X axis servo disconnect alarm 416

[theblacksmith22]

i need some help with a lathe, so about a week ago this lathe was crashed really bad, worst I've ever seen it was crashed so hard that the turret mounting plate that attached to the ball screw and the guideway's broke in half, I got it all apart and repaired and put all back together. go to start up the machine all goes well till i go to move the x axis up, then i gat the 416 disconnect alarm i have gone thru everything and cant find a problem I replaced the amplifier, same thing i replaced the pulse coder ohmed out the cables from the servo to the axis card. removed the belt from the ball screw and the motor, same thing. also getting x axis excess alarms following the disconnect alarm. one odd thing that I've been noticing is the servo motor will move by itself with no commands now i know there has to be compensation for axis drifting but that's not this, in creeping really slow until it throws the alarm. the machine is a 1994 marathon sl500. homing is by decel dogs its the weirdest thing next plan of attack is replacing the axis card on the control. an help or ideas would be greatly appreciated thanks.

p.s during the teardown after the crash the servo was not messed with at all until it started alarming

 

[Vancbiker]

Is there a scale on that axis?

 

[theblacksmith22]

no scales

 

[Vancbiker]

The symptoms you describe are typical of a machine equipped with a scale, when the scale signals are not being received by the CNC. If it doesn't have a scale then I'd scope the encoder signals at the control to see if they are bad.
Would be a good idea to post what control you have since some models connect the encoder signals to the CNC while others connect to the drive.

 

[theblacksmith22]

ahh i missed the control its an 0-t. and i just learned that the solenoid looking part on the shaft of the x axis ball screw is a heidenhain pulse coder which i guess is a scale as well i shut prm 37.0 off to eliminate that scale same thing that was recommended from Fanuc maybe there's more prms that need shut off to pair with that ?

 

[theblacksmith22]

I'm pretty confident that i narrowed it down to the cnc side. swapped x and z scales and x and z motors on the pcb and the alarm transferred to the z axis

 

[memphisjed]

Depending on crash could the glass in encoder have cracked?

 

[Vancbiker]

ahh i missed the control its an 0-t. and i just learned that the solenoid looking part on the shaft of the x axis ball screw is a heidenhain pulse coder which i guess is a scale as well i shut prm 37.0 off to eliminate that scale same thing that was recommended from Fanuc maybe there's more prms that need shut off to pair with that ?

Easier and proper to figure out what is wrong with the Heidenhain encoder and fix that. And yes, it is being used as a "poor man's" scale in this application. If you want to run with 37.0 set to 0 to not use the scale for position feedback, at the very least you will need to figure out new values for CMR and DMR parameters. Since the builder added the Heidenhain encoder, sounds like this axis is belt or gear driven. If so and there is a reduction ratio, you may need to figure out a value for the flexible feed gear parameters. That would depend on ratio, pitch of screw, Fanuc encoder count, and desired least programmable increment (typically .0001" on a lathe X resulting in .00005" motion).

 

[Vancbiker]

I'm pretty confident that i narrowed it down to the cnc side. swapped x and z scales and x and z motors on the pcb and the alarm transferred to the z axis

Since the alarm moved, it is an indication that the CNC side is fine and that the trouble is axis related.

 

[theblacksmith22]

Since the alarm moved, it is an indication that the CNC side is fine and that the trouble is axis related.

okay where would you recommend i go from there ? i already ohmed the cables and replaced the axis card on the control

 

[theblacksmith22]

okay where would you recommend i go from there ? i already ohmed the cables and replaced the axis card on the control

and yes this axis is belt driven

 

[Chimpwithastick]

okay where would you recommend i go from there ? i already ohmed the cables and replaced the axis card on the control

You tested resistance on the cables to what? End to end, individual wires to ground, wire to wire? Did you test the insulation with a Megger? Often an ohmmeter's power isn't strong enough to push voltage through damaged insulation, but the control power is.

 

[theblacksmith22]

You tested resistance on the cables to what? End to end, individual wires to ground, wire to wire? Did you test the insulation with a Megger? Often an ohmmeter's power isn't strong enough to push voltage through damaged insulation, but the control power is.

resistance wire to wire and wire to ground i do not have a megger

 

[Vancbiker]

You need to check the signals from the Heidenhain encoder at the control with an oscilloscope. Based on the alarm and your description of how the machine behaves, I strongly suspect you will find that there is a channel missing, erratic, noisy or otherwise defective.

 

[theblacksmith22]

one problem solved, turned out to be the secondary pulse coder on the ball screw. god everything put back together and running well during this adventure i replaced the servo one with a rebuilt one during the process of finding center and squaring the machine up had no problems. started running parts that servo motor thru an over load the temp on it was 118 C let it sit for a while came back to it after it cooled temp spiked super high again just sitting in idle any ideas ???