Desperation for a really 'big' dipole line-source can lead one to strange places, to wit, the Mighty DynaPleats. The DynaPleat driver, which came in 6" square and 4" square models, was sourced from a factory in Japan as an autosound loudspeaker. The DynaPleat's construction is highly unusual in that it is an 'Area Driver'. That is, the driving force of the voice coil(s) is applied over a much larger area of the piston than in a cone driver.
The 6" driver has an approximately 4.5" square diaphragm which is molded from a hard, clear plastic with 11 raised half-cylinders on the surface. Each of the half-cylinders has it's own voice coil. There are 11 voice coils per driver, each having a separate flat and thin magnet mounted upon posts. The suspension is fully symmetrical and pleated from a parchment-like material.
The driver is only about 1" thick! At the time they were available here in Southern California each driver cost about $8.50. So I took a chance a few years back and bought 32 of them. Using the IMP to measure them I selected the best of the lot and built three loudspeakers, with two configured as in the photo. Twelve DynaPleats are contained within the central section outlined by 2" width lumber (on the right). The loudspeaker assembly is properly termed a 'Discrete Element Line-Array'.
The other 'black' rectangular section to the left of the DynaPleats contains an unused mounting assembly for a 72" true ribbon that I was contructing at the time they were built (the ribbons are still in construction as of 4/98). The Mighty DynaPleat is large, with the base measuring about 27" in width, with the height approximately 6'8". The DynaPleat drivers are covered on the rear of the dipole baffle with 2" thick acoustic foam for damping purposes. A size comparison photo can be seen in 'The Baffle & Ribbon Gallery' section.
For the March 2, 1998 measurement session at the Gym I decided to bring a DynaPleat in addition to a Hsu 10" subwoofer. How many times does one get the opportunity to compare the responses of two 72" tall dipole, line-source loudspeakers?
Below is the merged full-range response of the DynaPleat compared with three RD75 dipole baffles (all using the same RD75 driver). The plots are 1/12 octave smoothed. The same power input was used with the DynaPleat as was used with the RD75 baffles, hence the output amplitude differences are comparable in regards sensitivity.
1/12 Octave smoothed.
Ahem! The DynaPleats *were* rated as a full-range driver. It is interesting to note that the >15KHz response of the DynaPleats is better than the RD75. Clearly the DynaPleats have greater sensitivity than the RD75s throughout much of their range. They do have problems in the upper frequencies, yet, all-in-all pretty amazing loudspeaker for the cost.
What was of interest to me is a comparison of the low frequency output of the DynaPleats and the RD75. What might we derive from the 1.9KHz sampling rate high-resolution low frequency comparison of the DynaPleat and the RD75 driver in the Triangle baffle (which was choosen for convenience only)? No octave smoothing has been used in this high resolution comparison. Please note the compressed amplitude scale used (12dB/divison).
No smoothing.
At this time I don't have much more information about the DynaPleats, not even their 'free air' resonance. Yet a large difference between it's low frequency response and the RD75 is evident. The RD75 driver is high 'Q', and does not exhibit the expected -6dB/octave slope from the upper frequency where the dipole baffle support ceases to provide sufficent width for flat low frequency response. The DynaPleat, in this regard, seems to show a classic response, falling about -12dB over the two octave span from 30Hz to 120Hz.
As in all FFT analysis with loudspeakers, the extreme low frequencies must be taken with several grains of salt. It would be nice to clearly see the ultimate theoretical expected -18dB/octave fall rate in the DynaPleat at it's low frequency limit. Is this what we 'see' in the response at about 26-30Hz, the hump below, centered at about 19Hz, being an artifact? My interpretation is yes (given the similar hump shown in the response of the RD75).
One of the outstanding questions regarding the low frequency dipole behavior of the RD75 driver is the apparent difference between Rudi Blondia's RD75 driver(s) and the RD75 driver I use for the majority of my testing (BG S/N ...54). Using the same baffle (my 1S6T) each of us has independently measured our drivers and found different low frequency performance in that baffle. One of our goals is to clear up this issue at the next mutual testing session in the Gym.
Acoustic Line Source Research - Table of Contents
