Difficulties in designing a waveguide

[willi_vintage]

Hi folks,
A few questions arose when attempting to design a waveguide for a tweeter. The primary goal is to adjust the delay between the woofer and tweeter. This results in a depth of approximately 15 mm. This should be used in such a way that the directivity is adjusted to the woofer and more strongly restricted vertically (e.g.: 120° horizontal, 90° vertical). The tweeter is a classic 25 mm dome and coupling should take place at 1500 Hz. A waveguide was designed using this data. What doesn't fit: the depth should be at least 30 mm. My primary design is limited to 15 mm, which is actually too little. The whole thing was simulated:

cross-section in the simulation (vert). For simplicity, the chassis is open at the rear:

and SPL @ 1600 Hz:

2860 Hz:

13550 Hz:

so far...but I don't see any differences in directivity (left guided, right unguided). The same applies to horizontal radiation behaviour. Wrong design, input or simulation? idk

 

[Heinrich]

Hi folks,
A few questions arose when attempting to design a waveguide for a tweeter. The primary goal is to adjust the delay between the woofer and tweeter. This results in a depth of approximately 15 mm. This should be used in such a way that the directivity is adjusted to the woofer and more strongly restricted vertically (e.g.: 120° horizontal, 90° vertical). The tweeter is a classic 25 mm dome and coupling should take place at 1500 Hz. A waveguide was designed using this data. What doesn't fit: the depth should be at least 30 mm. My primary design is limited to 15 mm, which is actually too little. The whole thing was simulated:

View attachment 475640

cross-section in the simulation (vert). For simplicity, the chassis is open at the rear:

View attachment 475641

and SPL @ 1600 Hz:

View attachment 475644

2860 Hz:

View attachment 475643

13550 Hz:

View attachment 475645

so far...but I don't see any differences in directivity (left guided, right unguided). The same applies to horizontal radiation behaviour. Wrong design, input or simulation? idk

The dome's shape, convex versus concave, depends from where you look? The FOCAL's concave dome is famous for its unconventional dispersion pattern.

 

[AnalogSteph]

To my knowledge, making the throat shorter at the same angles means that the waveguide starts becoming ineffective higher in frequency on the bottom end (and thus efficiency does not go as high either) but retains its directivity higher up. Presumably it's not doing much of anything at 1.6 kHz yet. There does seem to be an effect at 2860 and 13550. Maybe try looking at 4ish kHz or so?

 

[willi_vintage]

please excuse my late reply...still have to do too many other things
yes, it seems the throat is too short (as expected). Effect with depth of 14 mm is visible @ 4-5 kHz with increased efficiency. Seems lowest frequency to have a directivity is x= c/2*f. Did a sim again with 30 mm throat and elliptical opening (110 and 80 degree). With that parameter, it should be round about 5,6 kHz. Sim results:

@1,6 kHz, 110°:

@3 kHz, 110°

@5,6 kHz, 110°

@10,5 kHz, 110°

@15,5 kHz, 110°

same with 80°....