What's new
What's new

backlash, BAD!!

viper

Titanium
Joined
May 18, 2007
Location
nowhereville
I just started checking backlash. Holy sh$%. .0035in Z and .0015 in Y. X looks pretty good. I called Haas and got the rookie tech of the year. Can anyone help me on parameter changes and if this should even be attempted to comp out? I am going to put an indicator on the screws and see if they are moving by chance. This did not make my day for sure!
 
check your thrust bearings first, blindly adjusting backlash params is stupid. if the screw is bad replace it or repair it, if its the thrust bearings you won't need to adjust the comps after the repair. If you don't know how to check the bearings ask or get professional help.
 
I would have to agree with WILLEO6709 parameters are not gonna be the fix. I would first check the motor to ballscrew coupler, those things pop pretty easy and when popped cause alot of backlash, if that's not the problem check motor housing bearing pack, that can cause backlash as well. last resort would be ballscrew nut lost it's preload.
 
Yeah, I pretty much knew not to try and comp this out but thought I would ask anyway. GD. I just talked to Haas but was a little unclear as how to inspect the screw couplers. I did have to take the Z axis motor off to install the machine so curious if I did something wrong there. Will the coupler say "sprung, buy another" on it somewhere??
 
the Z axis is probably the easist to pop that I have found, the best way to keep them from poping when installing is to put a "pac man" looking shim or material or horseshoe looking thing in the middle of the coupler (kinda couvering 3/4 of the inside of the coupler)so it doesnt compress when you shove the motor on, it needs to be a pac man shape so you can get it out after install. To check if the coupler is popped remove it from the motor and ballscrew and lightly pull and push it. if it has a little step or jump when you do that it's toast, or look at the spring pack in the middle inside if the spring pack has any bows or doesnt look flat it's probably blown
 
lost motion can be in:
coupling motor to ballscrew (more so if the motor encoder is the only one in use)
thrust bearings allowing the whole ballscrew to slide axially ( indicate the end of the screw from a fixed point using a mag base)
ballscrew lash- (put the tip of the indicator in the side of the groove on the screw( balls ride on the sides, not the roots), move the table back and forth switching directions rapidly, if the dial moves more than 0.001 its toast) - in which case the only prudent course of action is screw rebuild/ replace.

I don't own a Haas, I don't know what kind of couplings they use. I know depending upon hours and bearing design thrust bearings are very likely. Some VMC manufacturers recomend replacement every 10k hours of machining. Couplings should not " go bad" without a crash but they can be mechanical and mechanical things sometimes break. Some machines use a simplistic soft steel tube clamped to motor and ballscrew as a coupler, but be aware some allow for misalignment of motor to screw so don't assume you can put a shaft collar on it and go.
 
Honestly, I am still confused what "popped" is referring to on the coupler. I thoroughly inspected it and checked all bolts and pinch collars and everything is tight. I did remove the Z axis motor to position the machine but it was a very simple deal. One keyway with a pinch bolt securing them together. I did not have to take the coupler all apart. Where the coupler from the screw and motor come together, there is a shim pack with three bolts there. All the shims looked fine and not all bulged up or something like that. I heard that might be an indicator.

What I did to get some of the lash out is tighten the mounts that secure the ball screw nut to the machine. They were a bit loose on the Z and Y. The Y is dead on now and the Z still have about .002 in backlash. Because if the age of the machine, we are just trying to get something going for now. We have tried adjusting the Z parameter that says "lead screw" but that actually throws things off when moving in increments. IE, we keep moving in .010 increments and it just gets further and further off in the same direction. Anyone know how to set the comps in these? Are we waisting our time with this much lash?
 
you need to determine the source before you go further.... you are adjusting blind. check the thrust bearings first, they are most likely. check the ballnut lash next, on the z you might have to rig a lever to lift on the head, I have used an 8 foot 4 x 4 with a fulcrum off the table, have a helper read the indicator or run the lever. keep moving the indicator until you either find you movement or confirm that its mechanically sound. Then you can adjust backlash. I'd go back to original parameters on what you changed, and call tech support to fully understand the backlash parameters before you go further. This is one thing you want to take the time to do right, a " quick temporary fix" will only cost your part quality.
 
OK, I think what you are talking about is forcing movement in the table and measuring it with an indicator? We did some of this. We pulled the Y all the way + and felt for slop +/-, you could not feel it, if we moved to the center of the table, you could actually hear it move. We put the indicator on the screw itself and only got .0003. If we moved to the table, we got .003 or so.

I also did this in the Z but with no special tools. I was getting 2-3thou on it as well. I have yet to check the end of the Z screw but will. I am pretty much with the mind that the ballscrews are loose. My next question is, please don't laugh,an we add some oversized balls to at least get going for now?? I realize that is not the best fix but we do not have 6K to throw at this machine right now.
 
if the screws are tight on the ends and loose in the middle good luck reballing it. It won't be compensatable with backlash params, because the wear will change. If your tolerances allow and you are not having finish issues run it, you can't hurt them much more. However, if you can get a 2 week window, I have good success with PGM products for ballscrew repair in warren michigan. Usually cheaper than the OEM and can work off your screws.
 
We are trying to find the official way to check backlash in a ballscrew but it spins free as a bird. I am told they should have a bit of tension like a pair of AC bearings. I am thinking we would like to at least try to replace the balls in one nut and see what happens. I guess they need reground anyway so as long as I do not get binding, "maybe" it will work. I guess if I cannot get the nut back together, the assy needs to come out anyway. I honestly just need to cut something with this machine to see what we have. I hate to throw 5K at it only to find out it needs spindle bearings, rails, and a transmission.
 
if the ballnut " gravity spins " the whole way its thrashed. It should have tension, about 5 lbs or so ona fish scale when turning slowly in a lathe IIRC. to check ballnut lash, fix a magbase on the sliding component to the " side of the groove of the screw. Theoretically the lead of the screw should be constant poisition, move the slide back and forth and/ or pry to overcome gravity while observing the indicator. To check the spindle insert a long tool/ test bar, 6" is my recomended min tool length. Psh by hand on the spindle, observing the indicator. This can be spindle bearings, drawbar springs, but if it moves noticably she may have an issue. If the rails are " pitted " where the balls ride they are thrashed, but some texture is liveable depending on tolerances. I don't have any transmission machines.... If you want to see what she does put her under power and cut it as is. From your first post I assumed it was a machine you'd had a while.
 
No, we just bought the old girl I am honestly trying to work out bugs before we put it to work since I spent days cleaning it. Much easier to work on it right now. What we did no was mount the mag base and indicator on the spindle head and touched off the table in Y and pushed and pulled table. .0025 or so. We did this in X with maybe .0002. We tried the same test with the indicator on the f the Y leadscrew and got .0002 or so.

The only other thing I am wondering is rails?? The Y looks very nice, shiny with no pits or bad looking areas. I have not taken this apart but curious how the X and Y rails meet and maybe we could have a gib problem? How does one check that stuff. I guess the weak motions of just moving the table .001 or .01 at a time should not cause the rails to slop should they?

I will try to find a better way to test the Y to ensure a nut problem but there is no way it takes 5lb of force to turn it. Even at the ends of the screw. I even tried to mount my mag base on the nut and touch off the screw and it does not work because there is radial play in the nut.
 








 
Back
Top