[SadieD]

Hi, my shop got a 1984 Makino FNC 40 A16, and I've done a bit of work on it to get it operational again. One of the repairs I made was replacing the X axis servo amplifier and I think that's probably related to my current issue. We just ran a program on it for the first time today and every thing was fine except for the positional accuracy of the x axis, it seemed to drift about .250" over 16 inches. Could the inaccuracy be due to not performing any adjustment to the parameters, settings, or pots on the amplifier?

[agering]

If inaccuracy appears in operation that you need to do on drive installing the driftage of the zero by resistor R2 (OFFSET). If it is parameter N007 #7(ADFT)=1, that necessary parameters N124-N127 to set in 0.
Change OFFSET possible to see using diagnosticses X=N800, Y=N801, Z=N802, e.t.c.
After setting the value greatly drawn near to 0, possible adjust the gain by resistor R1(GAIN). This will give acerbity of the positioning.

[underthetire]

I'd pull covers first and look at all the mechanics, especially if it has timing belts for the axis. Last thing you want to do is electronically "fix" this and have it be something mechanical.

[emexcee380]

You've got DC motors on that machine - correct? (Yellow cap on motor)

Initial thought is that your GAIN (stiffness) pot needs tweaked. Detailed info in
the Fanuc DC SERVO MAINTENANCE manual. Got one?

If not, simply look at the top pcb (printed circuit board) of the drifting axis. One of
vertically arranged potentiometers ("pots") will have a the word GAIN printed beside
it. Note the original position and tweak away one graduation at a time while watching
your indicator until the axis is stable.

Good luck !

emexcee380

(Just saw Underthetire's response. Always do what this dude says. He's forgotten more
than most of us know.....)

[cjfisher]

I agree with Underthetire... check your machine's mechanical connections. From pictures I have found, looks like the x axis may be directly coupled (no timing belt). One idea you may want to try: affix a block on side of the table and zero on the edge with an indicator attached on the z axis, then run the x axis back and forth by bumping the manual rapid - you want jerky starts and stops, finally jog back to your stop and check the indicator. If the indicator is off different amounts check the mechanical connections, if the indicator and control agree very time on position then you may have a parameter issue.

[SadieD]

The x axis servo is mounted directly to the ballscrew. They're DC servos.

[SadieD]

You've got DC motors on that machine - correct? (Yellow cap on motor)

Initial thought is that your GAIN (stiffness) pot needs tweaked. Detailed info in
the Fanuc DC SERVO MAINTENANCE manual. Got one?

If not, simply look at the top pcb (printed circuit board) of the drifting axis. One of
vertically arranged potentiometers ("pots") will have a the word GAIN printed beside
it. Note the original position and tweak away one graduation at a time while watching
your indicator until the axis is stable.

Good luck !

emexcee380

(Just saw Underthetire's response. Always do what this dude says. He's forgotten more
than most of us know.....)

I've had a lot of trouble finding manuals for any of Fanuc's stuff. The axis doesn't seem to be wandering, it's like it's scaled wrong because the error grows over distance.

[cjfisher]

Can't help on the manuals or the parameters... I would look at parameters though and not the servo tuning. As long the machine seems to drive smoothly where you want it to (well mostly) and back it is likely not a tuning issue. I would look at parameters regarding the ball screw pitch, and pitch compensation.

[SadieD]

If inaccuracy appears in operation that you need to do on drive installing the driftage of the zero by resistor R2 (OFFSET). If it is parameter N007 #7(ADFT)=1, that necessary parameters N124-N127 to set in 0.
Change OFFSET possible to see using diagnosticses X=N800, Y=N801, Z=N802, e.t.c.
After setting the value greatly drawn near to 0, possible adjust the gain by resistor R1(GAIN). This will give acerbity of the positioning.

I'm going to give this a try. The procedure is also mentioned here by tc429:
http://www.cnc-arena.com/en/forum/circular-interpolation-fanuc-6m-not-round--222446.html