Thanks much Bill. I will say I think the above 10k dropoff maybe looks a little worse on the first graph since it goes to 40k and the second graph only to 20k.
Good info on the off axis measurements. I haven’t assembled the test crossover yet, hope to this weekend. I will definitely measure off axis and wouldn’t have if you didn’t say anything.
based on off axis sims the outside vs inside off axis will be different, I assume due to the offset mid. It looks better panning right than left, which I guess makes sense as the coils are better aligned that way.
I’m going to keep tweaking on the computer tonight yet.
If I do measure and get bunching, what usually is the fix for that?
On the speakers that I have designed thus far, the bunching tends to be “baked in,” so to speak, for the most part. I say “baked-in” because the problem seems to be mainly determined by individual driver directivity characteristics. The best fix is to change to different drivers that have a better directivity match throughout the crossover regions. The shape of the baffle and roundovers, and individual driver placement decisions can also affect the bunching, but not as much. You can also modify the bunching somewhat by moving the crossovers higher or lower or by changing the crossover slopes, or by using asymmetric slopes. But, based on my limited experience, it is mainly determined by driver selection.
The bunching area is a concentration of acoustical energy that produces a bulge in the power response over a narrow band of frequencies. So one workaround would be to create a slight depression in the on-axis response over that same band of frequencies.
The net result should be a more even tonal balance at the listening position.