How to test backlash/gibs

[jderou]

I get alot of table movement on my boss whenever I try to plunge. Most recently it is with a 3/8 center cutting endmill, even when I slow it way down. Would this be caused by backlash in ballscrews and/or worn/improperly adjusted gibs? How can I determine if I have these problems and, most important, how do I fix them?

[rustyolddo]

Yes & Yes.
I've only adjusted manual Bridgeports, not familiar with a BOSS. You can use a DTI set up in the spindle or a centering indicator to determine where your movement is.

In a manual mill, you'll see it in both the gibs & the leadscrew. Manual mills have adjustments for for both. CNC machines are far more critical. In addition, in a manual mill anyway, you'll find wear in the center of the lead screws & ways and little at either limits. This poses a problem, when you adjust for minimal backlash where the machine normally operates, the gibs & leadscrew can lock up if you go to either limit. In a manual mill you adjust for best compromise and compensate by locking the gibs when plunging and running some tension on the gib locks when milling. Can't do that with CNC and you get the "shimmy shift" when plunging and you can some problems climb milling too. You can work around it to some extent but it takes rethinking your tool paths. I'm no expert by any means, just have some personal experience adjusting machines.

Again, not familair with a BOSS, but it may require a trip to a rebuilder to be reground, scraped etc.

[One of Many]

I would use an indicator and some feeler gages.

Set up a good responsive indicator held in the head with the business end against the face of something solid on the table for the axis you intend to inspect. Make .005-.010 moves in both directions and check for repetition with the control and see if the indicator reads direct to those movements. Any difference in indicated movement verses commanded movement at various places along the screw will show backlash. If the screw has a compensating nut, it may be adjustable, but for the same reasons Rusty points out as points of major wear. The ball screw will not like going into areas where there is no wear.

Then there is another option and that is if this control has a ball screw backlash compensation parameter where you could have the control take care of the backlash to some extent.

One other area that gets overlooked when excessive backlash is aparent is if the bearing support blocks have ever been removed, crashed etc. If the ballscrew is allowed to shift in the bearings or the compensating nut is allowed to shift, the problem may not be in the screw backlash itself. If the slop is coming from somewhere, it will be up to you to expend the effort to find it.

Checking for excess gib clearance can also be done with an indicator against a T slot. I prefer to check it as far out as I can go, but the outer edges of the saddle is as good as it will ever get anyway. Feeler gages can quick inspect the non-gibbed dovetail for slop. Grab the table, push and pull on the ends. Do this with the table at either extreme and in the mid-section. Detailed attention will give a good indication of the wear envelop.

Worst case scenario, when little to know correction can be made. If you need tighter toleranced parts, I would suggest offsetting your vice to run the table in a non-wear area. That can be rather limiting on a Boss table to begin with.

DC

[timlkallam]

http://www.kentechinc.com/tip2.html

[machintek]

The stepping motor versions of the BOSS series (3, 4, 5, 6) do NOT have any provision for backlash comp. The servo motor versions do (BOSS 8 and 9).

Gib adjustment is not any different that a standard Bridgeport. You must be careful not to make them too tight as any drag on the axis will cause a stall ( the stepper will whine but not go anywhere).

George

[One of Many]

( the stepper will whine but not go anywhere).

I've heard that called BP stepper motor constipation!

It takes a loosening re-laxative to cure it! heheh!

DC

[HuFlungDung]

A 3 or 4 flute center cutting endmill does not exhibit good chip clearing ability when plunging. They are better at ramping down into the cut, on a boundary ramp or a helical ramp.

Even two flute endmills are "iffy" when vertical plunging. If you look at a two flute endmill, you will see that the most common grind for the center cutting edge, creates a perpendicular face that does not mimic the helix of the spiral flutes one bit. This can and does plug up, and causes much shuddering and shaking.

I like to carefully grind (by hand) a bit of a back rake into that perpendicular face. Be careful to not actually grind the cutting lip back, just grind up to it and stop. Preferred wheel would be borazon cup for HSS tools, or natural diamond for carbide, because these wheels will hold their shape and you need quite a fine corner on the wheel to grind right to the center of the endmill.

Avoid the dust when grinding carbide and HSS.

[jderou]

I don't think im ready to start grinding tools just yet I do need to work on my plunging techniques, though. How do you plunge? Just ramp straight down into workpiece? How much of an angle does it take?

Well I measure backlash to be .006 in Y and .003 in X. I took a 4' bar, put it in the vise and tried to rock it back and forth but it didn't go anywhere. I'm thinking this means my gibs are ok.
So now how in the heck to I find replacement ballscrews? Or do I have these reground?
Joe

[HuFlungDung]

Well, I used to plunge straight down, like in drilling, but since switching to OneCNC cadcam software, it makes it quite easy now to ramp down on a boundary line of the part, or, if space is tight, a zigzag ramp, or a helical ramp (with small orbit radius). I use a ramp angle of 2 to 3 degrees.

If you don't have the software to make it easy to program these ramps, then maybe you can get lucky using a helical milling cycle out of your cnc controller. Or simply drill an entry hole for the vertical plunge.

The backlash figures you described are quite large. I would run the table to one end, and then check the gibs. If you can tighten them at the end of travel, then that is about all you do, and the center of travel will be looser.

The movable knee likely does add a degree of looseness that will always contribute to the shake and shudder. But more often than not, it will be the spindle itself that is swimming around the hole when plunging with a ordinary, unmodified two flute endmill.

[machintek]

A regrind of the ball screws is a lot cheaper than new.
BUT! Make sure your mechanical lost motion is not due to bad bearings, loose belt, bearing nuts loose, keys worn, etc.
As stated above, that is a lot of backlash.
Put a dial indicator on the end of the screw, if it moves in and out with a reverse of direction of motion, you need to replace bearings.

George

[jderou]

Something I found odd, in the Y direction, when I step it .001 I hear the motor step, but watching the dial indicator the table motion is 'soft'. The table takes longer to move than the stepper takes to step, kind of like a delayed reaction. Does that make sense?
I'm going to pull it apart today and look into the bearings/belts etc.

[One of Many]

Originally Posted by jderou Something I found odd, in the Y direction, when I step it .001 I hear the motor step, but watching the dial indicator the table motion is 'soft'. The table takes longer to move than the stepper takes to step, kind of like a delayed reaction. Does that make sense? I'm going to pull it apart today and look into the bearings/belts etc.

Sure that makes sense. It is within the float region of the mechanics within the axis. Backing off of the forces that moved the table in one direction will very well relax at a slower rate than if it were a zero slop connection. The rest of the float will need to be traveled +.001 before you see further perceptible movement in new direction. As has been stated, that float region can come from many contributing factors, not just ball screw backlash. The proverbial finding the needle in the hay stack where there may be a small accumulation from many parts. Divide and conquer!

DC