fanuc 18m 351 alarm

I have a Supermax v65A with a fanuc 18MC controller.
It used to produce a 351 x-axis alarm followed by a 414 alarm in X-axis. (most of the times, into a toolchange cycle).
After a little investigation I found that the previous owner tried to replace the X encoder cable. Eventually, he didn't replace the whole cable but only half of it.
I believed that was the problem together with some kind of cable-grounding error.
I have replaced the cable by myself, but I still see from time to time a 351 x axis alarm, even when the machine is on and stoped (controller into MDI mode ).

Any suggestions?

Thanks
Michael

I swapped axis cables and error disappears.
Any thoughts?

Is it possible to have the problem due to a broken spark killer or diode, because most of the times, the alarm is happening during M06 cycle?
Has anyone met anything similar?

Thanks
Michael

try to put some ferite core on your encoder cables if you do not have,might be some noise on your wires.

I have tried ferrites but nothing changes.
The fact that after I do a cable swap between X and Y axis and still the error is 351 x-axis alarm and not 351 y-axis alarm,I think it is not the servo drive.
I have already tried a new encoder with no luck.
Is this any possibility of a faulty CPU?

Michael

you should have 2 axis card control on the mainboard,1 is for 1,2 axis and one is for 3,4 axis.
Swap them and see if the alarm goes to Z or 4th axis.

I have a 3 axis card.Which cables should I swap? Cables to motor, encoder and control cables?

i was not talking about the driver that powers the motors,i was talking about the servo module cards from the main board,you should have 2 identical and u can swap them,just identify it and try.
If you do not know and is hard for you to take out the mainboard just call Fanuc.

I swap servo module a20b-2902-00700 of 1/2 axis with the same module of 3/4 axis.Problem still the same and most of the cases when in or after a while of a toolchange cycle?
The manual suggests when #5(STB) and #6(CRC) is 1 then
a) signal cable is disconnected
b) serial pulse coder is faulty
3) main cpu board is faulty (i don't have option board 2)

Maybe I must change the main CPU board or any module of it?

When the tool change is executed ,the X and Y go to reference points?

No, just where the program is interrupted, X and Y stop at whatever point they are,M06 is executed and then the program continues (if the error doesn't happen)
The error would be happen anytime inside the m06 cycle, or just a few seconds after m06 ends.
If I will run a program having m06 always with the same tool (avoiding toolchange) the error will happen again but some time later, let's say from 2mins to 15mins later.

did u have the machine grounded?i mean a solid wire from the main body to the centerline of factory ground?
do you have equipment that is welding using same power line as the machine does?

The machine is grounded.
Everithing is happening when no other equipment is on at the factory.
The strange thing is that 90% of the circumstances, the alarm is happening during or afterwards a toolchange.
I though it might be a spark killer of the motor that brings the umbrella of the tools forward and backward to the spindle, or this motor's brake.
I dont know how to check the spark killer and the brake is supplied ok with 90v when the umbrella motor is moving.

Another thing is that our mains is 410V but the transformer of the machine is 380V. I have already ordered a new tranformer so I would check with 220V and not 234V as is now.
The transformer is a little noisy, at the back side of the machine, and close by is the metal pipe whitch has the x-pulse coder cable inside. I moved the transformer about 0.5m away but still no luck.
The power cables go from the transformer to the machine and back and for some length we have 2 power cables near the metal pipe which has the x-pulse coder inside and the metal pipe which has the y-pulse coder inside.
I have checked DC bus is at the upper limit of DC.
24v is OK
5V is ok

Thanks
Michael

ok let me understand the atc cycle:
-magazine rotate to the tool number in spindle;
-z goes up to reference point
-umbrella has a pneumatic cylinder moving foward to spindle side
-tool unclamp
-z moves to second ZRN to go out from tool
-magazine rotates to desire tool;
-z moves down to the tool
-tool clamp
-umbrella moves backward
tool change complete

if this is happening on tool change i would try to isolate what moving from this tool cycle is generate the alarm
you should have M codes for magazine foward/backward;
if so,make a small program with some dwell time between and do some magazine moving foward and backwards for ,lets say 30 minute and see if it is happening.
also u can also do on the rotation ogf magazine by making a program with only calling a tool in standby with no tool change,like:
T1;
G04 X15.
T2;
G04 X15.
....

In this way you will isolate the umbrella and motor moving.

good luck and let us know about results.

The toolchange is happening exactly as you mention exept it has not a pneumatic cylinder but a 3phase motor.

When the alarm is generated during the toolchange macro, it doesn't happen on a specific instance. Maybe it will happen when magazine goes forward to spindle side, maybe when it goes backward, maybe when z leaves the old tool, maybe when z goes down to clamp the new tool.
It is possible to do 3 toolchanges in a row from MDI without an alarm but it will happen on the next.
The alarm never happens when the magazine rotates buring the toolchange macro.

I have tried to move the magazine, calling the M codes for back and forth as you said but I couldn't succeed.

Do you think the mainboard is ok? and we have not as a cause to the alarm some noise?
Maybe I should change the pulse coder and its cable again?

yes,the pulsecoder of X should be identical to Y axis so can be changed,also the cable from motor to cabinet side.

PS:for me sounds like the there is a cable issue for the encoder of X

The alarm is happening always with the x axis stopped. I have tried to move the cable from both ends in case any of its wires is loose although I checked its wiring and ground.

I have started from the begining to find maybe a pattern of what happens to trigger the alarm.
This morning I referenced the machine then issued a M19 spindel orientation into MDI mode and all the times after some seconds the alarm happened without any movement of the machine.
If a just reference the machine without doing anything then alarm doen't appear.

so if you try M19 several times will generate the alarm?

Every time, but the time the alarm follows it varies from 2 sec to 10 sec.
I found that my machine has a M sensor and a position coder because there are cables connected to JY2 and JY4 ports of the spindle amplifier.
My spindle motor is A06B-0854-B100 and the amplifier is A06B-6078-H211

so we isolate the problem that could come from the optical encoder that should be connected to spindle by a small belt?

I will check it tomorrow.
Thanks for the help.
Michael

I found out that the previous owner had replaced the spindle motor and this one has inside a MZ sensor although the old one had a M sensor, because the plans I have state this.
I checked the cable of MZ sensor and the only problem I found was that the shield is not connected to earth directly, but to pin 10 of the JY2.
I checked the parameters and as far as can judge seem ok.
I can not understand what that would be to trigger the alarm of X-axis communication even if it seems that is something with the spindle orientation.

Michael

what show diagnose 200,201,202,203,204 when alarm occur?

Only 2003 X crc=1 and stb=1, all others are 0.
This morning I tried some g84 cycles and alarm 351 continuous and also #200 x dca=1

that means mainboard or option board fault :)

Do you know how to change the mainboard?
I think I don't have option2.

Attachment 419498

this is my mainboard.
Couldn't reply to your PM. Your storage is full.

Goodmorning
Michael

I checked the pulse coder cable, OK.
I added a cable to ground from the replaced MZ sensor (the machine originaly had a M sensor in the spindle motor).
This way the alarm hasn't appeared yet.
Let's see.....

hi michael .
how to solve your problem # alm 351 ?