Fadal Backlash ?

[Hifeed]

Hi, I have a problem with my fadal 3016 FX. It has the fanuc oi MC control.
I have adjusted the backlash with a indicator and it repeats within .00015 everytime in feed and rapid feeds. When I mill a square, arcing in and out in the middle of the square, the square is .0005 different in size on either side of the entry point. Also when circular milling it leaves dwell marks on the Y axis only. the same axis where the size difference is. The machine is a year old and I have had trouble from the start. I actualy had to shim the column
to get this thing square and recut and hand scrape the spindle face to tram it. Would appreciate any suggestion PS when milling a hex its off .0005 in the 45 degree direction

[mniadna]

I have a 1994 Fadal 4020. I have the same problem. I am currently running a small production job on the machine through 3 setups. I only run 1 fixture at a time. I set the backlash to the travel extents of the fixture I will be running the part on, then I run the parts. When I move to fixture 2, I reset the backlash to that fixture, and so on... Doing this helps out tremendously with accuracy. It is a pain, though. I asked earlier in this forum if there was a way to update the BL values within the program, but apparently there isn't. I'm sure there will be people replying, telling you to make sure the machine is level, adjust your gibs, check your screws, check your bearings, couplings. I have done all this. I have all new bearings and zero backlash couplings, all around, new Turcite on the saddle and table, new gibs and straps! I believe the problem is with how poorly the machine compensates for BL. Setting 3 values for 40" (in my case) just doesn't seem realistic, and the fact that Z only has one value really blows my mind. I have a brand new NSK screw on my Z, and my Z will only closely repeat around the height I set the BL at.
I would be interested to see firsthand how a machine with scales performs...

[mniadna]

I just noticed you have a fanuc control. I have an 88HS. I'm not sure what I posted would be necessarily accurate for your machine regarding the BL, because I'm not sure how it is adjusted in that control, but I have the exact same problem and it is because of the BL compensation.

[Hifeed]

I may have found the problem. The fanuc control has a setting for accuracy level 1 thru 10. Mine was set at 1 (least accurate) I had set it at level 1 using the g code G05.5 Q1 R1 for a 3d toolpath and apparently it is a modal code so it stayed at level 1. On your machine you may need to tune your servo amps so that one axis does not lag the other. They should all arrive at the same point at the same time. This is what a fadal guy told me. It does not apply to my fanuc control however. As far as glass scales go, another fadal guy said they can cause problems and are not worth it.. He said unplugging the scales usually fixes the problems.

[mniadna]

Yeah, I set the position loop gain, and verified it. I also put new resolvers on X and Y.
Your problem really seems like a BL issue. Isn't the G05.5 for countour accuracy? Like chordal deviation of an arc?

[Hifeed]

I am sorry I meant G05.1 The guy at the factory told me it is for bolth 3D and 2D accuracy. It is a modal command as far as the accuracy level of the servos are concerned. G05.1 (command) Q1 (turns on) R 1 thru 10 (accuracy level, 10 being most accurate. What I dont like about using this is that the servos seem to bang when moving at high feedrates, just doesnt sound healthy. I intend to get the machine laser calibrated when I get some time. Maybe that may shed a little light on the problem