Need Help!- VMC-OMA circle mill

[monaro mike]

Hi to everyone
Having trouble milling 16mm dia. x 12mm long spigot in some aluminium components i am making, as in i cant get this dia. to the roundness tolerance required. .02mm t.i.r. should be achievable and ive achieved much better than this before, I am using 12mm solid carbide endmill 1200rpm 200mm per min feed, climb milling, have tried straight and radial approach and exit - no difference.
Parts are .08mm out of round , set up is super rigid, 6061-t6 alum. no prob there, m/c slideways are like new and gibs are all adjusted spot on and as i have known m/c since new -its never had a big prang and never done hard work.
Drive belts are in ex cond and tensioned correctly, ball screws look like new and are quiet, backlash checked with new dial indicator = .02mm max x and y axis.
Do i need to alter backlash parameters ? i have found them in fanuc manual and there are lots of them for each axis and i dont want to guess which to change.
Any ideas would be appreciated thanks

[barbter]

Find a tech + set the machine using a ball-bar test.
This is done dynamically, unlike setting backlash with a clock.
We did it on all of our machine and it's the best morning we have spent!

Failing that, i'd use a boring head and reverse the tool and 'turn' the od.
HTH

[monaro mike]

Never heard of a ball -bar test but i,m interested in hearing more, we have a couple of companies in australia that set machines with lasers, but they charge like wounded bulls. In reality theyre only setting the slideway gibs to the minimum clearance allowable, i reckon given that ive reconditioned many machines before that i can set the gibs pretty acurately, or are these tech people doing something in addition to this apart from checking squareness parallelism etc.
Thanks for reply

[Superman]

Hi Mike,

Having trouble milling 16mm dia. x 12mm long spigot in some aluminium components i am making, as in i cant get this dia. to the roundness tolerance required. .02mm t.i.r. should be achievable and ive achieved much better than this before, I am using 12mm solid carbide endmill 1200rpm 200mm per min feed, climb milling, have tried straight and radial approach and exit - no difference.

Why so slow ? 45m/min in Al should be up round the 200+ area , we use max rpm of the m/c S6000
What CBD cutters ? designed for Al - extra sharp, hi-helix and all that jazz ?

Parts are .08mm out of round , set up is super rigid, 6061-t6 alum. no prob there, m/c slideways are like new and gibs are all adjusted spot on and as i have known m/c since new -its never had a big prang and never done hard work.

Is the machine giving the same out of round bores / bosses on previous jobs, if not then it should be something in this part giving the problem
quite often this may be the way the part is done, if say 2nd Op and thin walls and floors exist then the possiblitly of flexing occurs.
Try ramping the OD of the spigot, only having a small amount of contact with each flute

ie

S6000 M3
G43 H1 Z20.
M8
G1 Z0. F2000.
G41 X14. Y2. D1 F550
Y0.
G2 X14. Y0. R-14. Z-2.
X14. Y0. R-14. Z-4.
X14. Y0. R-14. Z-6.
X14. Y0. R-14. Z-8.
X14. Y0. R-14. Z-10.
X14. Y0. R-14. Z-12.
X14. Y0. R-14. ( last is repeated to give flat floor )
G1 Y-2
G40
G0 Z3.

you know what I mean.

If you cannot interpolate a true diameter, then it should be considered a machine problem needing attention ( backlash, leadscrew, nut, bearings and thrust bearings )
90% of the time it is a programming strategy that needs refining

[monaro mike]

Thanks superman
Im ramping with roughing cut using 50dia coromill at 300m per min no problem and leaving only .5mm to finish with solid carbide endmill , i slowed it down to hopefully improve accuracy i,ll try ramping finishing cut - and see what happens cheers

[cncmts]

Hi, How did you check back lass with the pulse handle or did you write a program to feed away from then back to zero.

[barbter]

Ball bar testing shows everything.
Backlash, gib, yaw, squareness etc etc. It truly is worth doing.
It's not just the backlash you can adjust, but also the servo gain acc/dec etc to remove any spikes.
If you get north south east west lines on diameters (od or bores), ball bar tuning will also get rid of this.
Have a look on Practical Machinist web site and do a search.

