We have been cutting a series of angled 1" slots (18) in alum. , have run about 12-15 of these

Started the machine up , hit return to 0 , DRO went to 0's , put in new piece , hit run.

It started cutting the first slot where it should have ended , and cutting away from where it should , basically moved the slot 1".

Hit the panic STOP ! , return to 0 , it went back to original 0 , checked everything , all good , took the mill out, hit run , worked perfect.

I have cut about 20 since then NO problems  Any ideas ?

2nd ?
Calibrated machine , ran parts , all good


RECALIBRATED machine , ran same G code , same parts , cuts were .05" closer than the first run

Both times machine calibrated fine.

Not easy to say what could cause that especially as it only happened once.

Regarding the calibration, what are you actually meaning by that? If setting steps per unit then you should not have to always be changing them, if you do change them then you wil get diferent results each time you change them.
Hood

Calibrated machine to Mach3 , I guess that's steps per unit ?  In M3 you put in a unit of measurement , 0 the DRO , M3 moves the axis , then input the actual distance it moved.

Had to re-calibrate , because I was trying to setup M3 spindle speed control , and coolant on/off , but MFG'er gave me the wrong info to change and I had to reload M3 and lost original Cal. numbers.

But what stumps me is , I had to rewrite the G code to get it to work properly again , slot size stayed the same , the spacing changed?

ETA; I've been really watching the return to 0 numbers since , it's REALLY accurate (to me at least ) always goes back within .0002 of 0 ,many times back to 0 , I assume that's good ?

Calibrated machine to Mach3 , I guess that's steps per unit ?  In M3 you put in a unit of measurement , 0 the DRO , M3 moves the axis , then input the actual distance it moved.

Had to re-calibrate , because I was trying to setup M3 spindle speed control , and coolant on/off , but MFG'er gave me the wrong info to change and I had to reload M3 and lost original Cal. numbers.

But what stumps me is , I had to rewrite the G code to get it to work properly again , slot size stayed the same , the spacing changed?

ETA; I've been really watching the return to 0 numbers since , it's REALLY accurate (to me at least ) always goes back within .0002 of 0 ,many times back to 0 , I assume that's good ?

Ther is only one number for steps per unit that will be correct for an axis. I prefer to calculate it as that will make it as accurate as your hardware but if doing the set steps utility then as long as your measuring is accurate the steps per will be. Measuring however is unlikely to be as accurate as calculating as most people do not have the means to measure that accurately.

Slot size and space between should never change with respect to one another no matter if you change the steps per unit unless of course they are only single axis moves and slots run one way and spacing another.

Regarding the return to zero, that all depends on what you are talking about, if just referring to what Mach is showing in the DROs then that is no way to check. That is simply where Mach has told the axis to go to but as there is no feedback Mach can only assume the axis has actually got there,

Hood

Ther is only one number for steps per unit that will be correct for an axis. I prefer to calculate it as that will make it as accurate as your hardware but if doing the set steps utility then as long as your measuring is accurate the steps per will be. Measuring however is unlikely to be as accurate as calculating as most people do not have the means to measure that accurately.
Hood

You may find this calculator useful in determining the exact steps per unit.
http://www.machsupport.com/forum/index.php/topic,16315.0.html
Al

Thanks Al! 

Brett