In January 2003 a pair of 12 driver line arrays was completed using 6.5" Japan-origin "bazooka" subwoofer drivers. Following is a look at construction photographs, some construction details, driver parameters, and good intentions for the future for further research on dipole line array operation. The construction photos are taken in my new "Ribbon Shack" covered-patio area which I had constructed during late 2001 and early 2002 (with the hard won gains from my Vietnam Era veteran discrimination settlement from Cal State Long Beach). It is nice to be modestly retired now.
Routing the front baffle.
Nine-ply baltic birch in 12mm thickness was epoxied together to create the
needed length for the front baffle. The plywood was only conveniently available
locally in 5' x 5' square sheets . Spacing between the drivers is 1/16", which was
rather difficult to achieve reliably. Four of the webs between the drivers on the
two baffle faces had to be repaired due to breakout.
Without the Jasper Circle hole-cutting jig, installed on the router in the photo, the the job wouldn't have been attempted.
Not seen is a "Pro 8 Footer" 99" capacity clamping straight-edge from Griset
Industries. Combined with an aftermarket accessory platform, made of Deldrin, which
rides in the slots of the straight edge, and to which a 7" power saw is bolted, it
allowed the precision cutting of the long baffle pieces. No table saw was used in
the creation of the baffles.
The triangular truss assembly.
The triangular truss assembly was constructed from 1" x 1" cabinet grade stock
and 1 5/8" diameter closet-rod spines, which had been on-hand. The closet-rod
spines were from the now-cursed 16' tall microphone stand kludged together for
the 1998 Auditorium measurements for the RD-75 Dipole Baffle Study.
A $60 8" (205mm) compound mitre saw from Micro-Mark (The Small Tool Specialists) was invaluable for all the angle cuts on the truss angle sections. The depth of the truss assembly was designed such that the rear magnet assembly of the drivers would be about 1/16" from the closet-rod spine.
The primary method of assembly was with the West Epoxy System, but extensive
use of square drive screws is seen at the baffle-face to truss-angles intersections
throughout.
Epoxied rear spine of the truss assembly.
I had two primary concerns in creating this design:
1. To create a compact and relatively portable driver assembly that might be used for testing of dipole baffle issues - and, of course, as the woofers to my 3-way line array/line source system. The RD-75s planar-magnetics become midranges, and my DIY 6' true ribbon drivers tweeters.
2. To achieve the greatest possible strength and stiffness in the vertical dimension with the minimum possible weight using common materials. In my evaluation, considered as a standalone driver, the line array was most vunerable to flexing in the vertical dimension. A triangular truss assembly was the strongest possible design I felt I was capable of creating with my available DIY resources.
The rear spine of the truss is held in place solely with the West epoxy since a circular band of epoxy was created at the intersection of the angled truss components and the round spine. The end-grain joints of the 1" x 1" sections at the spine intersection were extensively soaked with undiluted epoxy before a much more viscous epoxy mixture was used to create the band. For those bands of epoxy the West System "High Density" filler material was added to the epoxy to acheive a very thick mixture which would stay in place as it was applied. All possible joint intersections between the front baffle and all the truss pieces, sides and cross-pieces, were epoxied! I love the West epoxy.
With the ends of the truss supported on blocks and suspending my full weight
of about 175 pounds on the middle of the truss, difflection appeared to be
less than 1/8" in the center. The assembly is very stiff.
A New Line Array for 2003 - Page 2 - Next Page.
A New Line Array for 2003 - Page 3.
The New Line Array for 2003 - Page 4.
Acoustic Line Source Research - Table of Contents.