I got this to work out, its notable that my 1000ipm job speed pretty much flies over my limits. I think it will be very necesarry to utilize software limits whenever operating machine. The debounce interval fixed the homing problem but amplifieds the high speed jog problem as the machine is over the limit before the limit even comes on.

Soft limits will help. And the Slow Zone.
Back to reply #5, did you reduce the debounce to where it JUST corrects the prob. + a little ?
The 2000 is just to pin-point the problem.
May only need to be 50 or so.
If set too high, and the limit is a straight in plunger type, hit it fast and CRUNCH.
RC

Its set to 400 right now, anything less and i get the error on calibration. The limit switch is plunger but it rides along a flat, so you can go right over it without damaging it.

Like Hood said in #6, you should check/shield your wires and check that switch. A clean noisless system should run at "0" debounce.
RC

It does the same thing on both axis and i've already replaced the Y axis switch (i jogged passed it and crushed it with the ball nut).

As far as noise, what noise are you refering to? The limits are wired to the control through an opto-iso board so i can see the relevance? What exactly is mach doing when the axis homes?

If  you get electrical noise in the cable from the switch it can look to Mach as if there has been a signal, it doesnt matter whether it is opto isolated or not .
 5v is very prone to noise as the difference between High and Low is fairly close in voltage terms, that is the reason industrial controls use mostly 24V for their I/0, as the difference between high and low is much greater. I use 24v on my lathe and Beaver mill and then just use relays or opto board to convert to 5v prior to passing to the BOB, this makes it almost impossible for noise to be an issue.
 Having shielded cables and having the shielding grounded at one single point in the control cabinet can make 5v work well enough, my Bridgeport has that and debounce is zero, but if I was re-doing the Bridgeport I would have 24v I/O.

Hood

When Mach "Homes" to a snap action limit switch, the plunger compresses until the switch "snaps" the contacts OPEN. Then the axis reverses direction until the switch SNAPS again closing the contacts. This distance varies between different mfg. of switches but lets say about .030". Now if you go to the HOME pos., you would be .030" from tripping the switch.

Edit: Provided you do not have a HOME OFF set

Home offset is zero.

The cable im using for my limits is shielded until the bridge where it splits to the various limits. Im running all the limits off 24V. The BOB is rated for 10-30V. I can't see it being possible for the wires to pick up 10V of interference from the shielded servo cables. ?

When the axis homes I get a "limit switch" error immediately when it touches the limit switch, no delay.

Well if you are using 24V then noise should not be an issue unless it is being picked up in the BOB.
So you are  ok with the debounce set to 400?
Would be interesting to see where the problem is, if you set the debounce back to zero and do a home move and trigger the switch with your finger and keep it pressed to see if the axis reverses for a bit before you release that would let you know if its noise in the BOB or just very twitchy switches. If it doesnt reverse then I suppose  it could still be the switch but it would have to be fairly bad switches but to be sure if you could take the wires off the switch and touch them together, do the home and open them and see if it reverses, if it does then its the switch, if not its noise at the BOB.
 Do you have a scope? If so would be fairly easy to see if there is noise from the BOB.
Hood

I can get a scope.

One thing i noted from what you just said, would it matter if the switches were NO or NC? I have them wired NO.  The switches are OEM parts off our press brakes, good quality. I dont think the switch would be the issue.

I tried pushing the limit switch in while the machine was homing and i get the "limit trigger switch error"