I finally was able to get my G0704 converted to CNC.  I can run all the axis and they are very smooth while jogging the machine.  When I loaded the roadrunner program, the y axis sounded horrible.  It was missing steps and making a loud grinding noise.  As soon as I stop the program and jog the Y axis, it is perfectly smooth again. 

I am running at 48 volts, nema 23 on the Y with a Keling 5056D driver.  I use an UC100 for motion control.  Dell PC with windows 7.  I have turned off and on the 1/2 sherline mode, tried different kernal speeds, different velocity/acceleration.  I am currently microstepping at 2000.  I have tried 400 (the lowest setting on the driver) and the grinding sound is still present during a program.

I thought maybe the Y axis was under powered so I turned the driver up to 4 amps and it did not help.

I am at a loss right now.  Any help is appreciated!

I am not certain but I presume the UC100 does the pulsing rather that the parallel port driver, if that is the case Sherline and Kernel etc will have no effect.

Have you got the steps per unit set correctly?
Maybe you have jog set to a % of full rapid and that is why you can jog but not run code.

If you attach your xml I will have a look but I may not be able to see all settings, it depends how the UC100 works.

Hood

Here is the XML file.  I believe the steps per unit are correct.  The X and Y axis are the same and the X did not have any problems, only the Y. 

I haven't turned the machine on in a couple days until today.  Odd thing happened.  When I powered up the motors they began to move very slowly and very jerky.  Hitting the estop in Mach3 would not stop them from moving.  Now, even when jogging, all motors are very jerky and noisy.  Not sure what is happening here. 

Ok afraid because I do not have a UC100 I am unable to view any settings in the plugin.
The issues you had today however are pointing to noise being a problem in your electrics/electronics. The motors should not be moving at all unless the UC100 is sending out pulses, I presume the DROs in Mach were not moving when that happened?
Check your wiring, ideally all wiring should be shielded and properly grounded, especially 5v signal wiring.
Hood

Thanks for the tip.  I am using shielded wire on everything.  Is there anything special I need to do to ensure it is properly grounded? (other than connect the wires to the breakout board/ motor driver.)

Also, any idea why the noise in the motors would just now be starting?  It has not happened before. 

Thanks for your help! 

Do not know if any of this applies to your wiring but here are a couple of thoughts.
You should route the shield to a  single point in your cabinet and that point should be connected directly to Earth. Only one end of the shield should be connected and your shield should not be getting used as a common/0V conductor.

Why it has suddenly started happening?, I would ask, has it really just started? You seem to have been having issues with the stepping so maybe it is just an escalation rather than a new phenomenon.

Hood

I have 18/4 wire with an uninsulated wire in the bunch as well.  The manufacturer refers to the uninsulated wire as the "drain" wire.  Is this the shield you are referring to?  The wire is shielded as well by a layer of what looks like aluminum foil.  Do I need to run the "drain" wire to ground and leave one end unconnected? 

I tried a parallel cable to see if the UC100 was the issue and I am having the same issues.  I think you are correct about noise in the system.  Now I am just not sure how to remove it.

I never heard of a drain wire! Maybe that's why my discipline was process / mechanical.

A shield provides for protecting what's inside of the shield from coupling with signals that are  outside of the
shield. Likewise it keeps signals inside the shield from being getting outside of it. If the shield was a continous metal tube
one would say that there is 100% shielding. The shield can provide a least resistive path to earth ground. You don't send the signals traveling outside the shield  to another circuit connection as that would be just coupling the interferjng signals
to some other circuit. Now what hapens inside is a different story. If an offending signal, noise ( noise is not understood by us),  is inside of the shield then it can couple and get mixed in with the signal on the wires. The uninsulated wire can offer
a less restrictive path to ground and help  minimise the the noise. So guess one could say it is "drain for the conduit"  ie; the conduit is filled with liquid and want to get rid of it...............

Yes leave the one end unconnected.

Noise can be a real bear to find and get rid off or control, but, a seperate ground system for noise is the first step to elimination.

As the saying goes is the above "clear as mud"? 

RICH
  

Thanks for the reply Rich.  I will work on getting the shielded wires grounded. 

I am using a UC100 Motion controller.  I noticed I do not see the noise issue (ie motors moving when not supposed to) when I unplug the USB.  Could it be the USB cable is the culprit?  It is a cheap cable and does not have ferite rings or shielding.

I grounded the drain wire on the wires coming out of the board.  I haven't seen the noise issue anymore.  Now with the UC100, I cannot get the motors to move.  Mach3 shows they should be moving.  I didn't change any of the wiring, not sure why this is happening. 

I tried a parallel cable and a different computer with Windows XP.  I can jog the motors now, but only for a few seconds.  Then the motors start skipping steps and eventually stall.  Mach3 still shows them as moving.  I am at a loss on what is happening with this machine.  I have tried different accel/velocity combinations with the same result.  If I turn off the power to the board and motors, then turn it back on, I can job again, but experience the same problem where they quickly stall out.