Accuracy of an inch machine

[sinha_nsit]

What I gather from Fanuc manuals, the minimum slide movement in increment system B is 0.0001 inch (which is called least command increment) on an inch machine. This is equal to 0.0025 mm. Now, when we work in G21 mode, how can an accuracy of 0.001 mm be maintained? In other words, G21 G91 G00 X0.001 will not cause any actual movement.

Either accept this statement or please contradict me. Either way, I would learn something.

[Al_The_Man]

As you say, your least commanded increment is .0025mm, so you can not position to .001mm with this system, what is the controller you are working with?
Al.

[sinha_nsit]

Finally, you decided to respond. Thank you.
I am referring to Fanuc 0i.
Now, I sort of understand the difference between least input increment and least command increment (as defined by Fanuc). The first is what we program, i.e., least input value. The second is the least slide movement, i.e., the least output value.
The actual movement is in multiples of least command increment.
Therefore, when working in mm mode on an inch machine (a machine with inch lead screw) in increment system B, I conclude the following
G90 G00 X10.000 would place the tool at X10.000
G91 X0.001 would not cause any slide movement
X0.001 would position the tool at X0.0025 mm
In other words, the tool would move only in steps of 0.0025 mm.
I definitely do not mind being contradicted. This is the way I learnt many things.

[CNC Data]

Hi Sinha.

Some Fanuc controls can go lower and this is set by param.

But in reality, forget anything under 5 microns on a used machine.

[sinha_nsit]

On 0i also, increment system C can be selected which changes least input increment to 0.0001 mm and 0.00001 inch in G21 and G20 modes, respectively.
But, in general, the mechanical structure of the machine is not rigid enough to ensure repeatability to this extent. What is shown on display would not be the actual position. I have not seen anybody using IS-C.

[sinha_nsit]

And, if nobody contradicts me, I will assume that the slide always moves in multiples of least command increment, whereas we specify the distances in multiples of least input increment. As a result, there may be a slight difference between the displayed position and the actual position, if least input increment is not same as least command increment.
I would be more confident, if somebody seconds me.

[stevo1]

Originally Posted by sinha_nsit Finally, you decided to respond. Thank you.

LOL. Wow Sinha...quite demanding today are we? I see a total time from your post to Al's response as 7minutes. I've waited longer at McDonalds.

Stevo

[underthetire]

Usually they should be set the same,yes. Most fanucs from the builders are set to .0001 increment and command. or the metric equivalent. All of it is really a moot point though, it's almost never a control unit input that is the limiting factor, it almost always mechanical. We have some newer machines with .00001 input and command, but they all have multiple scales and high count pulse coders, and were built to be high accuracy machines. Even they don't always position to millionths. The only machines we have now that can really do it are delta tau controlled machines with hydrostatic ways and spindles.

[sinha_nsit]

A better terminology for "least command increment" would have been "least output increment." This is yet another example of Fanuc's Jinglish.