 |
 |
 |
 |
 |
 |
|
| Tube & Speaker Tastings | 2007 | [PICTURE PAGE] | |||||
| 2006 | [PICTURE PAGE] | ||||||
| 2004 | [PICTURE PAGE] | ||||||
| 2003 | [PICTURE PAGE] | ||||||
 |
Jonathan's hemanrig |
The Fabulous 1428
Participants in the Oswald's Mill gatherings inevitably approach one of Jonathan's large loudspeaker systems to see what is
producing all of that effortless, lifelike sound. Lurking behind each of the giant, circa 1936 RCA 18 cell theatre horns are
two, count 'em, two menacing industrial devices, oriented vertically on a "Y" type divided horn throat. "Hey,
what the heck are these things?"
This is a fair and reasonable question, since these do not resemble other compression drivers. The answer? They are examples
of the rarely encountered and under-appreciated RCA model MI-1428B, the fully evolved version of the first compression driver
built by RCA. This driver, and a companion 15" paper cone woofer which used the same motor structure, were designed by
engineers Max Graham and John Volkmann in 1936 for RCA's first large scale two way motion picture theatre loudspeaker systems.
In the mid 1930s a team of engineers at MGM Studios had designed and built several prototypes of a new two way theatre horn speaker, based on earlier experimental work at Bell Telephone Laboratories. Their efforts became known as the Shearer Two Way Horn System. It set new standards in theatre sound reproduction, and soon almost every manufacturer would build a similar system. The first Shearer type systems built by RCA used drive units supplied by Lansing Manufacturing Company, supplier of the original Shearer drivers. RCA was understandably anxious to produce their own "Speaker mechanisms", as they called them, so Graham and Volkmann got busy. The first versions of their designs were in production by early 1937.
The 1428 compression driver was really a breed apart from other designs of the era. Whereas most other drivers used a dome shaped diaphragm pressed or spun from aluminum foil, the RCA used a cone shaped diaphragm. Early examples were formed from a fibrous material, really a stiff paper, but an improved version was soon developed which incorporated two layers of fine silk cloth and a phenolic resin, known in those days by the trade name Bakelite. As far as I have been able to determine, the 1428 was the first loudspeaker to make use of phenolic resin as a diaphragm material. Then as now, it offered a high strength to weigh ratio, resistance to heat, and admirable self-damping properties.
Most compression driver diaphragms had an integral dome and aluminum outer suspension, and a voice coil attached between them.
Graham and Volkmann chose a different path entirely, retaining some design features from earlier RCA paper cone drivers.
Their 4 3/4" cone diaphragm used a clamped felt outer suspension, which both damped the cone and gave it freedom of motion.
The voice coil was 2" in diameter, wound from two layers of round aluminum wire on a paper former. The cone formed a
flat disc in the center section, except for a hole for the center locating screw and three annular corrugations, which served
as the "spider", or inner suspension.
Most guests are amp jockeys, however, and they come to plug their kit into the Mill's big RCA horns. With just about everyone using the same speakers, the Tasting affords a fairly accurate comparison of gear. The day started off phono related- Larry Welsh was on hand to help tweak the tables, which included the Mill's Technics SP10 tt and EPA arm (in the original Obsidian Base) and Shure SE1 tube phonostage, and Randy Carter's Empire 208 with modified Rega arm and a Bottlehead Seduction Beta phonostage. The Empire both looked and sounded excellent and proved a popular source throughout the day. Jeremy Epstein's 6C45 phonostage was widely regarded as definitive, and I would have to agree. Visit Jeremy's WebPages to learn more. Jeremy also brought his dog and a 2A3 push pull with "iron concertina" phase splitter. I believe the dog was single ended, though.
Perhaps the crowning glory of the 1428 design was the unique phasing plug. In a compression driver, the phasing plug serves
to limit the cross-sectional area of the exit path to some small fraction of the diaphragm area. Another goal is to make paths
from different parts of the diaphragm into the throat as similar as possible, to avoid phase cancellation that results from
recombination of differing path lengths. The 1428 phasing plug was a magnificent structure, cast from aluminum, through which
a number of radially-oriented (think spokes on a wagon wheel) slits traveled from the phasing plug surface to combine at
the 1.5" diameter exit. The exit path was out the end of the driver, rather than the more common through-the-body path
of most other compression drivers.
The unique construction of the 1428 may have resulted in part from corporate legal maneuvering. Lansing had been building
his drivers with features very similar to the earlier Bell Labs experimental units, and he had been informed by Western Electric
that he was infringing on E.C. Wente's U.S. Patent # 2,037,187 which described the circumferential style of phasing plug.
