[laka]

I have a 1989 Sabre 750 with Fanuc 0-MC. Great, strong machine.

I've had it for about 7 months, and have been living with this problem but now it is holding me up on a job. It seems as tho when the torque to either the spindle motor or an axis servo spikes up a little the machine will feed hold itself automatically. If the load is constant (say face milling) it will take heavy loads all day long. Currently I have a job and am using some indexable carbide drills (1" and 5/8" dia) to peck drill some holes and almost at every peck the machine will feed hold itself. Sometimes it stops during a cut, sometimes when it's retracted to break the chips. I'm definately not overloading the motor to any extent whatsoever with these tools. To complete the holes I need to stand at the control and constantly press cycle start and let the hole finish. Machine will also do the same thing sometimes when say endmilling into an inside corner, or doing other work which may quickly load up the spindle.

Any ideas?

I've checked the Spindle Drive unit and no alarms are ever displayed when this happens. Same on the control itelf, no alarms are displayed. Only my feed-hold light turns on.

[mgnbuk]

This control will most likely have a built-in PLC to interface between the NC and the machine.

Fanuc PLCs use various addresses for physical I/O (X for inputs, Y for outputs) internal communication (G for signals going from the PLC to the NC, F for signals going from the NC to the PLC) and general use (R).

The PLC program controls the feedhold function using an internal flag called *SP, which is a G signal. The actual address varies according to the control - I "think" the 0-C control is G121.5, but if you look in your Maintenance manual there is a list of signals used with that control in the PLC program. The * signifies that this is a signal that has to be high (or 1) in the logic for the machine to run (like a normally closed pushbutton that opens when you press it).

Now you need to find out what is driving this signal & how easy this is depends upon what Cincinnati supplied. I seem to recall that the 0C had the ability to display the active PLC ladder similar to a schematic diagram, with the active signals highlighted. This is accessed using the Diagnosic/Parameter display key- press once for parameters, again for diagnostics and, if the ladder can be displayed, again to get the ladder. If this is present, you can search for the G121.5 signal & display the "rung" of the ladder that controls it - by monitoring this in cycle you should be able to see which signal is going away & causing the feedhold. If the signal has an X address, it is a physical input from the machine - refer to the electrical schematic to find out what the input is. If it is an F address, this is a signal from the NC - look it up in the Maintenance manual to find out it's function. If it is an R address, this is a "User flag" set by the PLC programmer - search for the R flag in the ladder to find out what is causing it to drop out.

If the active ladder display does not come up, then the signals can be monitored on the Diagnostics pages - just look for the numerical part of the signal address (the numerical values are all unique - you will not find X0.0 and G0.0 for eaxmple). The data is displayed in 8 bit words, with the numbers before the point being the address of the 8 bit word & the number after the point the bit required. The bit addresses run from 0 on the RH margin to 7 on the LH margin, so to look up G121.5 you would search for the 121 line in Diagnostics & look at the 6th bit from the RH side, where 0 is "off" and 1 is "on". The machine technical documentation should include a printed version of the ladder diagram - find the "rung" that controls the *SP output (for sake of argument G121.5, which will be down the RH side of the ladder). Then you will have to monitor each element of the rung in turn using the Diagnostics display to identify which signal is dropping out.

As each machine tool builder writes their own PLC program, it is not possible to guess what signals they have decided to include in the Feedhold line - when I was retrofitting these controls, I would include "critical" monitoring in there like, say, drive fault signals. Something less critical, like low axis lube, I would put into the Cycle start line - the machine didn't just stop part way through a program just because of low lube, but you couldn't restart it until the tank was refilled.

As the *SP signal has to be present for the machine to run & a momentary drop-out will force the feedhold, intermittent loss of a physical input incorporated in the relevant "rung" of the PLC program due to a loose connection would give the symptons you describe. You will just have to go digging to find out which signal you are loosing to make progress in finding the actual cause of the problem. Good luck !

HTH

Nigel B.