Fadal Circularity Issues

Alright, so I've started a new job, and at the moment I am mostly tasked with running proven programs on a Fadal VMC2216. Little BT40 box-way machine (so cute after those LeBlond Makino CAT50's!). However, as usual, it's got some eggy holes. Up to about .004" out of round at 90° (45° from the quadrants) - larger across 10 o'clock to 4 o'clock than 2 o'clock to 8 o'clock. I would like to improve this, as interpolated holes are usually going no faster than 20IPM, and I know it can get a bit better with proper comps, but I don't know where to set these or how to test them best.

What can a triangle-circle-square test tell me and how do I adjust after measuring? Is this a half-decent test at all?

I want to be a better machinist and troubleshooter, and consequently be able to get better results. I know there are a few Fadal gurus on here (bobw I'm looking in your direction), and I've got such a soft spot as a VMC15 was my first 3-axis mill experience.

Atom

There are no fadal gurus, just a buch of guys that got tired of calling a repair tech every other week and ended up learning by themselves how to keep their fadals in decent running condition

In Fadal land there are several known common causes for why your circles are not round: lack of servo comp, loose gibs, loose servo couplings, worn screws, bad trust bearings, innadecuate backslash settings. Just to name a few. You have to check each cause to be sure you are doing it right.

I would recommend to search in he OTHER forum: C*****E. Over there is a much larger fadal community than here in PM and eggy circles is a popular topic.

Cheers.

Atomkinder said:

Alright, so I've started a new job, and at the moment I am mostly tasked with running proven programs on a Fadal VMC2216. Little BT40 box-way machine (so cute after those LeBlond Makino CAT50's!). However, as usual, it's got some eggy holes. Up to about .004" out of round at 90° (45° from the quadrants) - larger across 10 o'clock to 4 o'clock than 2 o'clock to 8 o'clock. I would like to improve this, as interpolated holes are usually going no faster than 20IPM, and I know it can get a bit better with proper comps, but I don't know where to set these or how to test them best.

What can a triangle-circle-square test tell me and how do I adjust after measuring? Is this a half-decent test at all?

I want to be a better machinist and troubleshooter, and consequently be able to get better results. I know there are a few Fadal gurus on here (bobw I'm looking in your direction), and I've got such a soft spot as a VMC15 was my first 3-axis mill experience.

Click to expand...

.
some cnc machines you use a dial indicator and a gage block and move 4.0000" and see what dro says and often their is a compensation or calibration mode so you tell it it moved 4.0000 and resets scale. you then repeat test to confirm. this is done in x,y,z
.... other cnc machines using dial indicator you put in G91 G1 X0.100 F10. then G91 G1 X-0.1 F10. and see if indicator moved 0.100 both ways and if not there is a backlash setting that is changed. test done again and see if better. with servo motors (not steppers) there is usually some sort of feedback of actual position back to cnc
..... any special modes or parameter changes i would read the manual first and record the before and after settings.
.........machine squareness i check with a granite square and a dial indicator. sometimes if not too bad there might be a setting to electronically compensate for a out of square machine. again i would refer back to the machine manuals. messing with parameters can be a good way of messing up a machine.
........ sometimes there is a special gcode used and machine will follow path more precisely but might pause in spots as it confirms position. this can slow machine down so often is not on by default

If you are out .004 the first place to look is the axis thrust bearings and then the box way clearances. The nice thing about box way Fadal's is that you can adjust the tapered gibs to factory specs with a 1/4 inch socket. The accuracy and rigidity of the machine is lost without properly adjusted gibs, we check each year as part of our PM service. Check your manual for the procedure.

servo

in2glamisgirl said:

If you are out .004 the first place to look is the axis thrust bearings and then the box way clearances. The nice thing about box way Fadal's is that you can adjust the tapered gibs to factory specs with a 1/4 inch socket. The accuracy and rigidity of the machine is lost without properly adjusted gibs, we check each year as part of our PM service. Check your manual for the procedure.

Click to expand...

.
yes i have seen the ball screw bearing need replacement and we first noticed on a servo screen that showed servos going back and forth continuously when they should have not been moving.
.
by putting a dial indicator from spindle to part you could easily see 0.0010" movement back and forth every few seconds because of a bad ball screw bearing and servo going back and forth all the time

Sounds like servo mis match. You could verify by cutting your circle in the opposite direction and your "egg" should flip flop.

This is corrected by adjusting servo gain. Leave one axis alone and adjust the gain pot (if this is the type of servo you have) until you get a relatively round circle.

In Fadal land there are several known common causes for why your circles are not round: lack of servo comp, loose gibs, loose servo couplings, worn screws, bad trust bearings, innadecuate backslash settings. Just to name a few. You have to check each cause to be sure you are doing it right.

Click to expand...

For the record, this is not limited to "Fadal land". It's any machine.

0.004" does seem like a lot (what size hole and endmill?) but I also lean towards a servo gain adjustment as PMT has suggested. Our Fadals have never cut gnat's ass holes but our maintenance sucks too. As a standard practice on tighter tolerance holes, I flip my approach 180 degrees and mill bore the hole again and it tends to come out within 0.0004 - 0.0006" round. If it does fix the problem it is almost certainly a gain adjustment.

The Servo Gain Loop program is embedded into the control and can be retrieved at anytime. I forget how at the moment but it should be in the manual somewhere. If in2glamisgirl is correct however, the gain adjustment is not going to help. I'd check both if possible as they are both pretty easy to do.

PMT said:

Sounds like servo mis match. You could verify by cutting your circle in the opposite direction and your "egg" should flip flop.

