VMC15XT X Axis Positioning Accuracy

Hi All,

Looking for some troubleshooting help on a 1997 VMC15XT. I picked up about a month ago and been putting it to work right away. I had been doing most of my work on a Chiron FZ08S so having the larger travels and greater tool capacity is nice. I've noticed a positioning error on the X axis. Imagine cutting a 12" x 12" square frame from .25" 6061 where the frame is 1" wide all around. If I use the same 1/2" tool with same speeds/feeds/.005" finish pass, I can dial in to get 12.000±.001 on the Y outside and 10.000±.001 on the Y inside. On the X however, I get 12.004 on the outside and 10.004 on the inside. To me, that says that the X axis is either really sloppy or has a positioning error. I am measuring the part before removing the strap clamps from tooling plate so I'm not thinking it's a springing issue. I've checked the backlash values and they are spot on for both X and Y at about .0003". The survey for the X axis is set to what the factory paper in the control cabinet has on it. 0 at the -14 end and 13 at the +15 end. I've rewritten the survey to be 0 at all locations, written to the axis card, and read from the axis card to verify it's 0. I have checked both the X and Y with a 4" gauge block and Mitutoyo .0001 test indicator. The X will consistently read 4.0015" while the Y is 4.0000". I've even mounted my 24" Mitutoyo caliper on the machine, trued it up relative to X, and Z, and get a similar measurement. See the picture below. I understand this isn't a precision measurement by any means, but .006" over 18" lines up very closely to the .00015 over 4" and the .004 over 12". I'm taking all the measurements moving the X in one direction, so no backlash coming into play. I've also taken off the X axis motor side way cover and motor coupler shield, put a mark on the coupler, jogged the axis 12", and checked the mark. If the control was somehow mispositioning the X axis by .004" over 12", that would show up as a 7.2° error on the coupler (.004in/.2rev/in*360deg/rev = 7.2deg) which would be noticeable. If you look at the third picture below, the mark position is after .004" of travel, and easy to see it's not aligned with the pointer. I'm only using the test indicator stylus as a pointer in this case. It doesn't appear to be doing that as the mark lines up perfectly every 0.2" of travel.

So I guess my question is what are normal survey numbers for a 15xt? The Fadal guys I've talked to say 0-13 is pretty common. Based on my measurements (.0015/4", .004/12", and .006/18"), I figure an error of ~.0004" per inch of travel. If I plug that into the X axis survey, it'll start at 0 on the -14 side and end at 97 on the +15 side. Now when I measure things the exact same way as before, the X axis is spot on. If I cut that 12" part, it's now the exact same in X and Y. So what's the deal? One local Fadal tech I spoke with said survey values of 100 aren't even possible and the control won't even use them. That's not what I'm seeing based on the measurements improving.

Is it possible the ballscrew pitch is off by that much? I know rolled ball screws can be in the .004"/ft accuracy range, but all these Fadal people I talk to say it should be max of about .0015" over the full 30", or the max of 15 on the survey, not almost 100 like I'm using. Maybe someone replaced the ballscrew with one of lesser accuracy? I'm all for paying someone come and laser or ball bar the machine, but I'd rather figure out if I should put a new ballscrew in it first.

Anyone have any thoughts or suggestions?

Sounds like what a Bridgeport does with 'knee rock'. Shifting the weight of the table side to side takes the slop out and adds to the effective travel.

I don't know if a Fadal would suffer from the same problem from wear.

The TL;DR version of this is: replace the thrust bearings.

RC Mech said:

The TL;DR version of this is: replace the thrust bearings.

Click to expand...

I took off the motor and coupler box and spun everything and it feels fine. Probably not a bad idea to replace the thrust bearings anyways, but wouldn't a bad thrust bearing cause lost motion?

I would get an indicator and a pry bar and see if you can find any free motion in anything. I've seen loose knee gibs cause a similar issue on a bridgeport cnc.

I did some prying and indicating and didn't see more than .001 of movement. That was both table to Y saddle and then table to spindle.

I got a bearing kit ordered and will see if that fixes the issue like RC Mech said.

On my VMC 15XT the Y axis thrust bearing play was apparent by grasping the T-slot nearest the operator and pushing/pulling on it. Test indicator sitting on the way cover, probe on table front.

Thrust bearing failure is common on these machines. While you’re not necessarily isolating the thrust brgs by pulling the table in X/Y, it’s highly likely they’re the source of slop in a Fadal.

Aside- Max survey comp is 50 IIRC.

RC Mech said:

On my VMC 15XT the Y axis thrust bearing play was apparent by grasping the T-slot nearest the operator and pushing/pulling on it. Test indicator sitting on the way cover, probe on table front.

Thrust bearing failure is common on these machines. While you’re not necessarily isolating the thrust brgs by pulling the table in X/Y, it’s highly likely they’re the source of slop in a Fadal.

Aside- Max survey comp is 50 IIRC.

Click to expand...

I did push and pull along the X and didn't see anything over .0005. I also put indicator on the Y saddle and the probe on the end of the ball screw and jogged the X. The same .0005 backlash there so the screw isn't moving much relative to the Y saddle. I also put indicator on the table and probe on the ballscrew "thread" and don't see anything slop between the ball nut and screw.

But I do hear a really faint squeak in the bearing/motor area so probably good to replace bearings.