[cncmts]

I agree the ball bar test is the best to adjust a mill. We have had our QC10 system for about 2 years. This is the best way to check a machine its even lead use to bad turcite.

[monaro mike]

Thanks for the reply
I,m getting east/west north/south lines on the outside dia. and i checked backlash by using pulse handle i will try it with a program to see if there is any difference

[Dan Fritz]

I'm guessing that your 0M-A has the old 3-phase SCR drives or PWM drives with analog inputs. Have a look at the servo boards and see if you can find some little blue pots labeled "RV1" and "RV2". These would be used to adjust the velocity loop gain (RV1) and the "drift" (RV2) of each servo. You should also see some check-pins labeled "CH1", "CH2" and "CH3". CH3 is a ground pin, so all voltages on the other pins can be measured relative to CH3.

If these are analog servos, you may just have a slight difference in the velocity-loop gain between the X and the Y axis, resulting in a difference in the servo following error. A mis-adjustment in following error will show up as an "oval" circle. A circle that has "steps" at the 0, 90, 180, and 270 degree points are usually due to backlash or improperly set backlash compensation.

Try this test: With the machine at rest (servos on and ready), use a good digital DC volt-meter to measure the voltage between CH1 and CH3 on the X and Y servo boards. Adjust RV2 until you see exactly ZERO volts between those two check-pins on both boards. Now, program the control to make a 45 degree move in the +X and +Y direction at approximately the feedrate that you're using to cut your parts. Move a few inches in +X and +Y, then move back in -X and -Y. Make this program "loop" with an M99 so that the control just moves back and forth. While it's moving in +X and +Y, look at the voltage on the X servo board between CH1 and CH3. Compare this voltage to what you see on the Y servo board between CH1 and CH3. They should be approximately the same for both the + and the - direction moves. If they are not the same, try to adjust RV1 on either board until they are as close to the same as possible. Then, try to cut your circles again. These pots should not be turned very much, so if you find that you have to turn them more than 10-20% to make your adjustments, you may want to have a more thorough adjustment done of the servos by a good technician. There are a lot of other things on an SCR servo that can make them run a bit funky, and to make them "purr", you've got to adjust a bunch of things. These RV1 and RV2 pots are just "The basics"

There are also parameters for servo gain, drift, and backlash compensation, but I'm assuming that none of these have been changed much since the machine was new. If you've been adjusting the backlash comp parameters, you may want to zero those out and adjust them to this procedure:

With a linear dial-indicator on the X axis, make a short program that moves the X in the plus direction until the indicator stops at about mid-travel, then move about .100 further, then back to zero in the minus direction, then past zero about .100 in the minus direction, then back to zero again. Run the program in single-block, and set the dial indicator to read zero at the mid-stroke point. The indicator should come to zero when approching that mid-point from BOTH directions. If it does not stop at zero from both directions, adjust the backlash compensation parameters for the X axis so that it does. Remember that the backlash may be different in the middle of the axis travel than it is closer to the limits of the axis travel (due to ballscrew wear), so it's best to adjust it in the range where you'll be doing most of your work. Do not "over adjust" for backlash. If you see a lot of backlash in either axis (with the comp parameters set to zero), then you have a mechanical problem that must be addressed. The most common problem is a loose ballscrew thrust bearing or poor way lubrication that causes torque "wind-up" on the screw. Repeat this procedure for the Yaxis and you should be all set for backlash comp.

Of course, if you don't have analog SCR drives on your machine (as Rosanne Rosana-dana would say) NEVER MIND!

[gbowne1]

In both my 0M and 6M, I had several people look over my Servo tuning, backlash, etc. I am so glad that I did.

Once I get my 0T/6T control built and when my my new lathe gets here I will install the controls and do the servo tuning to it too.

I really suggest that people do the ball/bar tests, servo tuning, etc. I really worked wonders on my machines. I re-test every 3 and 6 months as a part of maintenance.

Greg

[monaro mike]

Thanks for replys
Just cant get over the wealth of knowledge all you people have ! some of these guys must sleep with their machines, and special thanks to Dan the guru of fanuc, of the next few days in between making lots of parts i hope to
try what Dan is recommending and will let you know what i find out Cheers everyone