I got borred and needed to take my mind off some life issues, so I decided to redo my crossovers in the Focal Towers I've been working on. These were designed around components I had laying around, rather than the best components for the job, never ideal, but surprising how well they worked out. There are numberous issues with these speakers which did not cause the expected problems. The first problem is that I was careless with a router and routed the recess for the midbass drivers too wide by 1/4" all around. Second issues was that I recessed all the drivers too deep, and fully expected bad diffraction from the lip it created. I also can't find my roundover bit, so all the edges are currently not rounded over at all. Again, I fully expected problems, but I didn't see as ragged a response as I expected. The tweeter is too high, I didn't have the 8 ohm resistor I needed for the L-pad and used a 6 ohm instead, as it was all I had. One problem I noted was that I tried to take into account the baffle step I was measuring before, and apparently over did it, because now there is about a 1-2 decible reduction in energy in the midrange. I actually like the sound like this, but will be ordering a new resistor to see if that balances things better.
Most impressive to me was that by crossing these over this low (~1.6khz) the response is very good off axis. In fact, the response I will show is the spatially averages response +/- 15 degrees. There was no change greater than .5 decibles from around 200hz to 15khz or so. I also noticed that my speakers previously had quite a reduction in the treble above 10khz, which I had assumed was caused by the absorption of my acoustic panels, carpet, drapes, etc. I thought, I wonder if I leave the top end lifted, will I see a virtually flat response or the expected lift in the model. What I found was considerably less reduction in the highs above 10khz than I expected, making me think something else was causing it.
Some things to note are that A) my speakers do not have bass down to almost 20hz, that is because I accidentally left the subwoofer on when I took my measurements. It also showed me that I maybe have some work to do in the bass response of my room. The sub level is too high I think, and the response is pretty lumpy. B) The spike at around 2.5khz or so is inherent in the tweeter, as you can see it in the factory response graphs too. I don't know what causes it, but intend to play around with felt, foam, that phase plug, and whatever else I can to get rid of it. Anyway, its a pretty impressive response I think, I'm happy to know I can design workable crossovers. The lowest point in the response is 86db's at 440hz and 86hz, and the highest was at 89dbs at 15khz. That makes it +/- 1.5 db's from around 20hz to 20khz, and more like +/- .5db's over most of its response (Ignorring the few glitches.
If anyone has comments on what might cause some of the glitches, or if things do actually look like glitches gaused by diffraction, etc. let me know. As I said, I plan to use some 1/2" acoustic Foam I have to see what effect it has, but I'm also wondering what else I might do to further improve things.
6th July 2008, 09:54 PM
Did some more crossover tweaking, here are my Focal Towers response
Similar Threads
