GRS PT6816 Improvements

I recently tested a GRS PT-6816 and am very impressed with the performance. From the get go I was reintroduced to what I consider to be the superior sonic presentation of a true dipole compared with conventional drivers. This is further benefited by the minimal and non resonant enclosure of the driver. The GRS seems neutral and fairly well behaved. It exhibits minimal audible distortion playing loud crossed over at 400hz-20khz. I consider it equal in performance to the best 3" full range and mid woofers that I've tested. An advantage I observe in addition to its dipole radiation pattern, and advanced elegant design and construction, is that it is extended like true tweeter, more so than any of the 2"-3" conventional cone drivers that I've tested. Though conversely, unlike said conventional drivers, 400hz is certainly the limit of its lower frequency extension. In addition a point source is a preferable design to a line source in my opinion and thus I wouldn't endeavor to use more than one or two of said drivers per channel in my explorations.

I observe a harshness in the upper mids and highs at louder volumes when compared with the highest performing small cone drivers and or a dedicated tweeter. While the clarity and istrumentation is equal only to electrostatic transducers, there is also a seeming albeit slight lack of dynamism and three dimensionality to the sound that the GRS produces.

I wonder if both aforementioned observations are reduced or resolved through the incorporation of a more elastic membrane and conductor as I intuit that "Polyethylene Naphthalate" is quite rigid, seemingly similar to mylar or kapton tape. In the interest of reducing high frequency resonances of the membrane, increasing low end extension while maintaining treble extension, and increasing effiency, I suspect that the incorporation of a slightly elastic membrane and conductor would set a driver of this design in a class above the rest. What would be such a material for both the membrane and conductor? Even at present, the GRS PT-6816 is my driver of choice for a near midfield two way or three way design.
 
Last edited:
As with all drivers to cover one bandwidth well or lower frequency.
Even on paper if it measures well at higher frequency.
More suited for lower frequency.

Real world a smaller driver is used to cover higher frequency.
Rather simplified but physics and driver size eventually come to play.
These drivers types are rather light mass diaphragm and with very powerful magnets
the distortion is very low and detail is rather clear.

Agree horizontal of course non issue for multiple units.
But vertical would approach 16 inches instead of 8 inches.
Bringing the crossover way down to 800 Hz or lower.
Rather silly since this is part of the ideal bandwidth we wanna hear.

Cross over to a smaller planar for upper bandwidth and the benefits will
shine just as much as the larger counterparts.

Even if the materials were improved more for a larger surface to produce
high frequencies. The physical size still has tradeoffs off axis. Much like any driver type.
No need to ponder how to improve or why it does so. The physical size is a tradeoff regardless.
Easily solved by smaller unit. The detail is well worth it.