A study on directivity... cones, domes, large and small, small and large baffles

[tktran]

I decided to fork out the study on directivity. And move midranges/fullranges/midwoofers etc. Basically anything that's 2" or larger.

Let's start with the B&W 5" Continuum FST from the 803 D4, released in 2021:



Data courtesy of @Dcibel
Here it's mounted in his test cabinet of 222mm x 414mm with a round-over radius of 10mm.

Here are the original frequency response and distortion measurements here:

https://www.htguide.com/forum/forum/primetime-a-v/mission-possible-diy/driver-testing-discussion/950349-some-random-driver-tests?p=950887#post950887

Now I imported the data into VituixCAD2 to look at the directivity:

Recall, here's the on-axis response:



Off axis:



Nothing can be done about that dip around 13KHz, but would the resonance at 16KHz benefit from a notch filter?



Let's straighten out the on-axis response and insert high and low pass filters for an acoustic LR4 @ 400 Hz and 4000Hz

Before notch filter:


After notch filter:




Suppose we didn't try to flatten out all those lumps and bumps, but still aimed for acoustic LR4 @ 4HKhz:




POLAR MAP for a straightened on axis, LP and HP filters for acoustic LR4 HP@ 400Hz and LP @4Khz:


POLAR MAP where on axis not straightened LP and HP filters for acoustic LR4 @ 400Hz and LP @4Khz:

[tktran]

ONWARDS and OUTWARDS!

Here's another B&W midrange. This time, an older model with the classic yellow woven Kevlar cone.

This one is in a cabinet 24cm W x 30cm W x 20cm D (9.5 x 12 x 8"). Driver centered on the baffle, near bottom edge of cabinet.



ON and OFF axis, Line Chart:


Straightened on axis, with crossovers for acoustic LR4 high pass of 400Hz and low pass of 4000Hz:




POLAR MAP:

[tktran]

The same 6" Kevlar midrange in a spheroidal enclosure with diameter 21.5cm (8.5")






POLAR MAP:



Notice the shape of the radiation it quite different - smoother contours lines, and although the both have -6dB beamwidth of 30 degrees at 3KHz, the region between 1Khz to 3KHz is not as wide...

[tktran]

The next FST is one from the 801 D4 and 801 D4 Signature-
B&W mount this midrange in an spheroidal enclosure with a diameter of about 11", with large 2+ inch round-overs, then gently tapered all the way back 24" deep. The midrange enclosure alone is some 40lbs,



The whole system is some 100kg (220+ lbs). Who could blame B&W engineers for not taking polar measurements?

But what does using a spheroidal enclosure do?

Here we have taken the midrange and placed into a spheroidal enclosure.
This one is made of fiberglass and is 8.5" in diameter, and effective round-over is only ~1".
The upside is that it's much lighter, so that can it can be placed onto a turntable for polar measurements:



On and off-axis measurements, Line Chart:



Straightened and corrected for acoustic LR4 @ 400Hz and 4KHz:



Resonance is 40dB down. What about a notch filter?
Perhaps unnecessary, but this is what it would look like with the notch applied:



POLAR MAP on a 30dB Y axis






Edit: corrected graphs, dimensions. 360 degree measurements to come…

[tktran]

NEXT UP:

A 3" dome that weight 20lbs. It's the the Volt VM752. This one definitely deserves a photo:




It's designed with a shallow waveguide, to be mounted from the rear onto a 3/4" baffle, sitting flush. So we have a situation that is somewhat the the opposite of the B&W driver above- ie. instead of having a round-over starting immediately at the frame and swooping backwards away from the cone, we have a waveguide swooping forwards from the dome and then meeting a flat baffle.

So how does this shallow waveguide + baffle affect the directivity/dispersion?

Here's a speaker designed by Matthias Brennwald, the Open Source Monkey Coffin project: GitHub - mbrennwa/osmcdoc: Documentation for the Open Source Monkey Coffin

The center of this 3" dome is mounted 14.5" wide (368mm W x 682mm H) cabinet, offset 1 11/16" from center.
Here's the on and off axis responses taken from the .FRD files shared by Matthias Brennwald:



What is that dip around 2.4KHz?

The project documentation said "The on-axis response of the Volt VM752 midrange shows a dip at 2.3 kHz. This dip disappears in the off-axis curves, which indicates that this is an uncritical diffraction artifact related to the driver/waveguide geometry rather than a problematic resonant effect."

I've applied a filter to result in an acoustic LR4 high pass 300Hz and 3KHz low pass to the on-axis. The pink is what it will look like:








So what does the POLAR MAP look like on a 30dB scale?

[tktran]

NEXT UP:

4" cone midrange- Faital 4FE42 in 6" spheroidal enclosure:

On and off axis measurements:

LINE CHART:


POWER and DI chart:



Frequency response now baffle step loss compensated, neutralized and with crossovers for acoustic LR4 high pass 300Hz and low pass 300Hz: In a passive XO, this gives a midrange sensitivity of a healthy 90dB/2.83V (8 ohm driver)