Fanuc 0iMC - Scales vs Encoders. Accuracy.

[Leha_Blin]

Dear All!

I've got the following situation. Milling machine has CNC Fanuc 0iMC and linear scales. I was measuring the machine accuracy using Renishaw Ballbar.

So when the machine has only ENCODERS activated after some measurements and corrections of parameters 1851 and 1852 I had Backlash error about 5-7 micron, Reversal spike - about 10 and out of roundness about 12-15 micron.

After LINEAR SCALES activation and setting 1851 and 1852 to zero Backlash error became even less but Reversal Spike and Out-Of-Roundess increased too much - about 20-25 micron.

Renishaw help explaines that there is some slow reaction to reverse of motion direction.

Do you have any ideas about this problem?

[xavierdemoura]

Leha_Bin,

You only will take out the spikes change the gain´s of the motor´s.

Daniel

[f-bu]

Lexa,
do not reset backlash to zero, leave 1 in 1851 for every axis, because if the machine has "backlash acceleration" set, it will not work. From my experience, using external feedback needs new adjustment procedure. "Servo guide" can help doing this, if you have it.

[Leha_Blin]

f-bu: Well, I think I don't have it. Could you advice where I can get it? And could you please describe "backlash acceleration"?

[f-bu]

"servo guide" is Fanuc tool for adjusting servos. Chinese sites have it for downloading. For a pity it works with Fanuc supplied PCMCII ethernet card. Working trough embedded port or data server port return strange errors. Maybe it is version incompatibility, who knows. Someone told me that 16 bit card can be used instead, but I had no success with different ones.
Backlash acceleration - see 2003.5, 2048, 2094, 2071 in servo parameters manual
Typical setting is 1, 600, 600, 100, but depends on input increment and machine, so start from 50 for 2048/2094
It is quite long to explain, I prefer to talk after you read it in the manual.
PM для смены языка.

[cliwar]

Leha Blin,

If you post an image of your Ballbar diagnostic report we may be able to provide some more insight on what exactly your machine is doing.

Clive Warren
Renishaw, Inc.

[Leha_Blin]

This is the result. If it's not enough to understand tell me what to add.

[cliwar]

Leha_Blin,

Thanks for posting the ballbar plot, unfortunately without having access to the original .B5R file I cannot do any kind of in-depth analysis for you. However, something that you do need to keep in mind is that if your machine is using linear scales, it cannot have backlash (unless the scales themselves are loose!). What you could be seeing is the effect of the machine table yawing as it changes direction. To check this, re-test the machine with the ballbar in different locations on the machine table. If it is the effect of yaw, the apparent "backlash" will vary depending on where the ballbar is located.

Clive