control choices

[surface]

Fadal now offers a choice of 3 controls. The 104/D, Seimens 840D and the Fanuc 18i MB5.
As with anything, each control has strengths and weaknesses. At this time, the 104/D seems to be dominating the weakness catagory. So what are the strengths of the other controls and which control would you choose?

[Scott_bob]

Where there's a will, there's a way...

The problem with chosing the way of "try before you buy" requires time, and costs. Money well spent though.
Claims to accuracy and speed and therefore value, can only be validated by doing a test cut. It's the only "way"...
Now, if a buyer wants to have a test cut done on say 3 CNC machines that he is comparing or 3 controls, that is going to take some time and patience is needed. Opinions are not substantiated without testing. Talk is cheap...

We recently had a test cut done by a CNC machine builder from China with a Fanuc 18i control, which is considered by many, to be one of the best controls. The results were very diaapointing, and we can remove that CNC from consideration with complete confidence and relief. What if this CNC made it to our floor? Reputaions are won and lost on performance, not opinions.

Choose the way of validation.

Sincerely,

[surface]

I don't think the machine's control was the limiting factor on the machine from China. When it comes to machining performance, accuracy and speed, there will always be a "weak link". CNC machines are a complex array of mechanical and electronic elements, any which could be the root cause of errors.
Scott_bob, in an earlier post, Fadals new Augusta control or 104d, "What is the most significant 'limiting' factor in High Speed Machining on a Fadal", it was the consencus that the largest limiting factor was the machine's control. Now that Fadal is offering 3 different controls, is the "limiting factor" still the control, and like your china-machine, the limiting factor is now mechanical? If however the limiting factor is still the power of the control, then one offering must be better than the others.

[Scott_bob]

I don't think you could possibly make an observation on what was the cause of this "Fanuc controlled" CNC test cut, without seeing. Mechanical factors could not account for the inaccuracies we saw. There was nothing wrong with the program, fixturing, cutter etc. This brings us back to the control that creates the inaccurate motion. This motion I am refering to is circular interpolation. It is disappointing how bad a lot of controls are at doing arcs. This kind of feature is easilly inspected on a machined part. Say a slot .200 wide with full radius of .1 at each end and 1.000 long. The code to produce the R.100 "depending on the end mill used" would be pretty small arcs, say R.0375 for a .125 end mill. On this test cut the programmed feed rate was F78. The material was 6061-T6 .125 thick, the .125 end mill, a 2 flute stub carbide end mill. Now, to most of you this feed rate would be way too high. But we use this "programmed" feed rate all the time and sometimes over F100. for small features on a Fadal CNC with a Rubicon control. The important thing to know is the programmed feed rate is F78. but a good control does not go any faster "than it can" in small radii or at a sharp corner. In this case "the Fanuc" control did have some feed reduction depending on geometry but, lost accurate control of motion due to either tuning, or lack of feedback loop resolution. True, there are many things to go wrong in a CNC machine but, this was a new CNC. I am sure, that a well tuned Fanuc controlled CNC machine could perform well but, I don't believe Fadal Engineering would be qualified to fine tune a Fanuc, they are not all that successful with their own control. Seimens, same thing, Fadal has not shown me the ability to develope a high performance set up. One thing I have learned lately, it is a lot more difficult at the higher performance end of HSM.

Our next test cut: Mori Seiki, Kitamura then ?

[ty1295]

Without knowing Scott particular issues on his circumstance I would still pick Fanuc for one reason and will depend on your situation.

The reason I would pick a Fanuc would be easier to buy future "other" brands with the same controller.

As and engineer at a place with 60 shop floor employees, and one that buys equipment for our facility I have found training of operators to be a bigger hassle sometimes than anything.

We mainly have Okuma's which have their own control, and I have stuck with it for the pure reason of operator training.

A few newer machines, Okuma didn't make have the Fanuc control. (got a 60" vertical lathe on the way)

I actually prefer the Okuma control to Fanuc myself but ........

I think it all depends on YOUR situation.

If it is just 1 or 2 machine shop, get the control you feel most comfortable using, etc. (assuming they all machine the same)

[Scott_bob]

I used to be a Fanuc enthusiast. Until I found a better CNC control... I have spent more time hassling with memory than I care to admit. In the Mold machining Industry, 3D programs are big, step overs are small and high speed motion is critical to success. Without small step overs or tiny scallop height settings in CAM we lose accuracy and rely too much on hand finishing. This is where a CNC control like the Rubicon can shine against so many others. Smooth, accurate high speed motion. Not all CNC HSM is the same. It is shocking to see the performance difference just by changing the control.

I remember the Okuma control; it was good, but out of business today. This is what can happen to a CNC machine builder even as in this case Okuma was a premium CNC machine tool builder.

As for training operators, I hear what you're saying, it is a hassle.
But, I have found the simpler the control, the less hassle you have.
I have evolved in my expectation of the CNC machine control. I used to consider bells and whistles important. Now, what is important to me is accurate, smooth motion control, if it cannot deliver on this expectation, what good are bells and whistles like multi tasking, back ground editing, graphic tool tables, not to mention internet access. I mean I did not buy a CNC machine for that stuff, I use a 400 dollar PC on the bench next to the machine for all that.

If you don't have accurate results from your CNC control, maybe it's time to replace it.

Sincerely,

[Kmed]

I think something that was left out on this was the machine in question that did not do weel witht he fanuc control (coming from my friend shop who useses fanuc and heiednhim (sp) controls) they had to replace the machine, it was nto the control as first thought. The servos could not respond fast enough to generate the hi feedrates required.. What control of fanuc were your tests performed on? Recently in s HSM test the fanuc 18i worked wonders even at 2000 IPM .... I know my control I can change the circular spline fit of the machine and greatly affect he concentricity of a interpolated cirlce arc from lees then .0005 to greater tha .001.. All with one little change in the control parameters..

[Scott_bob]

Download, modify for your needs...

So many benefits, Video file to follow...
You just cannot believe the smooth, accurate CNC motion till you see it for yourself.

Good luck,

[nervis1]

Hey you really shouldnt be in there without glasses.


Cool vid, little too short to really get a feel for it though.

[Scott_bob]

Cool vid, little too short to really get a feel for it though.

I could not get a bigger file uploaded...
I wanted to upload a 250 mb file in the "Downloads" area but could not...

Notice the 2 set screw holder which is too long - not ideal
Major improvement by using a stub, or better yet shrink fit...
Did not not stop us from going 400 to 500 IPM though...

I only wish you could see this in high resolution and full length.
This coolant manifold has 6 high pressure nozzles, & we upgraded to 100 PSI.

[nervis1]

Bummer, I'd like to see the whole thing. I love that stuff. I'm sure others would like to see it too.

I'll go and ask the boss.

[nervis1]

Pretty cool. Thanks for taking the time to post that up. That coolant system sure looks like it gets the job done .