:
'The MeasRadiator is linked to a section Pressure_Points or Pressure_Box of the Boundary Element or Direct Sound scripts.'
However I also found this from the 'LE Systems Example 1' of the help section:
'Def_MeasRadiator can be used in Direct Sound Analysis only.'

I read that the direct sound module ignores reflections and diffraction. Hence, I don't think it is suitable for modeling a horn.
I guess modeling a horn would be 'too easy' if I could just measure the driver directivity and use it to simulate the horn's response!

Anyway, good to know that I have company.
 
From what I've read the direct sound portion is quite limited and as you say not suited for analyzing the 3d sound field. Full BEM is necessary here and you'll need the driver parameters for the lumped element section. For analyzing the directivity of a mid horn I think "close enough" would get you there. You might even use the parameters for the driver they use with the mid horn example.
 
Agreed

Thanks Nate. I guess that is a good way to start. I am now trying to decide which software to use to draw the horn. I cannot draw it in ABEC as one of its parts is quite complex. I cannot seem to find proper instructions on how to import a geometry. Do you know how to do that?
 
I've never used ABEC myself other than playing with the examples so I'm not of much use. The best tutorial I've found is this one: http://www.diy-hifi-forum.eu/forum/archive/index.php/t-11174.html. It's in German I think so the translation is a bit iffy but it looks comprehensive. A couple diya members contribute to that thread so hopefully they see this. In that thread they get into bringing models in and mention some free CAD packages. I could see about drawing something for you if you want since I'll need to figure it out myself.
 
If what you need ends up being the driver parameters, I can probably help you to get those as they are never made public by the manufacturers. I developed some techniques that work for measuring them without having to destroy the driver. Let me know if you want to do that and I'll tell you how I did it.
 
A compression driver is just a speaker playing into a small ported box (when it is playing without its horn. I modeled this using the standard techniques and then fitted the parameters (mass, compliance, BL, etc.) through various tests by looking at the impedance. These tests are: the driver radiating open; the driver radiating with the throat capped. In some drivers you can also remove the back cover, but not all. You can measure the Re quite accurately with an ohm meter. Then you know the diaphragm area - or can measure it. The front gap can be found from the model and the known area and length of the throat (Helmholtz resonator). Iterating the other model parameters and you can nail down all the pertinent values.

One thing that is also possible that I have never tried is to put a mic right at the blocked throat and measure the transfer function from the voltage to the SPL. At low frequencies this is a simple function of the volume and the BL. From this you can back out the BL quite exactly.

None of these tests require anything destructive of the device.
 
Importing

Hi Nate. The link that you shared helped me alot in the drawing process. Unfortunately, the group members still have not got to importing models. I have figured out how to import the CAD drawings into ABEC. However, I still need to figure out how to assign the elements their properties.
 
Hi Nate. The link that you shared helped me alot in the drawing process. Unfortunately, the group members still have not got to importing models. I have figured out how to import the CAD drawings into ABEC. However, I still need to figure out how to assign the elements their properties.

Ahh yeah.....I hadn't read thru the whole thread until last night. Bummer!

Let us know if you figure it out
 
Importing

Draw the horn(or whatever you wish to model) using one of the drafting softwares suggested in the forum. Export it as a '.step' file. I used a meshing software (suggested by the ABEC help section- Data Files & Tools/Mesh Files(CAD)) to mesh the '.step' file. Meshing software is very easy to use unlike the drafting! The meshing software saves the geometry as a '.msh' file. Add this file under the 'mesh' section in ABEC. Give it a suitable name as Alias. To initially view the whole geometry, I suggest using the script provided in ABEC/Help/Data Files & Tools/Mesh Files(CAD). Edit the code so that you make only one 'elements' section and use 'include all'.Under the 'Drawing' tab you can view how the elements are distributed throughout the structure. Clicking the small elements will give you their tag/indice number which you can chose to either include or exclude in other 'Elements' section of the script. A diaphragm can be added by following the solving script for Horn Expo 800 example.
 
Hi. I managed to solve the 'out of memory' problem by using a larger edge length of 2mm at the meshing stage. Now I am able to run it at 10KHz. I suppose if I have more RAM, I can run it at 20KHz. I am aiming for good directivity upto 10KHz. I am investigating all the other parameters that I have to consider while simulating the waveguide. It is a simple spherical waveguide with a narrow 'arch' shaped reflector in the middle close to the throat. As for the driver, I have modeled a simple 'piston' right at the throat. I have managed to measure the basic driver parameters(Fs, Re, Le, Qms,Qes,Cms,Mms & BL).
 
Last edited:
German ABEC Tutotial - Translation

Hi,
As I started the 'ABEC Thread' in the German DIY-Hifi-Forum...
I've never used ABEC myself other than playing with the examples so I'm not of much use. The best tutorial I've found is this one: http://www.diy-hifi-forum.eu/forum/a...p/t-11174.html. It's in German I think so the translation is a bit iffy but it looks comprehensive.
...I'd be happy to help and translate specific posts, if of interest. Just indicate.

Beside that, I'm very interested in this thread, as I'm actually trying figure out, which free CAD software works best to set up a horn contour and then import to ABEC.

KR, Christoph
 
Hi,
As I started the 'ABEC Thread' in the German DIY-Hifi-Forum...

...I'd be happy to help and translate specific posts, if of interest. Just indicate.

Beside that, I'm very interested in this thread, as I'm actually trying figure out, which free CAD software works best to set up a horn contour and then import to ABEC.

KR, Christoph
Ahh cool, thanks for joining in here. Do you think the other participants in that thread will get to the CAD part? I use SolidWorks so Idk what works well for free sw.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.