[chopsmitty]

I have a NEW, never cut a chip, 2000 Cincinnati Arrow 500 with glass scales and Vickers Drives. The machine was installed in a clean room and never run. Pulled out a few months ago and brought to my shop in Hawaii. I've got it installed, leveled and sent the boards to Fives for batteries and reflash of bios. Fives also supplied a new SSD.

One of the machine parameters was set incorrectly, so it wouldn't drill to commanded depth (.100" shy). Got that figured out. Now it's working great! It's sounds like new - heck it is brand new!

My new problem is:
I've programmed (Mastercam) a contour cut rectangle with a 1/4" end mill. The part is about 1.5" x 9" and the depth of cut is about .45" Radial engagement is very light - clean up cut - maybe .015". About half way through one of the long sides I see (and can measure) a step of about .0015" It doesn't do it every time (I've got several to cut), but about one in five parts or so...

As this is essentially a new machine and may have not been set up by a trained technician, I'm wondering if there is some parameter that drives accuracy that works with the glass scales to maintain position? Anyone with one of these (glass scales) up and running, can shoot me a screen shot of parameters?

https://www.youtube.com/watch?v=7CRVgdva6wc&t=833s

[gwood]

Chopsmitty, The Kollmorgen - ServoStar Drives - Digital-No-Pots (Vickers) are very reliable. Instead of potentiometer resistors they used parameters that can be verified; or downloaded/uploaded from the control.
Hoping that Bill at Fives recovered the BCK file off the old drive for you, so that you can safely archive it.
In the scale feedback system the control closes the "Positioning Loop" and the Drives control the "Velocity Loop". If there is any mechanical backlash in the system; then it will cause instability which could cause
your bump. If this is the machine with the rust in the video, I would pull off all the covers and examine the THK linear ways for rust and the thrust-bearing housings for water contamination. These may be the Series-2 ways that need to be lubricated every 2000 Hours
with P-64 NLGI#2 (Lithium base) grease. These had no lube lines, only Zerk fittings. Be careful not to pump too hard or you'll break the THK QZ-Lubricator End Cap off and need new ways!
I've attached a procedure to check for mechanical backlash.

[chopsmitty]

Thanks for the insight. I have the original hard drive that still works perfectly. We have cloned it to a SSD. I also have the SSD as supplied by Fives. So, yes, we should have everything covered.

Although some of the "top" surfaces showed signs of water intrusion due to poor packaging and storage everything underneath was in perfect condition. I believe it was just rained on (Hilo, Hawaii) for some time. The ways were all perfect and from what I could see the thrust bearings looked new as well.

My quick way to check the machine periodically for backlash is to use a Haimer digital taster up against a vise or rigid fixture and rock it back and forth in the X and Y. This machine shows .0002" in both X and Y. My old machine showed .00025" (Y) and .0003" (X) using this method. I was thinking the glass scales would drive the machine to appear to exhibit "no backlash"? Where as the location of the encoders on the old machine prevented it from overcoming the backlash in the screw and thrust...

[gwood]

Since the Control is closing the loop with the scale you will not see a problem unless you use my methods in the Troubleshooting_Axis_Positions attachment

[chopsmitty]

After much discussion Mike Eckardt and I are thinking it may be a ground issue. We currently don't have the ground isolated - as per the wiring instructions. We will be pounding a ground rod and isolating it soon. I will report back as we tackle this issue.

[gwood]

As long as the machine transformer is ground to your supply ground it should be good. Also there should be no metal conduit or airlines attached to the machine.
These would create ground-loops.
Did you look for backlash on the end of the screw for Y-axis using a ball bearing? It only takes 10 minutes...
If the machine was shipped roughly with no shipping clamps it could have loosened the nut on the ball screw thrust bearing or worse.
This machine also has thermal comp for Y-Axis related to the Air temp sensor and the spindle carrier sensor; but I've never seen them jump as you are seeing.

[meckardt55]

Let us know what you find and what corrects the problem. I have also seen air compressors starting causing this problem. Was the X,Y, and Z axis secured during shipment?

MIKE CNC Machine Services INC | Cincinnati Milacron MachinesCNC Machine Services

[fdarling]

Would you mind posting photos of the glass scales mounted on your Arrow 500? Mine doesn't have any, and I'd like to see how difficult it would be to add them. Thanks!

[gwood]

FDarling,
If available for your machine there were 3-types of scales and scale harnesses. Much money involved for such an old machine....

Please provide:
Machine Serial Number: 70xx-Y0x-xxxx
Model Type (USA machines): i.e. Exx, ERO, ERD, EKK etc... or Year of Manufacture from separate Plate below Serial Plate (for Europe and rest of the world machines)
Control Diagram Number: 9120xxxxAyy

Machine Serial Number:
Model Type (USA machines): or Year of Manufacture:
Control Diagram Number:

[precisionatl]

shopsmitty,
That's an awesome find! I would like to find another low hour machine like yours! How did you adjust the as you referred ((to drill to commanded depth (.100" shy)) I have the same problem on all of my milling machine's. I have just gotten use to it. Something to do with the R value??

Thanks!!