Western Electric was beginning production of its own version of the Bell Labs driver, the 594A, at the time. Lansing's engineer
Dr. John Blackburn solved the sticky situation by designing the first radial phasing plug, described in his U.S. Patent #
2,183,528. Since Lansing enjoyed an amicable relationship with RCA, he may have freely shared the radial phasing plug technology
with them. Another possibility is that the two companies collaborated on the basic idea; Lansing had been supplying drivers
to RCA, and both companies' compression drivers conformed to a common standard of 1.5" driver exit diameter and 2.5" 16 TPI threaded attachment to the horn. In any event, RCA probably wanted to avoid any threat of patent infringement from
Western Electric, as the two companies had been in contentious negotiations over what RCA regarded as W.E.'s restraint of
trade in its domination of the theatre sound system business.
The early version of RCA's driver was the MI-1428A, which used an eight section phasing plug. Upon examination of the surviving
literature and examples of the drivers, it is evident that Graham and Volkmann continued to develop the driver in a quest
for lower distortion, greater power handling, and especially more extended high frequency response. Compared to the Western
Electric and Lansing drivers of the era, the early 1428 may have sounded dull on top. It was soon replaced by the MI-1428B,
with its 15 section phasing plug. Other changes followed: the Bakelite diaphragm, a second improved diaphragm with added compliance
ring in the center, and a felt ring attached to the diaphragm's center section intended to damp a 1 kHz. response peak. A
service bulletin was issued to modify existing drivers by drilling a series of 1/8" diameter holes in the phasing plug,
to vent the area near the voice coil into the exit path. This series of improvements was very effective, and RCA was proud
of the fact that the MI-1428B offered comparable bandwidth, lower distortion and greater power handling than the competition.
The 1428, in A and B versions, used a field coil powered by 13VDC. The MI-1443 driver was identical to the MI-1428B except
for a 115VDC field coil. Probably about 6,000 to 8,000 examples of these drivers were built from 1937 to 1942. The design
represents an uncompromising effort, made possible only by the engineering and manufacturing capabilities of an industrial
giant like RCA. Nevertheless, the 1428 must have been difficult and expensive to build, especially with that complex phasing
plug casting.
By 1940 RCA had introduced the MI-9443 (field coil) and MI-9448 (permanent magnet) drivers, which used a phenolic, dome
type diaphragm and exit-through-the-body type construction. At first they were used in smaller theatre systems, but after
World War II they and subsequent similar drivers were used in all of RCA's postwar systems. People in the know have lamented
that RCA never again built a compression driver as good as the 1428.
By 1940 RCA had introduced the MI-9443 (field coil) and MI-9448 (permanent magnet) drivers, which used a phenolic, dome type diaphragm and exit-through-the-body type construction. At first they were used in smaller theatre systems, but after World War II they and subsequent similar drivers were used in all of RCA's postwar systems. People in the know have lamented that RCA never again built a compression driver as good as the 1428.
)
So why are we concerned with the 1428 in the 21st century, and why does Jonathan use them in his system, with all the "botheration"
of finding these rarities, supplying their field coils with DC, etc? Because to many listeners, the 1428 provides the most
vivid, lifelike midrange reproduction they have ever heard! The sound produced by these drivers on proper large horns must
be experienced to be believed. They offer very wide bandwidth, superb clarity, and startlingly accurate preservation of dynamic
contrasts. Vocal reproduction can be downright scary. A pair of 1428s on large horns can form the basis of a truly "ultra
fi" horn speaker system.
Strangely, until fairly recently these drivers were not sought after by hobbyists and vintage audio collectors, sold for
little money, and had a reputation for poor sound quality. Perhaps they have been broadbrushed with the reputation (undeserved!)
that phenolic diaphragms do not offer good sound for hi fi use. As with most of the finest early industrial sound equipment,
the majority of these drivers eventually kept their appointment with the landfill. We are fortunate that at least a few examples
of these supreme industrial art forms have survived for our enjoyment today.
Postscript- modern developments
My friend Rich Drysdale and I have been working to develop a modern version of the 1428 driver. Our idea has been to adapt
the basic features of Graham and Volkmann's masterpiece into a driver ideally suited for modern hi fi use. We have been experimenting
with variations in the phasing plug and other parts of the driver to try and preserve the excellent performance of the 1428,
while extending the high frequency response even further if possible. We have been encouraged by the results so far. The high
frequency response of the MI-1428B nosedives above 8kHz, but we have achieved flat on axis response to 13kHz. in our prototypes.
For our 50+ year old ears at least, a separate tweeter is no longer required. Due in part to the incomparable phenolic diaphragm,
the highs are very clean, sweet sounding and integrated with the mids better than we have heard from any other setup.
 |
Frequency response of new 1428B |
Our plans for now are to continue to experiment and refine the driver's performance. Eventually we hope to produce and sell our version if we can tame the production costs a bit. For now, we are focused on the goal of building a system from scratch to demonstrate at the upcoming Rocky Mountain Audio Fest in October 2004. Come have a listen!
Steve