What's new
What's new

Mikron touch probe not agreeing with manual bore pickup

70olds

Aluminum
Joined
Jun 21, 2016
Hoping one of the Mikron experts can chime in here. We have a 450U machine and I am trying to verify center of rotation on my pallets. I took and endmill and plunged a small bore and then spun the C axis (after verifying my planes were flat without the pallet on the machine) to make a centralized bore. Next I took the probe (which was calibrated) and probed center of that bore using the "Probe CC" function in manual mode. If I then remove the probe and put an indicator on the face of the spindle with amg base and swing the center I am consistently off .0009" in Y and .0002" in X. I have tried creating a new work offset there and probing again and find the same problem. I mounted my lang plate and used the ground center on that to repeat the test, same result. I have re-calibrated the probe in manual using the "CAL R" function, and at the end of the calibration when i pick up the bore I get the same error. Is there a way to manipulate the values from the calibration? Am I doing something wrong? Probe has maybe .0002" runout in the tip, which appears to be correctly accounted for when the calibration cycle is finished.Thanks.
 

read through this thread.
 

read through this thread.
I read through that before. I'm not trying to set any kind of crazy offsets in the kinematics or even considering anything other than X, Y at this point with stationary B,C. For sake of argument let's say my machine is 3 axis. Let's also exclude me even cutting anything.

Imagine I put a 2" ring gage on my table and I manually pick up with an indicator by swinging it in.
Then I bring out my calibrated probe and pick up the same ring gage. it is off .0002" in X and .0009" in Y.
I run the calibration on the probe and pick up the ring gage again, same problem. Can I manually fix my probe pickup positions to agree with the ring gage or am I missing something in what I am doing?
 
Are you manually removing the probe and then putting the probe back in the same orientation to the spindle and not 180* out?
If its calibrated at 0* and then you accidentally reinstall it at 180* it will have 2x the error it is compensating for.
 
Are you manually removing the probe and then putting the probe back in the same orientation to the spindle and not 180* out?
If its calibrated at 0* and then you accidentally reinstall it at 180* it will have 2x the error it is compensating for.
The machine is HSK-63. It only goes one way. Either way I am calibrating it and checking it again without even removing it.
 
I know nothing about Mikron but, agree that whatever is going on does not seem to be due to kinematics of the machine. I was going to ask the same questions. Have you tried locating zero, boring a hole at zero (boring head) and then probing that location? That should be dead-nuts zero if the probe is working mechanically and the calibration stuck.
 
Have you also ruled out indicator error? Check with two indicators and make sure they agree.

Shaft wear in a dial indicator or loose pivot bearings in a test indicator could potentially cause enough lost motion for those kind of numbers.

What is the resolution of the indicator that you're using? Depending on the resolution and size of the scale, parallax error could also account for that kind deviation.
 
Are the probe and the indicator setup close in gage length? If the spindle center line is not parallel to the Z axis then gage length differences appear as XY position errors between two measuring devices.
And we have a winner.
I set my indicator at 3 different gage lengths and swung the same hole without moving anything but the Z axis.
Got the following numbers:
u450.jpg
To take it one step further I got my indicator as close as I could to the gage length of my probe and picked up a bore, then probed said bore, and I was out 80 millionths (.00008) in y and only 30 millionths (.00003) in X.

Moving on from there I wanted to verify if the head was tilted or the rails were tilted. I found that even at two different heights when swinging the pads that the pallets sit on all readings were under .0001", so the spindle axis appears to be square to the table, but perhaps the rails themselves are not at a perfect angle. I have a cylinder square to check the machine with that should give me the final answer but it appears that is the case so far from what I am seeing. I will update once I get that on there.
 
And we have a winner…….

What is the crash history of the machine? I don’t know Milton’s design and build techniques but having the spindle axis good to the table and suspecting rail or blocks being out strikes me as unusual unless somewhere along the way the table zero was shifted to get the spindle to tram well.
 
What is the crash history of the machine? I don’t know Milton’s design and build techniques but having the spindle axis good to the table and suspecting rail or blocks being out strikes me as unusual unless somewhere along the way the table zero was shifted to get the spindle to tram well.
why is it unusual? they are 2 different components and as such can and do have their own individual adjustments.
 
I unfortunately don't know the history. We bought the machine used and I am just getting into it.
 
why is it unusual? they are 2 different components and as such can and do have their own individual adjustments.
When the machine was built, the spindle and Z were probably parallel to each other and perpendicular to the table, X, and Y. Now the OP reports the spindle is still perpendicular to the table yet the gage length error suggests that Z is no longer perpendicular to X and Y. Unless some adjustment was made to the table, one would expect to see a table tram error reflecting the non-perpendicular condition. Shifting the table to get the tram good is not the proper fix for an issue with the basic geometry of the machine.
 
When the machine was built, the spindle and Z were probably parallel to each other and perpendicular to the table, X, and Y. Now the OP reports the spindle is still perpendicular to the table yet the gage length error suggests that Z is no longer perpendicular to X and Y. Unless some adjustment was made to the table, one would expect to see a table tram error reflecting the non-perpendicular condition. Shifting the table to get the tram good is not the proper fix for an issue with the basic geometry of the machine.
the spindle can be adjusted to be perpendicular to the table independent of the z axis rails.
 
the spindle can be adjusted to be perpendicular to the table independent of the z axis rails.
When speaking to a Mikron service tech, he told me the same thing and that this could have been done from a crash where a half-assed alignment job was done to say "yep the tram is good, so it's aligned".
 
how else are you going to ensure both are perpendicular?
I guess we must be misunderstanding each other…..

The OPs issue is because his spindle center line is not parallel to the Z axis. It appears that some misguided attempt to “align” the machine led someone to adjust the table to be perpendicular to the spindle centerline rather than realign the Z axis to be perpendicular to X and Y axes. The OP noted that a Mikron tech told him as well that his scenario sounds like “a half assed alignment job”.
 
I guess we must be misunderstanding each other…..

The OPs issue is because his spindle center line is not parallel to the Z axis. It appears that some misguided attempt to “align” the machine led someone to adjust the table to be perpendicular to the spindle centerline rather than realign the Z axis to be perpendicular to X and Y axes. The OP noted that a Mikron tech told him as well that his scenario sounds like “a half assed alignment job”.
misunderstanding indeed. you dont align the table to the spindle, or to the z axis, typically the spindle and/or z axis rails would get aligned to the table, but the spindle axis and z axis rails are 2 completely independent components that require its own alignment. yes, one of them was improperly aligned from what i'm gathering here.
 
I guess we must be misunderstanding each other…..

The OPs issue is because his spindle center line is not parallel to the Z axis. It appears that some misguided attempt to “align” the machine led someone to adjust the table to be perpendicular to the spindle centerline rather than realign the Z axis to be perpendicular to X and Y axes. The OP noted that a Mikron tech told him as well that his scenario sounds like “a half assed alignment job”.
I was not saying the table was adjusted to match spindle centerline, but rather the spindle centerline was adjusted independently (and no longer parallel to) of the Z axis rails to match the table. Either way, all 3 components need to line up.
 








 
Back
Top