Thank you, Steve! Another GEM for the LHS Archives!
Thanks also to LHS member 1audiohack. You have resolved most all the pertinent questions regarding ring radiators IMHO. Grumpy has also been a help too, along with many others including Widget, 4313B (Giskard), Subwoof, Zilch, and many others... a lot of input from many members. that's what makes this site so great!
The only thing that I want to add is that there is a distinct difference between the AlNiCo motors and "Tops" i.e. horn parts, and the Ferrite versions of the same:
1: The AlNiCo magnets have the guide pins, and corresponding holes in the "tops" that bolt on to the magnet assemblies.
2: The AlNiCo magnets are smooth across the top, where the Ferrite magnets have a machined recess of approximately .015"-.017" across the face of the magnet to accept and align the diaphragm.
3: The pole piece is aligned with the outer portion of the top plate in both the AlNiCo AND the ferrite magnets; meaning the diaphragm sits "flat" in either magnet- the diaphragms' inner clamping ring and outer clamping ring sit in the same plane on either magnet: the top of the pole piece is aligned with the diaphragm seating surface on both magnet versions.
4: the "Tops" are machined accordingly to mate with their intended magnets: the AlNiCo magnets mate up with "tops" that have a recess for the diaphragm that is approximately .035" deep, and the ferrite magnets mate up with "tops" that have a diaphragm recess that is approximately .015"-.017" deep.
IIRC, diaphragm clamping rings measure approximately .040" in all cases, I did these measurements last summer and was in regular communication with 1audiohack during the course of our measurements and other sleuthing regarding the distinctions among the various incarnations of JBL Ring Radiators.
I measured this on my AlNiCo 075 and 2402, and compared it to a pair of 2405H as well as the magnet for a 2404H-1 that I have. I used a machinists' straightedge and feeler gauges to measure the diffeneces, and I feel confident in my findings. I believe that 1audiohack would support me on this as a viably accurate method.
That being said, I feel that anyone concerned about "bastardized" combinations of parts could use this method to validate a proper combination...