You tried making your own gripper using Dave DeC's drawings, didn't you? Or am I confusing you with someone else.
Nope, wasn't me. I can make one easy enough, but then it needs to be hardened and there are not too many shops that will harden one item for anything approaching a reasonable cost . . and I can't blame them. I can get stuff nitrided by just 'riding the tailgate' when they are doing other stuff, but not precision hardening and tempering.
And a Belleville drawbar is a really tough thing to do on a converted knee mill....
I've watched a few of these PDB projects and they all seem to suffer from the same problem . . . I hasten to say 'in my opinion'. The bellevilles are not the problem. Early on I noticed that people had no idea how much force to use. Then if someone did discover a target force, they were unable to figure our how to generate enough force to release it.
A pneumatic PDB is very doable, even with R8, that's not why I switched over. A PDB is a stepping stone to an ATC, otherwise, not enough benefit for the effort, again in my opinion. I can find no way to make an unattended, reliable, ATC with R8, and I've looked at it a bunch of ways and studied a lot of different approaches currently out there . . including Tormac's stuff. Tormac's is workable, but has disadvantages, starting with the old R8 problem of needing gobs of force to retain and turn the toolholder. BT30 is short, only needs a few hundred pounds to retain the holder because the torque transfer is taken by dogs and not friction.