[narf04]

Hi all,

I am having a tough time getting this to work and not sure where else to turn. We purchased fixture tracking (CALL OO88) for our horizonal MB-5000H with OSP-P300M. I have Renishaw probing marcos working correctly from our CAM software and am trying to get the probing results on one face of the tombstone translated to other faces.

Here is a description the test file i have setup. We are programming from center of pallet with autosetting on the 0deg face. Then we index to face 210 and probe a part in the X axis while holding the part 1" to the right of where programmed. We then drill a hole in the center of this face. This hole is drilled in the correct position. Then, the portion we are having trouble with...we index +/- 90 deg (300 and 120 deg face respectively) and mill a contour. I can't get the results probed in x to translate to these index moves. I will add a photo of part and results along with the program most currently tested.

Does anyone have any experience with this? I've been working with our distributor but they have not done this before either. Any help much appreciated.

thanks, Nick

[deadlykitten]

hi, if i get it right, when you index at 210, the part face is paralel to xy plane ? and in this position you drill it

after that you mill a shoulder, by indexing it at ±90*; but isn't it possible to mill that shoulder without rotating ±90* ? the cam image is <> photo

... photo shows U groove
... cam shows L groove; L groove does not need 90* rotation

i guess you wanna probe the milling operation, so to be sure that size is correct, and that it is centered to the drilled hole ?


i need to know exactly what points you wanna probe, and what you wish to determine, based on those points : it imay be possible to read some system variables, that stores the output of the probe, and perform the math inside the program / kindly

ps : i have no clue what oo88 does ...

[narf04]

hi, if i get it right, when you index at 210, the part face is paralel to xy plane ? and in this position you drill it

after that you mill a shoulder, by indexing it at ±90*; but isn't it possible to mill that shoulder without rotating ±90* ? the cam image is <> photo

... photo shows U groove
... cam shows L groove; L groove does not need 90* rotation

i guess you wanna probe the milling operation, so to be sure that size is correct, and that it is centered to the drilled hole ?


i need to know exactly what points you wanna probe, and what you wish to determine, based on those points : it imay be possible to read some system variables, that stores the output of the probe, and perform the math inside the program / kindly

ps : i have no clue what oo88 does ...

This issue isn't about how to machine the part. I know i can machine that feature without indexing. This is a test file for probing in one orientation and then passing those results to different orientations. CALL OO88 is the reason for this thread and what i'm trying to figure out. As i understand it, the macro should update our index moves without having to use multiple offsets for each orientation. I am not able to find any examples on how to program with it and use probing. Here is a simulation of the job. as you can see in the results photo, the cut is deep and right of where it should be.
https://autode.sk/2vn9mfF

[deadlykitten]

hi, i think i got it : you wish to avoid multiple offsets by using oo88

you will have to deal also with the restart procedures, thus avoiding probing during a restart sequence, and still be able to activate last work coordinate system

i have done this, by using the soubroutines that came with the machine (vmc), and by digging some system variables, and writing/reading data from files, so to calculate the rotation of a part that was clamped at a random angle


in your case, if you wish to update the index movement, based on a probing sequence, then you must define the probing points in such a manner, that they will touch among a surface that is not paralel to xy axis : your part and those operations are all paralel to the axis

please show/share the points where you wish to probe, and the direction among which the probe is moving toward the part / kindly

[narf04]

hi, i think i got it : you wish to avoid multiple offsets by using oo88

you will have to deal also with the restart procedures, thus avoiding probing during a restart sequence, and still be able to activate last work coordinate system

i have done this, by using the soubroutines that came with the machine (vmc), and by digging some system variables, and writing/reading data from files, so to calculate the rotation of a part that was clamped at a random angle


in your case, if you wish to update the index movement, based on a probing sequence, then you must define the probing points in such a manner, that they will touch among a surface that is not paralel to xy axis : your part and those operations are all paralel to the axis

please show/share the points where you wish to probe, and the direction among which the probe is moving toward the part / kindly

I appreciate you looking at this, but i don't think we're on the same page. I'm not looking for workarounds. I should be able to probe as shown in the video and then index/machine.

