Hi Joris,
If a sound wave hits an object that is less than ¼ of a wavelength in size it goes around the object like it wasn’t there. Similarly if you place 2 sound sources within ¼ of wave length they will combine like they are one source. At higher frequencies they won’t combine and you will begin to get phase cancellation producing lobes in the coverage pattern.
The trick I’m trying use with the double 10 and the PM90/60 is to have a crossover point so that the 2 / 3 sources combine without any lobbing.
The problem at low frequency end of the horn’s operation is its directivity will not match the nominal coverage angle; it will be much wider and more or less controlled by its width and height, not the angle of its walls. If the low frequency driver is a similar size to the horns mouth its directivity will be similar to the horn.
When the horn has control of directivity at higher frequencies its mouth will be large with respect to the wavelength it’s producing and you will not be able to get the two sources to behave as one as mentioned above.
Because the HF horn and the 10” driver have some directivity at the crossover point it does help to reduced lobbing at the crossover when you model it compared to an omni source. What you find in practice with typical horns and drivers we use is you can get away with almost ½ wave length spacing and still get very acceptable results.
The obvious question is why not use a tighter pattern horn … well the horn has to be bigger to get the directivity control and then the driver spacing is increased and you have achieved nothing.
