The previous trapezoidal baffle tests performed on my dipole Carver did not take into account the "average" width of the baffle. The following measurements try to keep this more or less constant. "Average" baffle width is [baffle surface/baffle length].
The plot below is a little "busy" but shows a few interesting things. As a reminder, vertical scale in these plots is 3dB/division or 1.5 dB between horizontal lines.
The white curve was taken on a 15" wide rectangular baffle, fully symmetric. The yellow curve was taken on a baffle that had one rectangular half replaced by a triangle, 10" wide at the base and 60" tall. The baffle is 20" at the base and 10" at the top. Since 3 meter is still pretty close to the ribbon there seems to be some decrease in dipole baffle low cancellation but more importantly the dips at 400 and 800 Hz are less pronounced. In the 1.2 kHz to 2 kHz things are "reversed". Above 8 kHz a fraction of a dB more high.
The brown curve was taken using a fully symmetrical trapezoidal baffle, 12" wide at the top and 18" wide at the bottom. "Flattens" the low from 200 to 800 Hz but with a higher peak at 1kHz. The area from 1.3 kHz to 2.5 kHz is less bumpy and a fractional dB more high between 8 and 11 kHz. The green curve was taken with the trapezoidal section at one side made one inch wider. As expected some less dipole baffle low cancellation, the peak at 1kHz is gone and there is an increased level between 1.3 kHz and 2.5 kHz. As stated before, subtle differences, making choices difficult.
The next plot was taken at 5 meter distance from the driver, close to the limits of my current living and listening room arrangement. This position show a "room" problem, 6dB down at 200 Hz and peaking after that. (same problem as seen on the previous RD75 driver plots at this distance.) The difference between the green and brown curve is the geometry of the rectangular baffle. Green is asymmetric, brown is symmetric The white curve is made using a "rectangular" baffle with a triangle cut out at one side, starting halfway the baffle and tapering 3 inches inwards to the top. Makes the average width higher then the two previous curves but this is not a subtle overall difference anymore. Again, remember that the distance between two horizontal lines is 1.5 dB.
The plot below was taken on a B&G RD75 installed in a quarter round baffle. Warning, the dip at 200 Hz is a room effect that I've been experiencing at 5 meter distance from the driver. See also the plot above. (Different drivers, different baffle widths, similar location of the peaks and dips between 200 and 600 Hz.) The white plot is from a 17" wide rectangular quarter round baffle. For the brown plot I cut a 3" wedge out of the baffle at one side and attached it at the other side, creating a symmetrical trapezoidal shape on a curved surface. The differences again are not outspoken, but they are there ;-) No EQ on the curves.
