What's new
What's new

Fanuc 31i hangs sometimes at M code (SOLVED)

EKrol

Aluminum
Joined
Mar 2, 2023
Hello Everyone,

We have a problem with our Doosan with 31I fanuc control. Sometimes the machine stops at an M code. In this case it hangs sometimes at M52 auto door open. I don't understand why. I put all the inputs in a trace and when the m52 is active the door opens. The sensor door open gets activated. The door open command finish and the M fin (G4.3) is active but still the control stops at the mcode. Nothing in the ladder is waiting for something else it just stops. If i select single block i can go further and the machine wil continue. I put waitcodes in front and after the mcodes. I put the mcodes in a sub program and call them as a program but still the machine will stop sometimes. i tried G5.1 q0 q1 etc but still no succes. The first time i started the program it run for 15 parts and then stopped. Now i can only make 2 a 3 parts before is stops.

Does anyone have any idea where to look?

Gr EKrol

[EDIT]
Problem is solved. The axis move a little when the door opens because the axis lose power when the door is open. Servo error is sometimes above 200 what cause a in position check what stops the control with no error message. I don't know why this value did get bigger in time. I didn't want to change the in position error value in the parameters while normal machining. Now i changed my robot sub program to adjust the value to 500 before opening the door and set it back to 200 when the door is closed with G10.

Thank you all for you advise.
 
Last edited:
You might have some chips or other gunk blocking the door path or one of the many sensors verifying the activities of its movement. Perhaps you got lucky the time or two when your followed along in the ladder.

When the machine stalls is there ever a message in the Buffer Area on the screen suggesting waiting for FIN or anything?

Anyway... my point is, check for physical obstructions or even a faulty, intermittent wire. Wiggle a few wires as they say. A ghost is in there somewhere.
 
You might have some chips or other gunk blocking the door path or one of the many sensors verifying the activities of its movement. Perhaps you got lucky the time or two when your followed along in the ladder.

When the machine stalls is there ever a message in the Buffer Area on the screen suggesting waiting for FIN or anything?

Anyway... my point is, check for physical obstructions or even a faulty, intermittent wire. Wiggle a few wires as they say. A ghost is in there somewhere.
Tnx for the reply.
Mechanical there is nothing wrong with the door. No message when the machine stops. The autodoor open finish signal is activated and that signal activate the mcode fin signal. Those signals are getting activated. I did put all signals in a trace and it's looking good. But its still not going true. I think it's in the timing for the mcode fin signal. I have no idea if you can adjust anything on that subject.

Ekrol
 
May not work, but try inserting some buffering-preventing code, such as G31 without arguments, after the problematic M code.
And, specify M code in a separate block.
 
Capn dummy, do you physically open and then close the door? or is it just a random m52?

if so, then you might not be closing the door all the way, or if you are not opening the door, then it needs to be opened and closed so the lock mechanism has a chance to actually lock, the light might be off, but the latch has to be down or it will not move past, UNLESS in Single block, even then it might want to automatically go to 25% rapid.

don't tell OSHA, but the first thing I did as soon as the machine setup guys leave is remove the damned lock... its just complicates things and makes making tooling and setups a total PITA if yer running a CNC with the door open, just because you like the feel of chips in your hair... Darwin would like a few words with you anyway.
 
Capn dummy, do you physically open and then close the door? or is it just a random m52?

if so, then you might not be closing the door all the way, or if you are not opening the door, then it needs to be opened and closed so the lock mechanism has a chance to actually lock, the light might be off, but the latch has to be down or it will not move past, UNLESS in Single block, even then it might want to automatically go to 25% rapid.

don't tell OSHA, but the first thing I did as soon as the machine setup guys leave is remove the damned lock... its just complicates things and makes making tooling and setups a total PITA if yer running a CNC with the door open, just because you like the feel of chips in your hair... Darwin would like a few words with you anyway.
It's an automatic door. So the door will open physically. There is a robot cell attached. I'm used to work with the machine with a working locking mechanism. It did take me some time to get used to it.
 
May not work, but try inserting some buffering-preventing code, such as G31 without arguments, after the problematic M code.
And, specify M code in a separate block.
I did tried this but no luck. I want to try to add M52 in the m code buffering prevention parameters. (3411-3432)
 
Door opens? Is there a robot signal missing/waiting?
What is the next operation command? The hang can really be the next command/operation.
 
Problem is solved. The axis move a little when the door opens because the axis lose power when the door is open. Servo error is sometimes above 200 what cause a in position check what stops the control with no error message. I don't know why this value did get bigger in time. I didn't want to change the in position error value in the parameters while normal machining. Now i changed my robot sub program to adjust the value to 500 before opening the door and set it back to 200 when the door is closed with G10.

Thank you all for you advise.
 
Last edited:
Any particular axis that is drifting? Based on your attempt using G5.1 in your OP, guessing a mill. If it’s a VMC and Z is drifting, possibly a bad brake. If X or Y then look for chips compacted under or behind way covers.
 
Any particular axis that is drifting? Based on your attempt using G5.1 in your OP, guessing a mill. If it’s a VMC and Z is drifting, possibly a bad brake. If X or Y then look for chips compacted under or behind way covers.
Its a millturn. Doosan smx2600s. Its the x axis that is moving the most. I guess if i edit the servo off delay in the ladder it will be better. Think after 2 years the x axis brake is taking a few mili seconds longer to fully clamp. The x axis brake have to work the hardest because it holds the milling head like the z axis on a milling machine.
 








 
Back
Top