[deadlykitten]

hello again / in that video probing occurs at 0:29; each face is touched, middle point is used for driling, and work coordinate for milling can allready be calculated


do you need a soubroutine, that will probe 2 points, compute the middle and the work coordinates for both sides ?
* initial data : Z0, target part width, and a fixture with a maximum part shift that can be tolerated by the soubroutine ( you said 1" right )
* output data : center_point and 2 work coordinates, one for each side face

the soubroutine needs to calculate the excentricity and the distance between part face and table center, so to deliver positioning for each side ( check attached image )


... is not hard; can you read, inside the program, the coordinate of the probed point ? if yes, than consider it ( almost ) done if no, i may guide you to find the system variable that stores the answer
... also you need to know the location of the table center / kindly

[deadlykitten]

search oo88 on this forum; check also attached pdf ( shared by superman ) & maybe your pivot distance

I can't get the results probed in x to translate to these index moves

oo88 does not take as input the values from the probing, but from the work coords system, thus is possible to rotate the work coordinates values, regardless of the method used to zero the part

please consider that is possible to create a soubroutine that does what you want; even simpler for your machine, since there is only 1 rotary axis

if you wish to check the output of your current oo88, simply compute the math for rotating the origin and compare the results with the oo88; there may be a clue revealed by those differences / kindly

[deadlykitten]

hello again / if your part is drilled to center, and the millings are simetrical, then why oo88 needs to shift the X ? " CALL OO88 XX = 0.332 "

xx is 0 for probing, but it is <> 0 for cutting, and this happens inside the cam generated program

i would look into this on its own, without relation to oo88 behaviour, so to see what is causing the x shifting


in the end, i would test oo88 on a simple single point, and when it works, i would add the probing sequence / kindly

[narf04]

hello again / if your part is drilled to center, and the millings are simetrical, then why oo88 needs to shift the X ? " CALL OO88 XX = 0.332 "

xx is 0 for probing, but it is <> 0 for cutting, and this happens inside the cam generated program

i would look into this on its own, without relation to oo88 behaviour, so to see what is causing the x shifting


in the end, i would test oo88 on a simple single point, and when it works, i would add the probing sequence / kindly

The XX = 0.332 is the amount probed in X at the 210 angle. I thought this would be needed to get the correct location once we index to the milling. I have tested without CALLOO88 altogether and got the same results as the pics above. I tried with leaving XX = 0, same results. everything i've tried so far is milling to deep or shallow (depending on side) from what is programmed.

[deadlykitten]

program coordinate for drilling is at x0, and if it is executed in the center of the part, it means that the probing sequence puts x0 at the middle, and this is ok

the middle of the part is excentric, in respect to the table center, and this excentricity has to be passed to the index sequence, so to raise / lower the milling positions

i don't know what is behind the oo88, but i can create a soubroutine that will shift the coordinates in a much shorter time than what it takes me to dig the oo88; problem is writing the soubroutine in a manner that makes it compatible with your actual cam system, etc, so to have it there available anytime, for whatever setup

i have done this for my setups ( different fixtures, different machine, but same idea ), and i believe i could do it also for your setup, but i need to be near the machine ... or simply share the input+output parameters that you wish for, and i may code it for you; or share actual numeric values ( before and after rotation ), and desired numeric values ( before and after rotation ) and maybe i will make a wild guess and deliver a nice code

in this thread, at post 14, is a soubroutine that is ( somehow ) equivalent to oo88; maybe it will inspire you : https://www.cnczone.com/forums/okuma...ro-needed.html

i know that my answer is volatile, but this is all that i can do / kindly

[stevemarkovick]

This thread really helped clarify how fixture tracking interacts with indexed probing, especially on the MB-5000H. I’ve faced similar issues with orientation consistency across faces, and it’s great to see detailed back-and-forth like this. Having a robust solution like a custom subroutine makes a huge difference. For tracking accuracy in other industries, I’ve found Phonexa surprisingly reliable.