Somehow I never saw this thread promoted with a new message, although I usually check on my phone which makes it easy to miss. I still wouldn't have had an answer though
Anyway, hope someone gains some insight from all of this. Back to making sawdust...
I did. Interesting that the midrange seemed to be more affected by the facets than the tweeter. I wonder if the fact that the tweeter is in a LITTLE bit of a waveguide, maybe it didn't "see" the facets as much.
I was going to respond, but I couldn't understand the graph patterns. I was afraid I might stick my foot in my mouth, so I kept it shut. I kept looking at the first midrange polar set that you posted and was wondering why the curves were offset so much below 1kHz. Now, after looking more closely, I can see why. The scale is only 2dB per division (when I initially looked I thought it was 5dB/division). I agree with your assessment that the power response will probably be hot in the 2-3kHz region due to the combination of polar response bunching and the overall driver directivity mismatch. The question is, how much of a problem will this really be? And will you be able to correct it significantly using various crossover design tricks?
PS: I suppose you could look at it this way: The baffle shape introduces a certain amount of power response peaking & dipping based on it's overall shape and dimensions. The individual drivers start to beam at different frequencies, which, in turn, introduces another set of frequency dependent power response peaks and dips. The problem is that I am still trying to figure out how to properly select the right drivers and build the right cabinet shape to cause all these curves to self-correct each other.
@rjj45 said:
re: @4thtry - me too! Seems like we have to build boxes and measure, and only then do we find a problem.
That's pretty much what I've always done. I guess we could spend the time to learn all the fancy modeling programs and all end up with the same speaker? 😛
As for a crossover, looks like 2K is what the mid kinda wants (directivity wise), but I don't think the tweeter is gonna like that. I next need to run some distortion measurements.
Comments
Somehow I never saw this thread promoted with a new message, although I usually check on my phone which makes it easy to miss. I still wouldn't have had an answer though
I did. Interesting that the midrange seemed to be more affected by the facets than the tweeter. I wonder if the fact that the tweeter is in a LITTLE bit of a waveguide, maybe it didn't "see" the facets as much.
I was going to respond, but I couldn't understand the graph patterns. I was afraid I might stick my foot in my mouth, so I kept it shut. I kept looking at the first midrange polar set that you posted and was wondering why the curves were offset so much below 1kHz. Now, after looking more closely, I can see why. The scale is only 2dB per division (when I initially looked I thought it was 5dB/division). I agree with your assessment that the power response will probably be hot in the 2-3kHz region due to the combination of polar response bunching and the overall driver directivity mismatch. The question is, how much of a problem will this really be? And will you be able to correct it significantly using various crossover design tricks?
PS: I suppose you could look at it this way: The baffle shape introduces a certain amount of power response peaking & dipping based on it's overall shape and dimensions. The individual drivers start to beam at different frequencies, which, in turn, introduces another set of frequency dependent power response peaks and dips. The problem is that I am still trying to figure out how to properly select the right drivers and build the right cabinet shape to cause all these curves to self-correct each other.
re: @4thtry - me too! Seems like we have to build boxes and measure, and only then do we find a problem.
That's pretty much what I've always done. I guess we could spend the time to learn all the fancy modeling programs and all end up with the same speaker? 😛
As for a crossover, looks like 2K is what the mid kinda wants (directivity wise), but I don't think the tweeter is gonna like that. I next need to run some distortion measurements.