Equipment
Vendors

SITEMAP

BLOG

Last Updated:  07-12-13

Baker's Gas & Welding CandCNC Cascade Metal Designs Corel Free DXF HYPERTHERM LubeCorp Manufacturing Inc. ARTSOFT MACH3 PRECISION PLASMA LLC Steel F/X SHEETCAM COPYRIGHT 2013 DESERT FABWORKS LLC - ALL RIGHTS RESERVED- WEB DESIGN BY: DESERT WEB MARKETING LLC    -     Contact Us   -
BUILD PAGE 19
CONTINUED ON BUILD PAGE 20
PROBLEMS SQUARING THE GANTRY

We have had the table up and running for over 3 months now.  We have been staying pretty busy.    The table is averaging about 3-6 hours of actual running a cutting time per day.

99 % of the work has been plasma cut art objects.  We have not had very many structural component jobs until a few weeks back.

When we were cutting the plasma art we did not notice that our gantry had become out of square.   With the art files the small out of square problem could not be detected.   We noticed it when we began cutting square plates for some industrial size smokes and BBQ’s.

This is a run down of what we found and what we did to correct it.

This picture shows the we were about 1/8 out of square.

The problem we had in tracking the problem was that this was showing up on a small square only 4x4 in

1/8 out of square on a 4x4 would make you think that across the 48in table would mean the gantry is very far out of square but it was not.

Checking the Gantry we found that it was perfectly square to the sides.

So where was the out of square problem coming from?

We consulted a few sources including Precision Plasma and CandCNC.  We looked to rule out that the software was not the source by checking the G code and looking for any programming problems that could create the error.

The G code was good and there were not settings in Sheetcam or MACH3 that were creating the problem.

Because the squares were dimensionally correct ie 4 in wide by 4 in wide our kerf and other settings were dead on.

Talking with Precision Plasma we found that we had a very uncommon problem that had not been seen on other tables.   They told us about an upgrade that they had produced for the X axis Carriage and sent one out to try and eliminate that from the problem.

On the left is the original carriage and on the right is the new carriage.  You can see the added set of bearings that help stabilize and guide the carriage along the x axis.


This new carriage is a one piece welded assembly.  The design is great and ads lots of stability to the carriage.

The problem for us is that it did to solve the out of square problems that we were having.

We would perform a tweak and then cut a test square. This quickly became a waste of time money and materials.   So using some scrap brass tube and some misc supplies we created a pen assembly for the machine.

Below is a picture of the Pen holder.  Its made out of brass tube with a reducer and cap on the one end and a cap that has been drilled out to fit the tip up the pen.   It fits a Sharpie perfectly and we installed a spring to deal with the variance and take the place of a THC sensor.  We also made it fit a pencil.  Then mounted it to the existing holes in the Z assembly with tube clamps.   This set up was cheap and simple and worked great.

We laid down a piece of sheet metal taped some paper to the metal and began drawing right angles with the keys.    

Its important to use the keys instead of running a cut program to eliminate everything else as a problem.

Once the line is drawn I lay a square on the line and find out what I need to do to correct it.

I would then use the following procedure to make a correction.

Loosen the 4 bolts holding the gantry to the carriages on each side.  Then

I would unplug the drive motor I wanted to make the correction for from the Bladerunner box.  Move the now free spinning gear a small amount in the direction of correction.   Leaving the other side of the gantry locked.

This moves the gantry a slight bit.  I then plug in the drive motor.  Turn off the drives and back on so that everything is in lock and then tighten the 4 gantry bolts.

Then I re-run the line test.

As you can see from the number of lines this process took a bit to complete moving in very small increments.   Eventually it was drawing perfect right angles and perfect squares.  I confirmed this by running a G code file with THC off of course and then reinstalling the torch and running actual cut tests.

Once it was perfect everything was tightened to prevent movement in the future.