[IrishJ]

Hi All

This is my first post here so go easy on me,

I have a fadal 4020 and after running a ballbar test on the machine I replaced the trust bearings on the X axis and Leveled up the machine and set my backlash, It all looks prity good except the Scaling error is -219.1 PPM on the Y and -179.8 PPM on the X . Can anybody tell me how I could fix this

Thanks in advance

John

[mflux_gamblej]

Hi All

This is my first post here so go easy on me,

I have a fadal 4020 and after running a ballbar test on the machine I replaced the trust bearings on the X axis and Leveled up the machine and set my backlash, It all looks prity good except the Scaling error is -219.1 PPM on the Y and -179.8 PPM on the X . Can anybody tell me how I could fix this

Thanks in advance

John

you need to adjust your pitch comp tables. on a fadal this is done under the survey 'SV' menus. best way to do this is to have the machine laser calibrated using an interferometer.

that being said .0002 inch error is pretty much zero - get over it. i doubt you'll be able to control tool runout better than this anyhow. most parts are +/- 0.005", some +/- 0.001" so you're doing way better than necessary. maybe run that ballbar at many locations around the table, see what ppm's you get. you could use that to do some SV stuff rather than laser, but laser is going to be the best, good down to 1ppm measurement.

fadal machines are not a makino claiming 5 micron tolerances on parts. just make parts, don't get caught up on any problems like 0.0002" error. its a waste of your money

did i count that right? 1 ppm is 1 millionth of an inch.

0.001 thousanth
0.0001 ten thousanth
0.00001 hundred thousanth
0.000001 millionth (1ppm)

so 100 millions is two over so tenths... yea i did count it right. just get over it

[IrishJ]

Thanks for that Mflux

[mflux_gamblej]

Here are two ballbar tests on my 4020 machine. One at 40ipm, another at 80ipm

I just did a huge amount of maintenance. new waylube, new spindle, new gibbs....

also this machine has not been laser calibrated yet (because i knew it needed gibbs and lube done). I am working on laser calibrating it now. I will post new ballbar tests after laser cal. I will then attempt to tune up the servos by messing with the trim pots on the axis amp cards and matching it up on the following error readout / etc. I will then post a 3rd set of ballbar results which will be for after laser cal and after servo tune-up.


right now my machine looks pretty sick...

[mflux_gamblej]

Here's how laser cal went for the X axis - made a big difference. not sure why my ball screw is so messed up in the last few inches of the right side of the table on this X axis. the comp tables can only do 0.0008" / 1.0" updates so its as good as i can make it. you can set and verify backlash during a laser cal too...

Attachment 339910

[mflux_gamblej]

This morning while the laser was still setup on the X axis I decided to run the machine in to warm up the ballscrews. I ran it at 150ipm for 1 hour x and y axis' using M11, M12 reciprocating axis motions (just back and forth infinite).

holy sh*t the screw expanded a lot. it is 0.007" / 40" under this condition. I noticed that the fadal factory SV tables usually biased the ballscrew about 0.005" / full travel (in this case its a 4020 so 40" x) towards the short side. I see why now.

I don't always operate in a condition where my screws experience 150 ipm all day long, and there may be plenty of downtime during the day - so I will probably just pick a number in the middle between 7thou and 0 (cold cal). and I will bias my SV table accordingly.

here's the data of Cold vs Hot screw. what a pain in the a$$.

[IrishJ]

Thanks a million for all this info mflux_gamblej, you are sure putting a lot of work into this, Can you tell me which Laser calibrator you are using. Maybe its time I invested in one

[mflux_gamblej]

Hey I got myself a renishaw qc10 ballbar kit. hopefully it is easy to use. we can compare results on my fadals vs. yours.

i'll likely use it more to tune the servo amps for roundness at various feedrates.

[mflux_gamblej]

Here's the ballbar results after laser calibration. it looks like the backlash is different dynamically than when running slow and doing the laser calibration stuff.
what did laser calibration improve? 1.) scaling mismatch, 2.) positional tolerance, 3.) backlash 4.) circularity to some extent


Servo mismatch is the biggest problem now. its at about 2.0 milliseconds for both 40 and 80 ipm

Attachment 339990

Attachment 339992

[mflux_gamblej]

servo tuning made a big difference for my machine.

master clock program is supposed to take 72 seconds, mine was 77. i got that put right by adjusting pot on clock card.

then x,y,z following error adjustments to set inch ballscrews to 595 at 150ipm feed. x was 545, y 585, z 580. got it all to 595. made ballbar result go from 4 thou circularity to 1 thou. I also adjusted BL settings to get the ballbar test minimized - so its a dynamic setting not for static like when using the laser cal setup.

here's the final results - any more tweaking is right down in the weeds / you could spend a year doing this work...

Attachment 340140

Attachment 340142

so, after all this sh*t i went thru on my machine and ended up with 0.0012" circularity.... i think if it is even true that your fadal is down at 0.0002" thats probably better than factory - would love to see your plots to prove it.