I've found JBL tech sheet info is usually good to two decimal places.
800 Hz and 3 kHz looks like what it is.
Nearfield measurements of the three drivers below (all playing).
Port's next to 2118 and 2404. I stuffed it, but there's still leakage.
Don't know why this 2118's dying at 2 kHz. Kinda looks like a phase hole.
Indeed, the mid and tweeter are wired out-of-phase, but that shouldn't matter nearfield.
Second unit, same deal, so nothing seems FUBAR here.
http://www.jblpro.com/pub/obsolete/2118.pdf
Anybody have the original that can tell us what the legend in the FR curve says?
45° off-axis response shown falling off above 1+ kHz there?
In any case, 13.5 uf into 8 Ohms translates to something just under 1 kHz, and 2.5 uF with 0.4 mH is up around 4.5 kHz, both depending upon the actual driver impedance, but it's the acoustic response what counts, shown in the measurements, i.e. how the filter performs with the particular drivers in the total design package.
It's a little tough reading the values in the schematic, tho:
http://www.jblproservice.com/pdf/Cab...ries/4628B.pdf
Looks like 3.0 uF changed to 2.5 uF in the "B" version, raising the highpass frequency to the tweeter a bit.
I wouldn't pay actual money for those crossovers; they barely qualify. Woofer plays full range, substantially overlapping the mid. ACK!
[This has been a Zilchster stream-of-consciousness presentation....
]