How can I model diffraction + baffle step on a non-square baffle?

[fjhuerta]

Hi!

I was working on a new shape for a design. I found out that a pyramid with slightly offset tweeter worked great.

Unfortunately, I'd like to model the woofers' response using LSPCad's KOF tool (basically a diffraction + baffle step simulator), and I just realized KOF cannot do non-rectangular baffle shapes.

If I understand correctly, neither can SoundEasy.

I'm following these steps:

1) Far field woofer measurements.
2) Near field woofer measurements.
3) Diffraction + baffle step on near field measurements.
4) Merge Far field + Near field with diffraction and baffle step.

Unfortunately, I can't simulate step 3... I guess I could approximate things... but this doesn't seem right.

Should I just stick to rectangular enclosures...?

[augerpro]

Couple things you can try:
-just use the farfield measurement
-if you just need baffle step for the nearfield response, you should be able to use Edge and find a rectangular baffle approximation of your real baffle. Once you find that approximation plug the dimensions into the SE diffraction program and it should be close
-use Edge to do the real baffle sim and then somehow import that response into SE and add it to the nearfield response in the frequency combiner, kind of like you would adding the port response to the woofer response. I don't know if Edge can export the response to a text file, but SPLtrace for sure would work. Then just import it into SE

I haven't tried the last two yet but I have thought about how I'm going to deal with non rectangular baffles. The main thing to remember is that edge diffraction should be in the farfield response so all we really need is the baffle step characteristic, and that can be approximated by a rectangular baffle-you just have to find it.

[BobEllis]

Check out The Edge at http://www.tolvan.com models non rectangular baffles, dipole or monopole quite well.

[fjhuerta]

Hi Brandon, Bob,

I'm not sure Edge would work, since I see no way of importing an SPL file into it. Thus, SPLTrace could not be used (not that I have SoundEasy around, anyway - I just looked into whether SE could model non-rectangular baffles).

But I guess you are right, Brandon - I could certainly begin creating baffle step+diffraction models on Edge, and when I find something useful, I could use that on the KOF tool (or is it KHF? I forget) of LSPCad.