What's new
What's new

Cycle 800 on DMU, is part need to be perfectly at the center table ?

datletien

Plastic
Joined
Mar 10, 2023
Hi all,

I have a question about cycle 800. Do we need to locate the part perfectly at the center of the table in order for cycle 800 work correctly when rotating the table? I have my located part off the center table a bit within .005 -.008 and when the table rotate A-90 and C0 to drill a simple hole. The hole location is off .005 on X and .008 on Y. We did the test cut that we turn the table A-90 and C0, take the workoffset from the part and drill the hole without using cycle 800, the hole is precise to its true location. Before I thought the cycle 800 will be “smart” enough to calculate all the workoffset transformation even though the part off the table center, but I guess apparently it is not. Now we have a part that needed to mill on different faces and everytime we turn the table we need to take a workoffset again. I believe this is not how it is supposed to be with a 5 axis machine. Could anyone advise me on this issue ?

Thank you
 
Last edited:
Hi, what control on which DMU? What year, and what software version? Millplus, Heidenhain, Siemens, Fanuc (Unlikely but I've seen it)?
 
It is DMU 50. We use Esprit Cam software and the control is Siemens.
 
Last edited:
It is DMU 50. We use Esprit Cam software and the control is Siemens.
You certainly should be able to find your XYZ zero with A and C at 0 degrees, then run cycle800 to rotate A and C for side access and have an accurate workpiece coordinate system. That is the most basic killer-feature use-case for cycle800. If your probe and kinematics are calibrated correctly, it should reliably work.

If time allows, on the machine itself, I would go into MDI mode or write simple test case program that does nothing but run the cycle800. Then, jog to the faces of the block and verify that their coordinates make sense.

It's quite possible your post's expectations of how to manipulate the coordinate system with cycle800 aren't perfectly in agreement with what the control actually does.
 
If you are using CYCLE 800, the origin point can be anywhere on the table, and it will track it correctly. The only thing that is important is that your origin point is in the right spot relative to your part. For instance, if XYZ 0 is the top center of stock, then it doesn't matter where you have the stock located on the table, as long as the work coordinate system has the *actual* top of stock that you loaded set as 0,0,0. We do vise work on rare occasion, and the vise will often be off position by .125in or more in any given direction. The whole point is that with CYCLE 800, it doesn't matter.

However, in order for CYCLE 800 to work its magic, your center of rotation does need to be calibrated correctly under the kinematic table. Are you doing anything to verify and maintain accuracy of the COR?
 
So far we haven’t calibrated the machine for the pass 6 months. However, we dont have any issue with quality until 3 weeks ago. We started noticing that the true location of holes was off .005-.008 and parts were failed CMM inspection. So you are saying that I need to check with DMU Mori for machine calibration ? Also, other opinion said that if you put part off center table, your part should be put off center table in Esprit software too. It is also confusing me about that. In the software simulation we always program from center table, but in actual real life, part can be off a bit from center table too. So do we need to match the kinematics model between the software and the actual set up at machine ?
 
So far we haven’t calibrated the machine for the pass 6 months. However, we dont have any issue with quality until 3 weeks ago. We started noticing that the true location of holes was off .005-.008 and parts were failed CMM inspection. So you are saying that I need to check with DMU Mori for machine calibration ? Also, other opinion said that if you put part off center table, your part should be put off center table in Esprit software too. It is also confusing me about that. In the software simulation we always program from center table, but in actual real life, part can be off a bit from center table too. So do we need to match the kinematics model between the software and the actual set up at machine ?
Ah, 6 months is a while to go without running calibration, aka Quickset 3d. In that amount of time, .005-.008 inaccuracy building up for cycle800 and traori is normal enough, particularly if the ambient temperature has changed significantly. Definitely run quickset 3d on your A and C axis.

I think quickset3d should clear up the issue, especially if cycle800 used to deliver perfect results for you. It shouldn't be necessary for the CAM setup to exactly match the part location on the table, if cycle800 is working correctly. If cycle800 isn't working correctly, it is possible to use neither cycle800 nor traori, but to fall back on exactly matching the CAM and part setup, though it's a huge pain. You really want to let the machine do cycle800 and traori computations.
 
So far we haven’t calibrated the machine for the pass 6 months. However, we dont have any issue with quality until 3 weeks ago. We started noticing that the true location of holes was off .005-.008 and parts were failed CMM inspection. So you are saying that I need to check with DMU Mori for machine calibration ? Also, other opinion said that if you put part off center table, your part should be put off center table in Esprit software too. It is also confusing me about that. In the software simulation we always program from center table, but in actual real life, part can be off a bit from center table too. So do we need to match the kinematics model between the software and the actual set up at machine ?
As long as you adjust your work offset to match the position of the part, you do not need to shift the model in CAM.

Without checking center of rotation you really are flying blind. It's easy to find center of rotation of the C axis in X/Y by putting an indicator on the table and a test bar in the spindle. Once you find the center of the C axis, you can compare to the X/Y for COR in the kinematic table, and that will give you a rough idea if it is really bad. If the values are off by more than .0025in, then basically everything is suspect, and the machine needs a full recalibration.

If you have 3D quickset, check your probe calibration, and then just run it (quickset). Nothing to it.
 
Thanks a lot Erikh and Boosted for all your great advices. Today, we did a machine calibration and drilled some test hole with A and C axis rotated. It seems the machine gain back its accuracy. Location of test holes were within the position what we programed (X position is exact and Y is off about .001 only). Now I had a better understanding about cycle800 as it is only based on the work-offset point , not where the part located on the machine table.
 








 
Back
Top