Problem with 4th axis on CNC

I have an E bay router table that I mounted a dovetail on for a nice tail stock and have had it for about a year and a half. I have cut a lot of ring billets on it using cue billet from Kelly.

Suddenly I have an issue that I can't figure out what is causing. Cut a five slot billet with strips about .225 wide. The rotation between the first and second slot is supposed to be 72 degrees and mach 3 says that it did rotate that much watching it live. But it didn't rotate right down at the head. It comes up a few degrees short causing those first two slots to be too close to each other. Then it rotates and cuts the next three perfectly at 72 degrees which of course makes the last slot too far away from the first one by the amount the first one was short. So it looks double bad.

I thought it may be how I had the stock mounted using a single spur driver which I have used all along and thought it may have come loose but it didn't. I mounted another test piece up directly in the chuck and it cut exactly the same pattern. I then ruled out random steps missing by running the program a second time on top of a billet already cut and it cuts all five slots in the exact same spot with the first one short of 72 degrees.

The g code for this is really straight forward and cue billet generated it correctly I have checked it over closely. There is no other sign of issues with the hardware or the software that I can see. All the slots were perfect width for the strips.

The really shitty part is I didn't notice until I cut a ring and glued it up in a cue. It's pretty subtle until close inspection if you're not diligent on inspecting everything like many of us who get trusting of our process' and equipment. I was complacent because this setup has never failed me before now.

I rebooted the computer, restarted mach and regenerated the g code in cue billet even though I know it wasn't flawed and it's still cutting like this.

Has anyone had a problem like this before? I hate to buy a new 4th axis if it's not the issue. I'm stumped.

Thanks

MVPCues said:

Sounds like backlash. Set up a fresh dowel and Take all backlash out of your indexer by jogging the A axis in the same direction a CueBillet program runs. After jogging what should be way more than necessary, zero the A axis. Run the program. Check indexing in piece.

Repeat, with another dowel but jog the A axis in the opposite direction for a bit to introduce the possibility of backlash. Zero the A axis. Run a the program again. Compare indexing between the two dowels.

Click to expand...

Told you I learn everything the hard way Kelly.

Never had this problem until just recently when the timing gear set screw came loose on the stepper motor.

Thanks for putting me onto the cure.

ELBeau said:

Any luck tracking down the issue?

A loose set screw on the timing pulley could to it, just like backlash. On the first rotation it is taking up the space between the screw & the flat on the shaft. After that it the axis keeps moving forward the correct amount since that gap has been eliminated.

Click to expand...

That was the problem. The gear was falling off the shaft of the stepper motor on the 4th axis.

Without some help I may not have figured it out until it fell on the floor.

The idea that is was backlash pointed me in the direction as the stepper motor is connected to the drive with a timing belt and other than a loose gear there isn't much in that setup that can have backlash.