Hello All!

Hoping to gather some brainstorming ideas from all the big brains here!  :-)

I have a problem with my machine, running MACH4 and POKEYS57CNC.  I noticed after the finish cut was completed that the Z height of the cut was normal at the start.  However when the job was finish (at the final cut) it was ~ 5mm too deep.  The job looks great if it wasn't for the erroneous cut.  There is no obvious stepping in the cut or signs of a problem.  So I am assuming that the z depth problem is gradual on each pass.  The job is 183 mm wide with a 0.1mm step over.  So mathematically I am assuming that for each step over the z axis is out (negatively) by 0.0027mm [183mm wide / 0.1 mm step over = 1830 steps, 5mm error / 1830 steps = 0.00273~ mm]  At this exactly how the cut looks.  It's smooth and "flat" looking, otherwise a good cut.  Now the way I see it is the z axis error is on the positive travel direction (up), when it's not cutting vertically.  My logic is if the actual zaxis move up is less than the programmed move, then when the zaxis moves down (and there is no error with the downward travel) it will move down too far (the same amount as the error on the upward movement).  Now this really baffles me as there should be little or no resistance on the zaxis motor on the z+ movement.

A bit more context on the machine zaxis setup. I have a 8mm x 1mm pitch lead screw with a flexible shaft coupling.  The stepper driver is set to 800 pulses/rev with a Nema 23 Closed Loop Stepper Motor 2.0 Nm/283.28oz.in of torque.

I've checked for interferences on the zaxis travel, there are none.  I also ran a reliability test moving the zaxis vertically up and down in the same XY location for 1000 cycles.  That is also ok.  At the end of the 1000 cycles the cutter is exactly where it should be in the zaxis with no error.

I have also checked the flexible coupling set screws to ensure there was no motor shaft slippage.  All checks out there also.  When I installed the motor I made a mark on the end of the shaft so I could align the set screw properly on the shaft flat.  It's right where it should be and the set screws are still tight. (there are 4 set screws total, 2 on the flat and 2 90 degrees off the flat).  There is 1 flat on the motor shaft.

I was also thinking about motor amperage.  However the stepper driver is outputting their max 5 amps which is the rated current for the motor.

I've got a lot of cutting hours on this machine and never had this as an issue.  I'm scratching my head here.  Obviously there is something I am missing.

Any feedback / help / ideas greatly appreciated!

Hope everyone is safe and healthy.

Best regards.

Anyone want to bite on this one?  :-)

What about electrical interference?

Have you considered the possibility that the cutter is being pulled out of the collet or chuck that is holding it. I use collets to hold my cutters and if I take a heavy cut or try to cut too fast the cutter will get pulled out of the collet. See if that is a problem.

That's a good  one stevehuck.  I'll mark it with marker next time and see if it moves. 

I'm chasing other gremlins at the moment.  Seems like my entire system is a moving target of stepper motor issues (or communication to them)

Thanks,

Scott

I am reviving this thread as I am still having issues.  And wow time files, I see the original post was 2 years ago.  I have been doing a lot of laser work with the machine so the Z-Axis problem has not been an issue.  I am getting back into cutting and I am still struggling with the original problem as posted here.

Problem:  Z axis gradually drops (in error) over the entire cut job. It is so gradual that the job looks perfect until you measure the actual thickness at the end of the cut.  I just finished a piece that is ~ 550mm x 200mm, finishing pass along the 550mm length.  The job is 4mm too thin at the last path.  I have cut multiple jobs with the same error, always negative Z.

What I have tried/checked to root cause / fix:

• Compared holding position of spindle (in holder) and cutter (in collet) to ensure they were not slipping.  No problems, all check out ok
• Ran a 500 dry cycle test on the Z axis.  Each cycle moves the full range of movement. Compared cutter tip height by probing when dry cycle is done.  No issues.
• Changed coupler type between Z axis stepper motor shaft and lead screw to a style that should have more holding power
• Replaced anti backlash nut for Z axis lead screw
• Verified the work piece is not moving
• I checked the gcode cutter path to ensure that the correct z height value is being generated.  All ok.
• I've put a pretty beefy stepper motor on the Z axis to limit the possibility of stepper motor slippage.  I installed a Nema 23 2.0 Nm/283.28oz.in

My thoughts on other possible issues / fixes:

• I have a system with v slot rail and wheels.  I am thinking about changing the Z axis movement to ball screw with linear rails.  I am wondering if the v slot system is binding, albeit very slightly.  However seems very weird that the error is only in 1 direction (-z)
• I do not have shielded cables.  Although it would be a huge job to tear apart the machine and replace and/or shield the existing cables of course I would do it if I thought it would fix the issue.  I ran the test with and without the spindle running.  I am making a strong assumption shielding is not the issue as it is irrelevant whether I am actually cutting a piece or not.  (There is no Z axis height issue with the 500 dry cycle test)
• I am also questioning the holding power of any coupler between the stepper motor and lead screw.  The lead screw does not have a cylindrical, solid shaft on the end with a flat.  I am seriously questioning how much holding power the coupler can have with just 2 set screws and friction against the threads of the lead screw.  One thing I haven't done is measured how much lead screw is protruding from the holder bearing.  I will try another cut test and measure this value before and after.  If the lead screw is slipping inside the the coupler than it should show the difference.  However with 4mm of z delta per job it wouldn't take long for the lead screw to totally come dislodged.  Which is hasn't.  So the likelihood of this being the problem is low in my opinion.

I am baffled honestly.  I'm just grasping at straws now.  I am not one to quit or say I can't do it.  But honestly I don't know what else to try.  For sure something mechanically, electronically or with the software / firmware is an issue. 

Other than scrapping the whole machine and money that went into the machine build, I'm stumped.

Any thoughts / brainstorming is appreciated.

Thanks everyone.