--- [EMAIL PROTECTED] wrote:
> Having soft and hard limits seems redundant. Redundant is good as long as the overlapping safety systems are properly aligned. The biggest advantage that I know of with soft limits is that you can not crash into them in a well configured EMC2. The commands to move an axis -- after it has been homed -- honor the soft limit by setting it as the endpoint for the move. That's why you will see decel to the soft limit. Set a real slow accel and test it for yourself. Nice 'eh. This is not true of a real limit switch. You can easily power full speed right into one. That is why there is some space between a real limit switch and the end of travel for the axis it protects. It's a space to bleed off some of the inertia of a quickly moving machine. Now you can NOT safely use that space as a part of the normal machining travel. Now manual jogs before homing do not quite honor soft limits. The fact is that all jog motions before homing are constrained by the total length between negitive and positive soft limits. So for example let's say that min limit is -1 and max is +1. The total length of a jog before home is 2 units. It would still be 2 units if the limits were 0 and 2. It will do that 2 unit move either direction regardless of how the real numbers are set. When you reach a jog before home limit the axis stops but you can make another move in the same direction the same length. So the worst case of killing off an EMC while the machine is at the way wrong end is that you may have to make several moves to get it to the correct zero position. This should not be much of a problem since the prudent integrator will maximize axis travel between soft limits so a properly limited control should be able to get any axis to nearly the other end in a single move. That it for manual jogs through the normal NML channels. MDI commands on the other hand look at the soft limits as set, whether homed or not and use that value as the extent of commanded machine travel. You can't MDI past the axis position that matches the soft limit value. I believe this is true regardless of the coordinate systems or g92 offsets in force. All of this soft limit stuff assumes a repeatable home position. I don't really care how you define that or move to it or .... If machine home wanders at the whim of the operator, and many small stepper driven machines do just that, I'd recommend putting the full length of travel as both negitive and positive soft limits. That way you never hit them. For one of my Sherline mills, I've got -9 and +9 for the soft limits on the X axis. Now limit override is NOT soft limit override. There is no such animal. Limit override means let me move the offending axis off from the real limit switch it has hit. This can happen any time you have real limit switches hooked up properly and you are jogging the machine around before you home it. HTH Rayh ------------------------------------------------------------------------- This SF.net email is sponsored by the 2008 JavaOne(SM) Conference Don't miss this year's exciting event. There's still time to save $100. Use priority code J8TL2D2. http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
