Integrex B-Axis center of rotation (E-410)

[MazatrolMatrix]

Hello!

So we are trying to adjust the rotation center because Z0 is different in B0 vs B90 by 0.2mm. I know you adjust by manipulating BA61 and BA62 parameters. But how do you go about in getting the right amount to shift in each parameter?

We have two indicators set up to Zero in both axis in B0, then we switch to B90 and look up the value difference in the machine zero coordinates and so some calculations. However we seem to be doing it wrong still since the values still dont match. And yes the B Axis is straight.

 

[noname777]

what is the calculation? formula?

 

[MazatrolMatrix]

I solved it. Turns out that like everything else it isn't really difficult. Just that the information regarding how it's done is scarce or people are unwilling to share it.

 

[Nagol]

He says while not sharing the information...

 

[noname777]

it is a very basic calculation. Basic geometry. What is secretive here? You have to plus few digits and after divide them few times.

 

[MazatrolMatrix]

it is a very basic calculation. Basic geometry. What is secretive here? You have to plus few digits and after divide them few times.

As simple as it may be when you know how it's done, your explanation there is not quite correct.

Edit:
The picture is correct though.

 

[MazatrolMatrix]

He says while not sharing the information...

I will.

 

[noname777]

I can add more pictures ( besides one I did upload few days ago, above) where everything is explain even better

 

[MazatrolMatrix]

I can add more pictures ( besides one I did upload few days ago, above) where everything is explain even better

No, it’s ok. I know how it is done. Thanks

 

[noname777]

okay. As long as it is NOT a top secret anymore

 

[MazatrolMatrix]

okay. As long as it is NOT a top secret anymore

For others it may be, I thought I'd make a guide with pictures and everything when I have the time. Along with a step 1. xx 2.xx and so on. Mostly for the company I work for but might as well post it here.

 

[noname777]

we have already all this information available. When people ask, we help them. That is why we all are here in a first place. To support MAZAKs guys

 

[MazatrolMatrix]

we have already all this information available. When people ask, we help them. That is why we all are here in a first place. To support MAZAKs guys

Hi noname,

We have adjusted the rotational center now but when drilling or milling in B90, the holes position is still 0.25 wrong in the Z(X) axis in relation to the turned face of the part.

Would it be possible that what's wrong is some kind of Zero point shift parameter instead? BA64 or BA65, maybe.

 

[Nagol]

Not sure how to verify or how relevant it is because as I have stated I'm a mill guy. Tool length plays a big role in these calculations on a mill. Maybe you can verify your automatic tool measurer accuracy by turning a face of a part and check to your chuck with a gage block?

 

[MazatrolMatrix]

Not sure how to verify or how relevant it is because as I have stated I'm a mill guy. Tool length plays a big role in these calculations on a mill. Maybe you can verify your automatic tool measurer accuracy by turning a face of a part and check to your chuck with a gage block?

The tool eye is measuring wrong also but we know the length from spindle center to it's nose is ~ 250 mm, then the adjustment tool we use is "exactly" 65mm precision ground. So we know the values there. And since every tool is measured from the tool eye in our case they should still be correct in relation to each other.

The tool we use to do this only has a radius, the other side of it has a flat instead. I have assumed that this flat is exactly at the 0 point of the tool, thus the tools radius could be measured from the radius to the Flat. But perhaps they don't meet at the tools center, and the distance from the radius of this tool to spindle center could be 45.2 instead of 45, for example.

Maybe this can cause the calculation to be wrong. At a quick thought the rotation center would be off by the amount that deviates from the presumed value of 45..
I have a hard time believing this shouldnt equal exactly 45 on a tool such as this though..

 

[noname777]

what Mazaks service people say about this issue? Apart from MAZATROL parameters you are allowed to change, our geniuses from Japan have heaps of PLC parameters for tool path compensation.... and most of the time ordinary humans have no access to this information.

 

[MazatrolMatrix]

what Mazaks service people say about this issue? Apart from MAZATROL parameters you are allowed to change, our geniuses from Japan have heaps of PLC parameters for tool path compensation.... and most of the time ordinary humans have no access to this information.

We aren't particularly impressed by them, generally they have no idea. In fact it's them that have calibrated the machine to this error, somehow. We crashed it and they installed a new Milling spindle, and after that COR was wrong.

 

[noname777]

heard it before. Send me your email in PM, I will send you something one person gave me ages ago. You just double check it against what you are doing atm.

Ideally you should follow these steps, cannot skip them.
1. check machine level
2. check if upper turret is mechanically square in relation to main spindle ( in all 6 planes B0,B90,B180)
3. Compensate Y axis when B0 goes to B180
4.correct all X,Z,Y axises home positions by clocking empty mill holder in milling spindle.
5.Set your tool eye, preferably with test bar, not just OD turning.
6. Make sure machine is a perfectly square mechanically ( you do it while all software compensations are set to ZERO), and only after, you try to compensate irregularities with parameters. ( in general )

 

[noname777]

I doubt BA64 or BA65 is the problem. They just determine a single point in space. Your problem is when you B0 to B90.

If I was you I would be checking if the turret mechanically square, after removing all BA compensations, checking what the error was.
After this I would be using big numbers for BA61 and BA62 to see if they actually have impact on rotational axis compensation, and if they do impact and you can see it, only than I would be calculating them to the right amount.

 

[MazatrolMatrix]

I doubt BA64 or BA65 is the problem. They just determine a single point in space. Your problem is when you B0 to B90.

If I was you I would be checking if the turret mechanically square, after removing all BA compensations, checking what the error was.
After this I would be using big numbers for BA61 and BA62 to see if they actually have impact on rotational axis compensation, and if they do impact and you can see it, only than I would be calculating them to the right amount.

This is a good idea. It seems the machine moves but the 0.25 deviation is there still.

I'll let you know if I solve it!