New CNC user here. I’ve assembled my Elite Journeyman and working on flattening my spoilboard. I have a simple file setup in Carveco to flatten using my 1" bit from Spetool. The toolpath is set to start in the middle so I’m X/Y zeroed at the center of the cutting area. I zero Z using a piece of paper (no probe yet).
I started with an aircarve and all went as expected. I had zero’d Z at 1" above the surface and left the router off during this process. Once the aircarve completed, I zero’d Z at the surface and reran the toolpath with the router on.
The carve is set to remove .06" of surface with a safe height of 1". When the program starts, Z moves to 1" and X and Y move to the center of the cutting area. Z slowy lowers to the desired cutting height of .06 and X/Y start moving (circular pattern). All looks well on the Masso screen.
After about 2-3 revolutions, I notice more material is being removed. I thought this might be due to differences in height of my MDF wasteboard. The problem continued until it was taking off almost 1/4" of material. I E-stopped at this point. The Masso screen still said Z height was -.06.
After much troubleshooting and headscratching, I started the carve again over the now destroyed center portion of my wasteboard. This time I watched the Z axis ballscrew coupler and saw it move more than once, each time the carve getting deeper. Masso still says Z height is -.06 despite the movement in the ball screw. I E-stopped again and turned to the internets to look for solutions.
Sounds like you did step one and two, recognize a problem & identify the issue causing the problem. I would start with checking to make sure the screws on the coupler are tight. There’s 2 of them and they face opposite directions. The top one clamps onto the shaft of the motor, and the buttom one clamps onto the screw. Both need to be tight to be effective and accurate.
I’ve done more testing on this and am thinking it may be EMI. I originally had the Makita power cable in the drag chain and plugged into the Onefinity power supply. I’ve now removed it from the drag chain and am powering it from a completely separate circuit.
After doing this, I was able to run a flattening pass with no issues. I did do some further testing though and found that if the router is under a higher load (cutting deeper or wider passes) then the Z axis will still occasionally move. One odd thing is that when it moves, it always goes DEEPER - I’ve yet to see it move in the other direction.
Another thing that bugs me is that the Masso display doesn’t indicate that any movement has occurred. I thought the Masso closed-loop steppers were positionally aware - if this is the case, how could the ballscrew advance without it being reflected in the machine value for the Z axis on the display?
The Masso closed loop stepper is self contained with the driver in the stepper motor itself, The input from the controller is only the step/direction with no encoder feedback to the controller.
In your case there could be a few sources of the movement, EMI/RFI as you’ve mentioned could be inducing voltage on the step line that the driver interprets as commands to move (with the direction value low so it moves down).
Another potential source is that the stepper is malfunctioning and the encoder thinks it has moved and is attempting to counteract that movement - this would not be reflected in the DRM output on the controller. If you have eliminated EMI as a source I would suggest emailing support to troubleshoot further.
a router power cables does not emit the same type of EMI than a VFD-to-spindle cable, but the recommendations in my VFD manual could be considered as generally valid:
Use shielded wiring for the motor cable and all analog and digital control lines.
Allow the effective shield area of these lines to remain as large as possi ble; i.e., do not strip away the shield (screen) further away from the cable end than absolutely necessary.
Avoid parallel arrangement of low-level signal wiring and power-carrying or noise-prone conductors.
On the Original Series, all stepper motors cables are unshielded , I don’t know for Elite Series.
I agree to what Tom and Derek say. A closed-loop stepper driver reads the encoder and tries to correct it by itself for smaller deviation, but it’s only if there is a major difference between step commands and read steps, that it triggers an error condition that stops the CNC controller program.
Also Tom is right, if the stepper driver does not receive a STEP command due to cable fault or EMI, it will not expect to read it on the encoder.
Great feedback all. Additionally, I notice that during heavy load operations I get occasional flicker in the LED lights in my shop, and at least a couple of times this occurred at the same time as the random drop in the Z axis. This occurred AFTER I moved the router to a completely separate circuit.
If EMI is the culprit, what else can I do to protect from it? Could this be caused by an issue with the Makita router? I’m thinking EMI from the brush contact maybe? And if so, could this become less (or more?!) of an issue as the brushes on the new router wear in on the commutator?
a spindle and VFD make more dirty EMI, but it can all be shielded with shielded cables and putting the VFD into an earthed control cabinet. And a spindle (an induction motor) has no commutator. The EMI that comes from the carbon-brush commutators of the router (a universal motor) cannot be shielded for the part that comes over the air, because it is air-cooled and would not like to be enclosed in a faraday cage. Some people used a shielded power cable however, for the part of the EMI that goes through the cable. Also snap-on ferrite cores on the cables can help.
Another thought is maybe not EMI, but dirty power. Is anyone running a UPS on their controller to clean up the incoming power?
1 Like
Aiph5u
(Aiph5u (not affiliated with Onefinity))
10
Hey Charles,
for this it is always advisable to use a line filter. But the EMI that comes from the commutators over the air remains.
But as you confirmed it got better when removing the cable from the drag chain. Question: Are the cables to the steppers shielded? If not, that’s where I would start.
The motor characteristics of a so-called universal motor (commutated series-wound motor) is that it is slowed down by the load (unlike a VFD on a spindle) and the more it is slowed down, the higher the current, thus the higher the EMI it emits (see also here).
Bone stock Elite Journeyman so the shielding I have is what Onefinity provides (which looks to be none - I see a jacket on PART of the stepper cables but the rest are loose wires).
Hey! Where are you at with this problem? I just got my Elite series setup, went to surface it today and am experiencing the EXACT same thing as you. I’m taking off 1mm and by the time the router goes across the board once it drops about 1-2 mm. Same for me as well the Masso screen doesn’t say the Z dropped. Also when I went and tried a second time to surface it took out even more material.
, I don’t know for Elite Series.