Z Axis Missing Steps - Zero height moving

[damomac]

Hi All,



Just wanting to see if anyone would have some advice to help solve a problem i am having with loosing steps on the Z Axis.

I am in the finals stages of a TEKCEL Router conversion. I have kept the original Pittman Servo motors, I have installed the MACHDRIVE Servo Driver boards & MASSO G3 controller. I have Tuned the servos, The motion of the machine seems good, I have just started the first couple of Cutting projects. During the cutting sequence, I am losing Step count and by the end of the cutting project the Z zero changes from being at the top of the cutting material, to approx 2mm above. This is over approx 5 min cutting file.

I have checked the following:

• Checked the Drive Pulley (Belt drive) on the Servo Motor and it is tight, no chance of movement,
  
• Checked the Drive Pulley on the Ball screw - it is tight also with no movement
  
• Ran the project multiple times - Same result
  
• Connected to the MACHDRIVE driver board to see the DRO, It does not show the error, stays consistent with the MASSO controller,
  
• Checked the Earthing circuit of the main components, all are tight,
  
• Possibly some Noise from the VFD coming through, Spindle power cable is shielded and earthed to the VSD chassis & not at the spindle.
  
• I have reviewed the Soft and Hard limits and dont think there is any issue with the set parameters. These should not be hit during the project,
  
• I have looked over the NC code, nothing seems a miss, using the same generic post processor as I use on my smaller machine, all works OK.
  
• I have reduced the acceleration, it was 100 mm/sec, now 50 mm/sec (Not sure what is the correct speed here) X & Y @ 100mm/sec seem to function well, will need to test further.
  

I am at a loss as to the cause of the variation, i am leaning towards that it could be potential Noise from the VFD but not sure how to test and would like to hear from anyone that has had a similar issue and resolved it, open to suggestions.



Regards



Damien.

 

[breezy]

Damien,

One possible cause could be cutter slipping in the collet. The downward cutting pressure is moving the cutter up. What may seem to be a tight collet is not so when cutting forces come into play.

Generally missed steps is in the downward direction, the motor has to work harder lifting the spindle than moving it down.

Regards,

Arie.

 

[evermech]

Yah the servo drive should have a parameter that allows its position to be only out by whatever is set. So you could set it to say 50 pulses. If the servo position is outside the 50 allowable tolerance it should kick out the servo drive. If the servo alarms are wired properly to Masso then it will shut down the whole system .

the whole setup is called following error in cnc machining, and is an important part of the closed loop positioning that is a major difference between servo as opposed to stepper systems.

having this implemented in your servo setup will tell the difference between position error issues with the servo, or some other mechanical issues such as slippage in the drive system or even the tool holding as Arie suggests

Guy

 

[damomac]

Hi

@breezy, I will double check the Collet, It did seem tight, at the start of each time there was a homing seqence and the tool seemed to be at the correct zero Height, Thanks for the suggestion, I will check and confirm.

@evermech sounds like it could have something to do with it, I will check, initially I would run the same program multiple times in the smae position with the same material, even when no cutting load It wold also change, I will look into the following Error, not sure what the MACHDRIVE setting are.

Do you know if EMI / Elctrical noise would potentially cause it also?

 

[evermech]

I've ran into noise problems twice. One time the servo was jerking around erratically and the other time was with the Masso and it was dropping communication with the USB both program stick and keyboard/mouse.

I think you have a problem more on the line that Arie is suggesting something mechanical, be it slippage in the collet or???

Setting the following error will give peace of mind that everything went were it should go, and eliminate position as the cause of your problem. Plus you should have the servo alarms wired to Masso so that if any servo kicks out because of whatever problem, it will shut everything down at once. I had my servo alarms hooked up incorrectly at first and when one servo drive had an error the other two drives kept running and ran the cutter into a setup fixture ( not good )

Guy

 

[damomac]

I have conducted some more trials, I have checked the collet/tool and ran through, i don't believe this is the issue.

I have ran the program above the deck and with no material, At the start of the program I have checked the 0,0,0 position, via G0 X0 Y0 Z0 using the MDI.

