Closed loop control using encoders?

[X7JAY7X]

I have read quite a bit and from what I understand Mach 3 uses encoders to update its DRO's.  But these encoder DRO's are seperate from the internal DRO's mach3 calculates from sending motor step pulses.  If Mach3 sees there is an error between the encoder DRO and the internal step pulse DRO it will feedhold and allow you to use the encoder DRO to update the step pulse DRO.  Is this correct?

Is there a way for Mach3 to use the encoder DRO only for positioning so it would be true closed loop control?

[Hood]

No and No

There is a board you can buy which does the first thing, well sort of, it wont pause and update then carry on. What it will do is fault Mach and then update DROs to encoder position. But is this useful? probably not as you have lost position by the amount of following error you have set up, this has probably ruined your work anyway if you are using a metal cutting machine with fairly tight tolerances, on a router the tolerances in your work are normally less so it may help.
Having said that you need to think why you are losing position and it all comes down to one thing if it is a regular occurrence, that thing is the motors are not big enough for the task you are asking them to do.

Now your second part, Mach is a buffered system so software closed loop is not possible.

Hood

[Promech]

Question: If one uses servos (not steppers) then in theory no steps should be ever lost. Is this statement correct?

[Hood]

If you use Servos then the loop is closed between the drives and motors.
Hood

[stirling]

Question: If one uses servos (not steppers) then in theory no steps should be ever lost. Is this statement correct?

Well the crux is the "in theory" bit.
Put it this way. Are servos able to produce infinite power? No. Are steppers? No.
If the servo or stepper is correctly matched to the job required of it then NIETHER will ever "in theory" miss a step. But the contrary is also true - and yes - I mean for servos too.

Ian