The ghost is back in the machine

[martyscncgarage]

Would a stepper create these kind of errors?

Yes, open loop steppers (No encoders) could create errors if they are not sized correctly, proper power supplies not chosen, or you are pushing them too hard. (Post the cut sheet on your stepper motors, stepper drives and what voltage power supplies you are using)
Drop your feed rates way down and cut your MAX RATE (rapids) in half and retest.

How did you size your steppers to the machine?
How did you tune for MAX RATE?

Marty

[tuffduck]

Here are the cut sheets.
Jon from stepper motor Canada helped me size the motors. I followed a guide from the ACorn forum to tune for the max rate using a few calculators online.

[martyscncgarage]

Here are the cut sheets.
Jon from stepper motor Canada helped me size the motors. I followed a guide from the ACorn forum to tune for the max rate using a few calculators online.

Your motors can be configured several ways. How do you have them wired to the drives?
What is the DC voltage of your power supply? With an inductance value of 8, the optimal motor operating voltage should be about 90VDC
The power supply spec sheet lists a 24VDC and a 48VDC power supply.
What stepper driver are you running? Make/model/cut sheet?

[tuffduck]

The power supply is 48volt., 20 amps. The drivers are gecko 201x.

[tuffduck]

Sorry the motors are these. The attached image from the website was the wrong one. Model 86J18156-845.

[martyscncgarage]

The motors and drives look fairly reasonable.
Let's define errors. You have MPU/PC resends in your error logs. Those are most likely caused by EMI noise in the cabinet. That really needs to be found and corrected first.
Did you try another PC, ethernet cable and or Beagle Bone?

Artifacts in a part are usually caused by lost steps in an open loop system such as yours.
This can be caused by noise in the step signals between Acorn and the drive, steppers/drives/power supplies that are not matched up together. Poor choice in sizing the motor to the application. Tuning the stepper to the application as well. Many folks worry about rapids and shoot the axis as fast as they can until the stepper stalls or is making noises and back it off a little. I back it off by 30% and do test cuts. Too aggressive a cut can also stall a stepper.

Because closed loop hybrid motors and drives have come down significantly (yes, they are still 2-3x more expensive than simple closed loop stepper systems) Is why I no longer use open loop steppers for the most part.

I believe I suggested unplugging everything from Acorn (the terminal blocks unplug from Acorn) except the ethernet and the Centroid supplied Acorn power supply. Then modifying the Inputs to not look at drive faults or spindle faults, inverting the Estop input so Acorn can essentially run a program. Then run the program a few times or loop it. Then every so often stop and look at the logs to see if MPU/PC Resend problem has resolved.
Essentially you are taking Acorn back to a basic bench test. Make sure any VFDs/spindle drives are powered off. Look at the error logs, F7 Utility, F9 Logs, F1 Errors, if the resends persist then it is likely between the CNC PC and Beagle Bone. If the problem goes away completely, you should have a look at the wiring in the cabinet.

Just some suggestions.
Good luck,
Marty

[martyscncgarage]

Have you left that VFD turned off and run some programs and then looked at the logs?
(All axis connected and in control of the machine) Be sure to invert the SpindleOk input so you can run this test so the Acorn thinks the VFD is happy.)

[RogDC]

In addition to Marty's recommendations I have a few questions/comments from your photo of the cabinet configuration .

• Are you using a common earth ground point on your cabinets - not just your back plane

• The acorn appears to be mounted on a painted back-plane, is the back-plane grounded to the common earth ground?

• Are both cabinets bonded to earth ground at the entrance to the control cabinet?

• You have armor flex running in the bottom of the control cabinet which does not appear to be bonded to earth ground, is the is the VFD AC or VFD output cable running thru the armor flex?

• Are the control/alarm cables running from the VFD to acorn in shielded cable grounded at the common ground point or entrance of the control cabinet?

• Appears you are using aluminum diamond plate for the control cabinet back-plane, is this anodized or does it have a clear coat?

• Do you have an RF noise filter/suppressor between your VFD power input and VFD breaker/contactor?

[tuffduck]

Hi,
There is a common earth ground to the steel cabinet. The back plate is grounded to that as well. The Acorn is mounted to a piece of laminate and is on computer motherboard standoffs, grounded to the common ground. Both cabinets are grounded to common.
The armour flex carries the VFD output cable. Does the this covering need to be grounded?
The aluminum plate is anodized.
There is no FR filter on the VFD power input from the breaker.

I have replaced:
The computer
Ethernet Cable
Beagle Bone

[ShawnM]

Am I the only one overlooking the fact the the Gecko STEPPER drives are hooked to the SERVO headers on the board and not the DB25? Per the Centroid wiring schematic for the Gecko 201 they are wired to the DB25.