Hello all
  At work we have a new 4x4 cnc router for about 3 months now and have not got it up and running properly.  It was basically a fire sale price to good to pass up.  This is a HD machine with a 5hp HSD tool change spindle and a 6 place finger style changer rack.
It is a servo drive machine with all Delta products in it.  I am guessing it got stuck in a warehouse and never got inspected or run.  We had to replace the controller card and one of the X axis servo motors. 
  We finally got everything working on it but with the type of work we do in it we are seeing position errors after the programs are run.  We are cutting 2mm alum sheet to be used in computer enclosures.  This requires a lot of parts and small holes for fasteners.  I generally run a 1/8" bit to cut anywhere from 3.25-6.35mm holes in the sheet to be used as hold downs for the parts before they are cut out.  This can add up to several hundred holes.   At the original acceleration setting in Mach3 I was getting about a .5mm out of round condition on holes under 10mm and after runing the program and returning to zero it would be off 2-4mm from my original start point that I would mark on the sheet.  But if I was cutting larger holes or parts without the rapid movements like in the smaller holes then I did not see the positioning errors.  I have tried turning down the acceleration rate in the Motor Tuning for the X and Y and it had reduced the error but not completely.  I can do a REF all Home and GOTO Zero and it will return to the original zero setting but that not practical.
  Any help here would be greatly appreciated.

James 

PS - here is a test cut I did,  the small holes were all out of round but the larger ones were fine.

Looks like it could be servos requiring tuned correctly.
Hood

Hood
  I did a few more tests.  Using the procedure in the V3 config manual using the indicator and gauge block.  I ran the short program they used 50 times and I get an incremental error while testing the X and Y axis.

  With a back and forth movement of 4"(101.6mm) and a feed rate of F1000(mm) at the end of the 50 reps I was averaging 0.015"(0.381mm) on the Y axis and 0.035"(0.889mm) on the X axis.
  Backlash on the Y axis was 0.013"(0.3302mm) and 0.001(0.025mm) on the X axis.   
  I havn't tried to get the mechanical backlash out of the Y yet.  When I tried using the Backlash settings (putting in 0.3302mm) the accumulated errors on the Y went up to over 0.060" an the machine did not go back to the original zero point at the end of the program.  it went to 0.3302, the same as the backlash input.

  If this was just the motor tuning then the length of movement might be off but it should still return to the original zero and it is not doing that.

Any ideas??

I would love to start using this giant paperweight.

thanks

James

I would think it has to be encoder error. Clearly the servos are not going where they should and the encoders should be reporting that and correcting. I would look at how the encoders are driven. If you have a loose coupling the error might show up more with a rapid speed change than a gentle one. Also could be electrical noise interfering with the encoders.

Gary
  The Servo motors are Delta ECMA C30807PS  750W ac.   The Servo Drives are also Delta - ASD 0721-AB  750W.  All run through a eCut V1.1 breakout board.   The electrical cabinet is pretty good But with the long runs of cables I wouldn't know where to look for interference.   Here is a shot of the cabinet.  the cables have been tidied up since.

  With a back and forth movement of 4"(101.6mm) and a feed rate of F1000(mm) at the end of the 50 reps I was averaging 0.015"(0.381mm) on the Y axis and 0.035"(0.889mm) on the X axis.

    If this was just the motor tuning then the length of movement might be off but it should still return to the original zero and it is not doing that.

Any ideas??

Using AB servos and drives, my mill did exactly the same.
Going from Single Ended connections to Line Drivers on the step and dir. lines corrected it for me (Thanks Hood)
Russ

Russ
  I am a bit out my league here :-)  where are those changes you referred to??

Thanks

James

You can check the terminals at the BoB to see if they all share a common. if so, they are single ended. I'd check that first.
Also, check the docs on the drives to see if there is an option for either, mine were optional but highly recommended line drivers.

If needed, there are line drivers that you can wire in.

Some lit.:
http://www.omega.com/techref/das/se-differential.html

Russ,
Yes with digital servos the effect would exactly like a stepper missing steps. I think you probably have it right.

We finally think we have found the problem with the machine.  working with the service people,  we switched the y and z axis servo controllers and the error went away, so it seems to be a bad servo controller.  Taking a closer look at the controller we noticed soot on top of it :-(
Anyway,  the unit is a Delta ASD-A0721-AB,   it an older unit not being produced anymore.  If anyone knows a supplier in Canada or the US that has some of these,  please contact me with the info.   I have found then in china but would like to get them a little closer to home.
Thanks

James