Machine accuracy Testing

[pertneer]

Anyone have any suggestion on how to check the accuracy of my table over the length of travel. With it being a 4x8 foot table it make is hard to check it precisely. I am a Machinist and I have checked it with gauge blocks and an indicator, but this is limited to like 6 inch range. There is quite a difference between 6 inches and 8 feet. With a tape measure it looks very close, but with building this from scratch I want to know for sure its accuracy before I scrap out a bunch of steel.

[RClippa]

I just went through this on my table. I was convienced that I needed to use dial indicators, blocks, and gauges to get the machine accurate. What ended up providing the best results was the tape measure. I basically used a razor to mark the position on one end. Then ran it to the other end and marked again. using a measureing tape and a friend I got the table to a point where the marks were on the same edge on the tape. For repeatablility I put a dial indicator near the home location and ran the table around and brought it back to see if it lost any steps. once I did these two setups to the machine I am moving the plasma more accurately than it can cut. In the end a plasma table is not a precise tool, much better than by hand, but it will never be considered precise by a machinest.

[plain ol Bill]

Put a mark 6" from 0. Tell your machine to move 6". If it is short increase the number of steps, decrease if it is long. Do this until you are dead on - then move your mark to 2 ft. and repeat, then 4 ft. and repeat, then 8ft. and repeat. By the time you get out to 8 ft. you will be making very small adjustments. My table is very accurate out to 8 ft. and beyond.
I can cut a 3/4" hole and have to screw a bolt into the hole - that is close enough for me.

[pertneer]

I just went through this on my table. I was convienced that I needed to use dial indicators, blocks, and gauges to get the machine accurate. What ended up providing the best results was the tape measure. I basically used a razor to mark the position on one end. Then ran it to the other end and marked again. using a measureing tape and a friend I got the table to a point where the marks were on the same edge on the tape. For repeatablility I put a dial indicator near the home location and ran the table around and brought it back to see if it lost any steps. once I did these two setups to the machine I am moving the plasma more accurately than it can cut. In the end a plasma table is not a precise tool, much better than by hand, but it will never be considered precise by a machinest.

So you held the razor blade on the z axis and moved to make the measurement. I guess that would be a good way and should be closer than I have so far. I do understand it is only a plasma cutter and will not be as Precise as CNC mill. I just want to make sure it can move to a position that i tell it to, to start with. I wrote a program to rapid around the table to make sure everything worked properly I will just set my indicator up to make sure it returns to the same position after i get the length tuned in. Thanks for the ideas.

[SignTorch Vector Art]

To check squareness you can cut across a sheet then flip it over, cut a strip parallel to the first cut, if the strip width is different on each end then the machine is out of square by half the difference.

Digital motors maintain super accurate step distance, so if it works at 8" it will work at 8' with equal accuracy of plus or minus one half step (on distances that are not multiples of the step distance).

Any inaccuracy is usually due to machine slop, backlash or misalignment. And some limit switches are more accurate than others when it comes to detecting the home position exactly.

Stepper motors are subject to miss steps and get out of sequence if overaccelerated or overloaded, and they get weaker as they go faster. You can experiment to find a machine's speed limits by making many series of rapid starts and stops and turns and reversals, while making sure it lands on zero when homed. Then keep your operating speed and acceleration settings as far below the machine's maximums as possible. On a heavy machine axis, a stepper can also jump steps if over-decelerated. Servo motor systems don't have slippage like that.

But then with plasma cutters there is a whole host of other torch speed/height/material variables to worry about. Not the least of which is accurate kerf compensation, which determines whether you cut on or inside or outside the line, which can have a huge impact on the final appearance of art.

[FiveORacing]

I use to have to take all the Police tape measures to Weights and Measures for certification. In all the years I did it, not one tape, usually Luftkin, Craftsman or Stanley failed the specs which I think were 1/8 for 25 feet for steel tapes. Nylon tapes were a little less stringent. A couple failed for loose tangs. Not sure how accurate my table needs to be for what I'm making, but a 1/8th in 25 feet will do for me.

[pertneer]

I mounted a scribe in the torch holder. Went to a spot on table frame touched and scribed a line on table, moved to other end and did same thing. Then using a tape measure I put the scribe line in the center of the 1 inch mark on tape (wife holding the stupid end(of the tape measure )).I did this to take out any error of the movable thingy on end of tape. Then I checked it to the center of the other scribe line. It was about half the width of the line off over 101.5 inches. I think it will be close enough. I also did the same thing for the other axis and it was as close as the previous one. I did this with moving the axis with a "g0" move at 450 in/minute and a "g1" move at a much slower feed rate, around 10 in/min (this one took awhile). So I do not think it lost any steps along the way.