X Axis shifting during program

[SBC Cycle]

I've held off on posting this until I could get this resolved with my machine dealer and I'm still working with them. But I would like to know if anyone has ever had an issue with the X axis losing its place.

I have ruined dozens of parts because of this problem. Inexplicably, the machine X zero shifts (EDIT: To clarify, my offsets aren't changing. It is the Cold Start X position), sometimes in the middle of a program and sometimes between parts. If you send the machine home for power off, you can actually see that the witness alignment mark is off on the X axis.

I was standing by the machine once when it happened, I was in the middle of a cut and I heard a "clunk" so I stopped everything thinking I broke a tool. Tool was fine and I didn't notice anything amiss so I hit start and walked away. When I came back to load a new piece, I noticed I had a mismatch on my part (3D surfacing). Sure enough, I sent the machine home and the witness marks were off again.

I have had this problem since the machine was installed last Feb. (2008) Fadal has been working very hard to correct the problem with CPU replacements and various other things but since they came out early Oct. (2009) with another replacement card, it has happened again. I am at wits end now and I'm not sure what to do.

Again, Fadal has been EXCELLENT with their support and so far it hasn't cost me a dime in machine repairs but I'm getting killed with scrap parts. Some of the blanks I'm using are over 300 a piece. I need to find a solution quickly. I'm at the point now where I have to run this machine myself and stand right by the button if I hear the telltale "clunk".

If you have had this experience, please post here or shoot me a PM. There are further details that I could post to try to track down the problem but I would like to know if this is a common problem. Fadal says they have never heard of this before, I just want to keep them honest.

[HuFlungDung]

How many minutes or hours is it between glitches? Would it make sense to swap motor cables or swap drives, then run it and check for the error moving from one axis to another?

What is the exact amount of the error? Is it repeatable?

[Techman]

Could it be possible that it is a mechanical problem? I think I would be tempted to check the coupling between the servo and the screw and also the connection between the resolver and the servo. Also you might want to check the mounting of the bearing block that carries the ball screw bearing. Hope this helps. There is nothing worse than an intermitint problem that scraps parts!
Jim

[scadvice]

Its a new machine? If it was me I would request a conference call to be set up with your dealer, Fadal, and yourself. Tell them in writing certified mail to do this. During this meeting tell them that you feel that they have had plenty of time to solve the problem and have not. Tell them you are willing to set a date in the near future to have the problem completely solved and if it is not, then you want them to replace the machine. Leave it left unsaid that you will do legal action. They can read between the lines. But be sure to do everything by e-mail that is responded to... or certified mail. Keep records of the dates, times and who did the work.
I have gone through this sort of thing once, not with Fadal but another machine maker and learned. )I happen to like Fadals also... but......) It was a near thing for going to court because I didn't keep good enough records, however, they finally settled and took it back(with the help a good lawyer writing a few letters). By the time they took the machine back, more than a year had passed. BTW-It was a BP interact machining center.
Your case should be easier, assuming you will take a replacement machine. I suspect they will get more active in fixing it, if you start getting firm about telling them their time is running out. If they are not listening to you, hour's worth of lawyer time can write a nice letter. Fadal wants the problem fixed and sometimes it's just getting past the dealer and letting them know there is a serious problem afoot.
Steve

[holbieone]

i'd check your drive card for that axis , almost sounds like it's running off

is it in the same direction every time

[SBC Cycle]

Ok, just to answer a few quick questions:

The shift is repeatable and it is precise. The last 3 times it has happened (twice since the latest service call) it has been off exactly 0.0495" The interesting thing is that the operator seems to think that before the latest service call, it was shifting in the negative (-) direction. Now it is in the positive (+) direction. I have no way to verify this but maybe that means something.

There doesn't seem to be a time or number of auto cycles that this happens. Sometimes I will fire the machine in the morning and go straight to checking the offsets and find that they are 0.05 off right at the get go. What puzzles me is that sometimes it's as simple as realigning the marks and doing a CS, I check my offsets and they are back to where they should be. Other times I have no choice but to change all the X's in my offset table.

I'm leaning toward mechanical but the Fadal tech has never checked any of the things that you guys have mentioned so far to my knowledge. The last thing done to the machine was a new CPU early this month. Apparently the machine was shipped with the wrong one. California waived the $2250 cost of the CPU with no hesitation. Again, I can't say enough about the effort they have gone through to get me fixed up.

[SBC Cycle]

Originally Posted by Techman Could it be possible that it is a mechanical problem? I think I would be tempted to check the coupling between the servo and the screw and also the connection between the resolver and the servo. Also you might want to check the mounting of the bearing block that carries the ball screw bearing. Hope this helps. There is nothing worse than an intermitint problem that scraps parts! Jim

I'm a bit embarrased to say but I have zero knowledge of the internals of the machine. And quite frankly, as long as they are still working with me I don't want to mess around with anything that might hamper them from nailing down the problem. I always call my tech when this happens. The latest suggestion he had was to make sure the RS-232 cable is removed from the machine when it's not in use sending programs back and forth. I have a hard time believing that the cable could be corrupting the CS zero position but who knows.

You know how the intermitant thing goes, it never happens while they are here

[scadvice]

One thing to look at:
Was the machine grounded as they state in the machine setup manual when it was installed? A bad ground can cause real trouble sometimes.
Steve

[Neal]

SBC--
To clarify the RS-232 issue for you, If electrical noise of the correct frequency is received thru the RS-232 port, the machine can THINK it is getting a command and it could react to it causing some strange things to happen. This is not normally an occurance but it has been know to happen. That is another reason why we are so picky about proper grounding and keeping AC ripple and electrical noise down.

Neal

[HuFlungDung]

Originally Posted by Neal SBC-- To clarify the RS-232 issue for you, If electrical noise of the correct frequency is received thru the RS-232 port, the machine can THINK it is getting a command and it could react to it causing some strange things to happen. This is not normally an occurance but it has been know to happen. That is another reason why we are so picky about proper grounding and keeping AC ripple and electrical noise down. Neal

Maybe the neighbour's remote control garage door opener is talking to the machine

[iess525]

How old the machine? is it AC/DC?

[AccurateCalibra]

It sounds like the coupler between the resolver and the servo is cracked or loose. It is a very easy thing to check. I just found one that was loose after a motor replacement was done. The couplers very fragile. Chris ACI