HAAS VF3 irregular rectangle

[MrFish]

This is with our local Haas service tech but thought I would hit up the forum too as there is a wealth of knowledge on here that can be tapped into.

 

We have a Haas VF3 that is cutting rectangles out of spec from one axis to the other. That is if we finish cut a rectangle with a single contour cut we will see the right size in the long side but about 30 thou of a mm difference in the short side. This is only happening on shapes with a large variance between length verse width (rectangles) . We can machine nice and round bores and bosses and squares come out the right size in both directions. We have had the back lash checked and there is none, and we aren't seeing any lag marks at changes of direction like in bores or on bosses.

My only guess is that there is an error in some type of thermal compensation so we are getting a difference in adjustment that highlights on parts with large differences in dimensions in x verse y.

 

Any thoughts ?

 

 

[gcode]

get the machined lasered

the X ands Y axis are not square to each other

[MrFish]

get the machined lasered

the X ands Y axis are not square to each other

Thanks for reply Gcode,

 

Axis are square and we are cutting a square part ok just getting irregularities in rectangle sizes.

[Matthew Hajicek - Singularity]

I'd still get it lasered.  You may have a worn spot in your screw or something.

[Roger]

Had our 2 year old Haas VF5 SS lasered right after Hass did the yearly maintenance. You should get it done. They adjust the parameters, backlash, etc.

 

Just my 2 cents worth...

[Mick]

If the part is repositioned on the table and machined, does the same error occur?

 

I recall someone encountering a similar error many moons ago, and it turned out to be a worn ball screw from memory. (That is from memory though... )

 

.030 is a big error.

[MrFish]

Thanks for imput Mick and rodger.

 

This is a test we are going to try next, different positions on the table and also try cutting the rec with long side in X and then rotate 90 and try in Y and compare results.

[Matthew Hajicek - Singularity]

 

 

.030 is a big error.

 

He said "30 thou of a mm", which I assume to be .03mm or .0012".

[MrFish]

He said "30 thou of a mm", which I assume to be .03mm or .0012".

Yes that's right Matthew, me and Mick live in the land of Metric, ie anywhere other than the USA

 

But I still consider 0.03mm a big error, a lot of our parts are toleranced to 0.01mm.

[Mick]

He said "30 thou of a mm", which I assume to be .03mm or .0012".

 

Yeah, I just read that. That's the problem, when I have one eye on the screen with Vericut, and the other eye on the screen with the forum.

[Müřlıń®]

Your sine/cosine board in the control could be messed up....it is all about endpoint calculations....

 

Might  have to get a new motherboard...since it will cut round and square, I would say all your tach's and encoders are good.

 

Doesn't sound like a mechanical problem...

 

So the sine/cosine board is making errors on the info coming from the encoders...

 

Use to, all those boards were separate....now it is all combined on the MB.

[MrFish]

Your sine/cosine board in the control could be messed up....it is all about endpoint calculations....

 

Might  have to get a new motherboard...since it will cut round and square, I would say all your tach's and encoders are good.

 

Doesn't sound like a mechanical problem...

 

So the sine/cosine board is making errors on the info coming from the encoders...

 

Use to, all those boards were separate....now it is all combined on the MB.

Thanks for this info Murlin, could be helpful.

 

I did some test cutting of a rectangle at varying places on the table and could get a repetitive error at different places and interestingly when I turned the rectangle 90 degrees the error moved with it. That is we were getting an overcut in the short side aligned with the Y axis but when we turned it the error stayed with the short side and moved to the X axis.

So definitely seems to be some sort of control issue rather than mechanical.

[Orvie]

Is it possible that your using cutter comp & your lead in is not proper?

In your "distance to go" can you see the machine making the comp along the length of the 1st contour?

[MrFish]

Is it possible that your using cutter comp & your lead in is not proper?

In your "distance to go" can you see the machine making the comp along the length of the 1st contour?

No comp applied and we have double checked the numbers, there is definitely a machine fault.

[Machineguy]

How old is the machine?

DO you use the whole table or one side all the time?

Do you do a PM every year? 2 years? 5 years?

Are the drive cards checked for a voltage difference?

 

If you run the machine hard, do a laser check on all the ball screws. It will tell you what's going on as far as wear, or just need to fine tune the comp table. 

Next, IT A HAAS. NOT A MAKINO, MAZAK......ect

Change the way you do a final pass on a part and see if that helps.

.0012 is not a lot, but may not meet the spec you need. 

 

How are you checking the parts? Calipers, CMM?