[BuiltByBrooks]

Its easy to set up axis calibration on the mach router program but how do i do it in the plasma, my table is way off and in the plasma setting their is no calibration...

[BuiltByBrooks]

anyone?

[stirling]

what screenset are you using?

[Torchhead]

You don't need axis calibration to setup MACH. All you need is a tape measure, and a calculator.
Chuck a small drill bit or pointer tip
Zero the X axis and use Z to mark the table where it is.
Use the MDI window and send the axis out 10 inches using G00 X10 [enter]
Now for the math part:

Measure from the start point to where it is stopped (center to center)
write down the number 10.000 (your distance you told it to go)
write down what the actual distance is (use decimal notation)

Now subtract the smaller number from the larger one to get the difference.

Lets say you read 10.125 on the actual tape. you find there is a difference of .125 (error)
Now open the motor tuning for that axis and write down the Steps per unit number.
With our calculator divide the error (.125) in this case by the original 10.00 you called for. .125/10.000 = or .0125. What that represents is a ratio of the error.

Take the steps per unit you wrote down and multiply by the error ratio. Now you have the number of steps to add (if you are short) or subtract to get closer.

BEFORE you change the Steps per unit send the axis back to zero using G00 X0 THEN put in the new steps per unit.

Run the test again and you should be much closer. Here is the magic part: The longer you make the distance to easier the error is to see and correct. After a few tries you will be dead on at 36 or 48" .

Do it a couple of times and it's really fast to calibrate the axis. The calibration macro in MACH does essential what I outlined but does the math for you and lets you type in the numbers.

If it's a plasma table lay a piece of MDF out on the grid and clamp it down so you have something you can mark (and even use as a scratch pad!).

It's basically taking a reading, finding the amount of the error, converting that to a percentage of the total and using that percentage to adjust the steps per unit. It's pretty easy to see that is you are off by 5% over a distance, adjusting the step count by 5% would correct it. It's also obvious that the longer the test run the greater the accuracy of the process.

TOM caudle
www.CandCNC.com

[garydunlop]

hi there has anyone found out a way to speed up the x and y movements of a table as it travels from cut to cut