I ran the program and stopped just before the homing sequence, gone to Go Xo Yo Zo, the tool had 'creept' up 6.5mm

I isolated the 3ph supply to the VFD, and repeated the above, ie, Homed, G0 0,0,0, ensured it was the same position at the start as to the previous run. Ran the program, stopped just prior to end as per the first run, G0 0,0,0, there was NO CREEP, Z axis kept its count.

I now turnred the VFD back on, I installed ferrite coils on each end of the Spindle Cable and each end of the Encoder cables for all 4 drives (Dual X).

I repeated the above process, cycled through, stopped and G0 0,0,0, had creept 17mm up. Thought there must of been an error, repeated and then got the same result, 17mm

Any ideas @evermech & @breezy i am at a loss, I do believe it is the HF interference from the VFD.



I have included a screen shot of the profile i was running.

Also, as mentioned earlier, I have always had the Drive Fault output connected from the drive to the Masso, These have tripped before, mainly from over travel and hitting the end stops.

Please let me know your thoughts.

 

[breezy]

Damien,

I isolated the 3ph supply to the VFD, and repeated the above, ie, Homed, G0 0,0,0, ensured it was the same position at the start as to the previous run. Ran the program, stopped just prior to end as per the first run, G0 0,0,0, there was NO CREEP, Z axis kept its count.

I now turnred the VFD back on, I installed ferrite coils on each end of the Spindle Cable and each end of the Encoder cables for all 4 drives (Dual X).

I repeated the above process, cycled through, stopped and G0 0,0,0, had creept 17mm up. Thought there must of been an error, repeated and then got the same result, 17mm

It seems installing the ferrite coils made the error worse!

It definitely is a case of AC interference on the encoder leads.

Questions

• Is the spindle cable shielded?
  
• Are the encoder cables shielded?
  
• Are the servo power cables shielded?
  
• How are the shields terminated?
  

Most cases of interference is caused by incorrect installation of the shielding and that can introduce more interference. Properly shielded cables don't require ferrite coils.

Shields should only be terminated at the control end of the cable and all shields should be connected to the same earth terminal strip, matter of fact all earth connections should be on the same strip and go back to one earth point in the building wiring.

Regards,

Arie.

 

[evermech]

@damomac. Hi Damien,

maybe if you sent a pic of your wiring setup, vfd location, etc. that might be helpful too

Guy

 

[damomac]

Hi @breezy & @evermech,

Thanks for the assistance, I think I have almost got there, I did the following:

• Replaced the VFD Cable with a better quality cable, It also had a join in it, direct from spindle to VFD without joins - Did not solve the problem
  
• Tried a different VFD, The installed unit has some age, thought it might contribute to the issue, Did not solve the problem,
  
• I installed new Earth wires from the Z Axis Motor & Spindle to the common earth in the cabinet, This fixed it, It does not appear to be loosing steps any more, thanks @lilow for the tip,
  

There is still a little bit of noise in the Monitor when the VFD is running, but now the steps appear to be consistent.



Regards

Damien

 

[evermech]

@damomac

thats great. I once had an issue with a servo motor jerking erratically and the electrician finally grab a handful of teck cable running inside the box and just hung them out the door problem solved. Now we always put all 3 phase wiring and vfd in a different box than controls

Guy

 

[stewey]

Damomac, that earthing solution reminded me of our router- an old Esab, converted to a Tekcel in 2002.
We bought it in 2007.
I had similar problems with Z having a party of its own, and tried reterminating the shielding and assorted other ideas.
Finally what solved it was drilling a hole through the concrete shed slab floor, belting an earth rod down there, then earthing the router frame to that earth rod right there. (not via the mains earth wires)

The thing is after you've solved them, you tend to give no further thought to the idea or solution...!

 

[grindaur]

I am experiencing similar problems with my Sherline?Masso setup. The last three attempts have resulted in the z axis losing zero (twice) and attempting to machine a hole to China (my tooling plate will never be the same). Today it was the X axis that lost zero and moved approximately 25mm in the positive direction, then continued to run the program machining air but emphatically showing on the screen that no unplanned movement was made. I was concerned so I had simulated the program with both Fusion 360 and GWizard, and nope no glitch in the G code.

Has anyone come up with a line filtering fix, I am assuming this is a conductive emmisions most probably on the AC line. Is there anyway to loop in an absolute encoder to verify movement/or report unplanned movement?