The best sounding HF driver you have ever met

How loud and how low something can go is largely a function of how big of a volume of air you can displace and how much distortion (largely IMD) you think is tolerable.

A JBL 2412 used in the JRX series has a 1" diaphragm, they sound like shite at high volume.

Each doubling of area will allow 6dB more output (for the same diaphragm excursion).

A 1.5" diaphragm is the next step up (2.25x the area or 7dB), then a 1.75" (3x area, 9.5dB), then a 3" (9x area, 19dB), then a 4" (16x area, 24dB).

Odd sizes, 1.25" (EV type, 1.56x, 3.9dB), 2.84" (Altec type, 8x, 18dB).

Going down one octave requires 4x the excursion, or 4x the area (representing 12dB more output capability).

The double 15 design JRX 125 sounds like it needs another 8dB~10dB more HF output (in a live music situation) with its 2Khz crossover, that looks like a 1.75" driver is needed. It also sounds like shite to take the 15s up to 2Khz too, 1.2Khz would be a lot better, what's that going to take? That 2/3rds of an octave is going to need another 8dB or so, so 16dB~18dB above what the stock 1" can do is needed. Now we are up in the 3" diaphragm class.

this time, science wins 

From Acoustical Engineering - Harry F Olson

Would this be the graph you are referring to?

_djk_ wrote: Which simply means that it increases at a rate of 4x each time you go down and octave. Olson has graphs showing this too. I have two copies of this Olson book (and some others). I think I am going to donate one of them to be scanned for posting, the process requires the book to be cut apart to facilitate the scan. Harry went to school about 20 miles from where I live, although the bulk of literature is at the other state engineering school about 100 miles away (where my father taught). It should also be pointed out that IMD is largely driven by diaphragm motion, and the increasing amplitude at the lower frequencies for a constant spl. In general, the IMD will be proportional to the square of the bandwidth in octaves. "so what's a typical max excursion of a 1.75" diaphragm?" Most hit the phase plug with 0.050" motion, some with less.

that corresponds with the akabak model of a compression driver I have been using (apparently based on physical measurements) which has the gap between diaphragm and phase plug defined as 1mm. 70W input is giving 0.17mm displacement at 1.3kHz. so the volume and therefore the compliance of the front chamber must change significantly with displacement which will give some sort of distortion, won't it?

For a direct radiator in 2Pi the diaphragm (cone) motion increases at 4x per octave (as checked in BassBox v6.0).

It looks like the diaphragm in a horn may only increase at 2x per octave, I'm trying to find a way to double check this as Hornresp uses the data from Olson. I'm trying to remember where my Beranek is.

Assuming 6dB is correct for the horn (as it seems to be in Hornresp), then my figure of 8dB for 2/3rds of and octave lower (from 2K to 1.2K) would be only 4dB. This still means that the 1.75" is a little short on output and that a 3" diaphragm class is probably in order. This corresponds with my experiences with live music, and reducing a few drivers to confetti. You haven't lived until you have put pin-holes through the suspension diamonds of a JBL 2445 from sustained high volume.