[DerHammer]

I have a Fanuc 0ma controller built in 87 on my knee mill and I just experienced a new problem. I only have a couple weeks experience so far, so it seems like I'm constantly trouble shooting new issues but I figured thats all just part of getting to know your machine right? Well today I hit a snag more mysterious than most of the time and I'm hoping someone more experienced with my controller can help.

Long story short, I was in the middle of a part run and after inspecting one of my pieces, I realized the work was all shifted in the cross travel direction by about 1/16 of an inch. So now curious, I broke out the edge finder and checked how much off the zero position for that axis was and sure enough, it was out by about .065 from where I originally set it several parts back. So now I have this hunch that needs another's opinion.

I will sometimes Jog the table by hand towards me to make a part easier to remove when finished and I'll accidentally trip the soft limit switch on the cross travel much more frequently since theirs less travel there naturally. Like described, this had been the case once or twice already since zero setting my first part. Am I right to trust my gut here because now I'm wondering if hitting a soft switch could possibly uncalibrate a set zero position? If not, what should I check first? Thanks

[HuFlungDung]

There is nothing like running a batch of tests to find out (without ruining parts). If you have mechanical dials to check the travel with, that would make it easy to determine whether your position display gets out of sync with your actual table position when you hit a limit and then back off.

Can you perform a return to home procedure after hitting the limit? Does it return to the correct home position?

[dertsap]

do you rehome everything after youve hit your limit switch , some machines can and will loose track of where they are if interupted by over travels and such

[DerHammer]

actually.....I did not. Now this is going to sounds really dumb but I haven't really been using the homing feature at all because to be honest, I have no idea what it does .

Yeah I know, I'm a total newb but I guess I was under the impression it was only for setting the travel limits of the table, but then again I was just guessing. I suppose I need to learn about what its for sometime right. How exactly will homing the machine correct a zero position thats out? I'm having trouble understanding how they relate?

[dertsap]

all machine axis should always be sent home upon startup

http://books.google.com/books?id=Wf6...VArIA#PPA24,M1

[DerHammer]

Thanks dertsap, that helps allot. I read through the link and ill be sure to get into the habit of homing but I am still a little confused on something. Whenever I boot up my machine for the first time of the day, it will always zero the DRO where the table sits. I usually return my table to my part zero position before shutting down the machine so I still have it the next day when I go back out to the garage.

Now if I understood the link correctly. If I don't machine home the mill first, then some of the measurements of my parts could be changed because all the machines internal calculations and its accuracy is relative to the accuracy of the machine home position. So heres my question... Does this mean that I need to reset my part zero position every time immediately after homing the machine like in the event of tripping an over travel limit for example or booting the machine back up the next day?

[trubleshtr]

Your macine might be moving a little a power up/down.

After power up, home machine, then set-up your part zero position and zero out your DRO. should be good to go.

If problem arises during your "production run" and machine has not been turned off or hand wheels moved manually, then I would suspect backlash.....and I would definetly follow Hung's advice...

[dertsap]

Does this mean that I need to reset my part zero position every time immediately after homing the machine like in the event of tripping an over travel limit for example or booting the machine back up the next day?

no !

if you have to pickup the workshift each time you startup the machine you have problems !!
if all is good with the machine you should be able to startup ,home everything in ,warm up the spindle (very important) , and then press cycle start

[DerHammer]

I tried to home the machine and for some reason the soft switches trip before the control panel lights up indicating that its homed. When I first bought the machine this was giving me trouble then also, and I think there was a key combination during start up one could use to reset the soft switches in which you then immediately home the machine to permanently set it? Is this right and if not, what should I try to remedy this? as always, help is always appreciated. Thanks

[HuFlungDung]

You might have to disable or extend the soft limits temporarily in order to allow the machine to move from its 'power up zero' to its true homed position (as allowed by various methods). You might have to do this at any time whenever the power is interrupted and the machine has been left sitting somewhere other than at the home position.

I would try various abort scenarios and return to home procedures to establish what the machine is capable of attaining. If it will return to a given position with .0001" accuracy in all axis, then that is great, but if not, then you'll want to know about the magnitude of the variation, and analyze why, and if anything can be done to improve on it.

[Dan Fritz]

Watch out for a switch or pushbutton labeled "ABS" or "Manual Absolute". If ABS is turned off, then when you jog or MDI the machine, the absolute position zero is SHIFTED by the amount you move. If ABS is turned on, then the absolute position zero should stay put, and the next absolute move in your program should move right back to where it should.

The ABS signal exists on all Fanuc "M" (mill) controls, and it causes a lot of grief. If I had a Fanuc with an ABS switch, I'd permanently jumper the switch so the signal is always on.

Hitting your software OT limits should not affect your absolute machine zero at all.

[DerHammer]

First of all, I want to say thanks to everyone helping on this thread. So now down to business. Dan, I have seen that ABS button on my machine an wondered what its all about but I've always left it on the absolute setting. At least I think I have. By that I mean that on my DRO, I have I can select from what appears to be 3 views. Absolute, Relative, and All where it shows all positions at once. The problem is that I'm not sure if selecting between these is actually changing a mode or rather just what I'm viewing but I suspect that I am in absolute because if the zero position was indeed changing by jogging the table, It seems like the shift would be occurring in much greater increments if I understand this correctly. I spoke to a tech today on the phone who was convinced that I was probably in incremental mode, but I check my part programs and there is a G90 in the beginning of every program which should be absolute I believe, so I don't think thats the issue.

I was thinking that if this were an issue due to shot bearings, belts slipping, etc, there should be error all over the place but there isn't. The whole pattern is shifted by small fractions of an inch but everything else is still accurate it would seem and spaced correctly like pocketing, etc. Heres the next part which bugs me. If the zero point was slowing crawling across my parts program to program, I would think that there would be a slow and steady compounding of error. you know starting at .005 to .010, .015, etc but rather out of six parts, The error is split. Three parts about right around .032 off and then next three right around .064 off. I might be off on these numbers a little because I don't have state of the art measuring tools yet, but more importantly, I can make 2 stacks and the parts in each are identical like its happening in stages. So this is starting to make me learn towards the table shifting during power up theory mostly because these 6 parts were cut over the course of two days. How likely would this be on these old OMA machines and is their an easy way to narrow it down further? Thanks

[HuFlungDung]

Have you kept track of the direction of the error? It is possible for noise, or fading electronics to cause positional drift. I've had this happen on old cnc's just before they really go bad. But, inevitably, to track down the error, one has to know which axis is causing it, then try swapping motors, or cables or drives to try to determine which component seems to be the problem.

Swapping components to try to resolve a bad one is risky. Motors may run quickly to the end of travel if encoders get reversed or disconnected from the motor they are normally coupled with. If you can disconnect the couplings or belts before testing, runaways are less serious.

[DerHammer]

Electronics failure is a possibility but I really don't think its likely in this case. Until recently, I haven't been homing my machine at all and I probably have tripped the over travel alarm 20 times or more on the cross travel during that period of time in which I didn't know its importance. Seems unlikely to me that its is purely a coincidence because this happens to be the same axis that the shifts on. That and I'm only tripping that alarm in one direction of jog and my part shift also only happens only in one direction. Seems like that would be an odd coincidence if it were. So yesterday I ducked under it, found that my limit switches had a fair amount of chips around them, which I cleaned off and I'm now homing the machine during startup. I'm just gotta do another batch of parts now to see if its resolved. Heres to crossing my fingers, ill keep you guys posted