Yep, I set the Z value appropriately after switching the bit. It ran through and cut almost the entire toolpath which has just maybe 1/2" Z variance, then drops. I actually just tested it again without a bit in it and without the dust collector running. It drops in Z way below where it should around the 92,000'ish line number. I think I'll have to send it in to Vectric as a bug report, because the software shows the Z staying where it should, the simulation looks good, but the program that is generated tore down into my table at this point. I was just curious if anyone else had seen this. It's a fairly complex cut (i.e. that's a lot of lines) with a single bit and almost makes it to the end when it's like the bottom just drops out.
In the image, you can see it looks like it's making a profile cut around the outside of the sword on the right side here, but I do not have a profile cut in this toolpath or even on this side of the piece, it's just supposed to continue cutting at pretty much the same depth level to give some additional offset (boundary offset set to 0.2 inches) , then I flip the piece over and do the profile cut from the other side.
Also, I did get some kind of "Inertia Position Error" (or something like that) on the A11 controller the first time it failed. That could have been because it tore the piece out though I guess.