Mike,
When I built my router I used servo's but I am thinking back to some of the issues I had to figure out when I considered steppers. If my math is correct (please check me on this). You are going (200 ipm x 5 TPI x 200 steps/rev)/16 micro steps = 12,500 pps. This is good and within the limits of Mach 3. What is concerning is you are running your Stepper at 1000 RPM which sounds high since they are best run at lower speed for peak torque. I believe steppers are best run less than 100 rpm range for peak torque (depending on the torque curve of the motor). Do you know if your motors have enough torque to move your machine at the higher rapids given the torque drop at the higher RPM's?

First I rechecked my setting and I forgot I had reset my resolution to 1/8 microstep not 1/16th. This puts me in the 8000pps range where the torque drops from 3nm at 200rpm 500PPS, down to about 1.2nm at 1000rpm at around 8000pps. I should be able to get 350 to 400 IPM with the 40vdc supply and my torque would still be exceptable for the mass I have to move. I could change from unipolar to bipolar parallel wiring and gain some torque but I don't see that advantage at this point. I'm more concerned about the Resonance issue. I just hate to waste the money when I'm thinking that the resonance issues won't change. I can get 300IPM now but again I get resonance issues at anything over 200IPM usually from direction changes only. These most likely are caused by mass/acceleration at that High of feedrate/rapid. Nema 23s are notorious for these problems. I can dial out some of it by lowering the acceleration but that defeats the purpose. I need the small rapid moves to be at max in order to see any gain. I really don't want to redo the entire machine to go to nema 35s. I would be better off selling this one and building a new one from scratch using servos.

Don't get me wrong 200 IPM rapids are fine but faster rapids would speed things up for me. My milliing feedrates would most likely not change anyway as I have found the optimum Feeds and speeds for most that I do. But I do cut a lot of small parts and there are many rapid moves made when milling multiple hundreds of the same part out of a single sheet. By speeding that up to say 300 IPM, would increase my productivity by 15%, in an 8 hour day that's a 72 minutes saved in that 8 hour time frame. 72 minutes is approx 80 more parts a day. At $0.75 each that adds up to dollars.

Brad, Thanks for the food for thought on this and you may have solved my issue as it does look to be more of a mass/torque problem. I just needed someone on the outside to look in from the other window to see it.

any suggestions or pitfalls are welcome for discussion please.

MK