This is corrected by adjusting servo gain. Leave one axis alone and adjust the gain pot (if this is the type of servo you have) until you get a relatively round circle.

Click to expand...

Also, with the machine off, you can reach under the chip guards and hand turn the screws. If they turn smoothly with little resistance then your not having a serious problem with lubrication or a severe turcite wear problem. Ball Screw to servo coupling looseness should be checked also.

RTM said:

0.004" does seem like a lot (what size hole and endmill?) but I also lean towards a servo gain adjustment as PMT has suggested. Our Fadals have never cut gnat's ass holes but our maintenance sucks too. As a standard practice on tighter tolerance holes, I flip my approach 180 degrees and mill bore the hole again and it tends to come out within 0.0004 - 0.0006" round. If it does fix the problem it is almost certainly a gain adjustment.

The Servo Gain Loop program is embedded into the control and can be retrieved at anytime. I forget how at the moment but it should be in the manual somewhere. If in2glamisgirl is correct however, the gain adjustment is not going to help. I'd check both if possible as they are both pretty easy to do.

Click to expand...

Today the hole was 2.875" for a matching part. Eggshape was about .0035" the same directions, finished with a 1/2" AlTiN carbide in a helix bore path at 20IPM. Mating part is being welded in, but it still had to go in first! I'd be plenty happy with roundness at .001". Again, these are proven programs, so not a lot I can do to edit except what I can do at the control. I will say that the guy who used to do the programming had some real quirks though.

Talked with the shop manager today too, and he told me that the turcite has come loose twice on the gibs of this machine, last time was about eight months ago and it was repaired, so it shouldn't be the gibs. The X-axis ball screw was replaced within a couple years ago, but the Y is coming up. I will have a good hard look at the manual and run some tests on Saturday when I can get back in the shop.

I like this job, they like me, and I want not only to make better parts than they are used to, but to prove that hiring me was a worthwhile investment.

I had a fairly bad "out of round" problem on a Fadal 40x20. It turned out to be the leveling (or lack there of) of the machine.

If somebody can get a dead nuts interpolated hole out of a Fadal, I'd surely like to know the trick... If its critical, I just bore it.

I don't much anything about servo tuning.. But the instructions are in the manual.. And I'm sure that has something to do with it.

As for just general machine condition... Gibs and straps and all that... The instructions are in the manual.. And they are in english
and they are actually written pretty well...

Lost motion.... Checking it is pretty simple. Indicator on the end of your ball screw to start.... If that's good, then
into the back lash setting.... FOLLOW THE DIRECTIONS!!!! Seriously... FOLLOW THE DIRECTIONS in the manual...

Don't just move the MPG one way and then the other... FOLLOW THE DIRCTIONS... I've found its easier to put a pin
in the spindle and put the indicator on the table... but I FOLLOW THE DIRECTIONS!!! And then when you have
all of the backlash parameters set for left,right,middle... Go back and do it all again since left right and middle effect
each other... You'll probably find that you will need to decrease the numbers in the backlash table a little on your second
go round.

Then just to make life difficult... Do it with different feed rates and see what happens. I found a loose motor coupling that way
once. At low feed rates, it was fine on direction reversal... At higher feedrates, on reversal, the coupling slipped a bit... The way
the couplings are attached to the screws and the motor is rather pathetic, one tiny little key and a 1/4" setscrew... Mine now all
have the original key, the original setscrew.. Two extra flats and 2 extra 5/16 set screws, and I had one of those loosen up on me
a while ago.

Well as it turned out the X-axis amp was well on its way out, causing a servo mismatch. The tech that came out swapped it to the Z axis and we ran it at half rapids for a few days until they could get back and replace it. Once replaced and everything adjusted properly, I'm getting .0002-.0004 circles. Certainly good enough for me on a Fadal! Thanks for the help folks.

Tuning procedure solves alot of out of round issues. You should also dial in your backlash because it WILL fight your pitch error comp table. Most folks never run a servo gain program they just bitch about it until they sell the machine. You can make VERY good parts on a Fadal VMC.

Atomkinder said:

Well as it turned out the X-axis amp was well on its way out, causing a servo mismatch. The tech that came out swapped it to the Z axis and we ran it at half rapids for a few days until they could get back and replace it. Once replaced and everything adjusted properly, I'm getting .0002-.0004 circles. Certainly good enough for me on a Fadal! Thanks for the help folks.

Click to expand...

Heck, I don't know if my Fadal was any better than that when it was brand new. If memory serves me correctly, I think that's what I got when I tested mine new.

Today was the tightest tolerance I've seen on any job yet here at ±.001 on a bushing (the bushing is plastic) bore. Held the diameter within about six tenths towards the high end of the tolerance across all eight parts (the bushings were about .002" over nominal). Granted it's not production, but I feel pretty good about it. Bushings pressed in straight and easy. Finish feed was 25ipm with a .5" 4-flute end mill on a 2-3/4" diameter. With G8, almost no visible witnesses on the quadrants (actually I think the only one I noticed was X+Y0).

Atomkinder said:

Today was the tightest tolerance I've seen on any job yet here at ±.001 on a bushing (the bushing is plastic) bore. Held the diameter within about six tenths towards the high end of the tolerance across all eight parts (the bushings were about .002" over nominal). Granted it's not production, but I feel pretty good about it. Bushings pressed in straight and easy. Finish feed was 25ipm with a .5" 4-flute end mill on a 2-3/4" diameter. With G8, almost no visible witnesses on the quadrants (actually I think the only one I noticed was X+Y0).

Click to expand...

If I were you, I'd be feeling pretty good about that too. Sounds like you've got a nice machine there. I love my Fadal. As slow as it is, it's just like a timex.......