G54.2 vs G54.4 for HMC

[SlaveCam]

Shop's got some new four axis horizontals that are equipped with both options. I want to use either option to compensate for the mismatch of workpiece position to the true position of part attached to a tombstone and when the pallet rotates, the correction vector would rotate too. With the older machines I've had to use multiple work offsets and calculate the correction for each B position using a macro.

If I've understood correctly, G54.2 is only able to correct X/Y/Z while G54.4 can also correct the rotation of the part on tombstone's face along any axis. Because the machine's got only four axis, the correction can only be applied to two axis instead of three.

Is there anything G54.4 cannot do that G54.2 can? Which is easier to incorporate into post? G54.2 seems very simple and I rarely need to take the incorrect orientation of part into account. All I need for now is to probe the true position of part and use the result for correction.

[cncappsjames]

G54.2 (Rotary Table Dynamic Fixture Offset - RTDFO) and G54.4 (Work Setting Error Correction - WSEC) are two VERY different functions and should not be confused or even thought of as similar in any way, shape or form.

If anything, G54.2 and G68.2 could be thought of as "mostly" comparable. Each one's setup/usage is much different however. One needs a separate offset table (G54.2) the other(G68.2) uses standard offsets. Anyway, I say mostly because the behind the scenes functionality is similar though G68.2 is FAR more powerful than G54.2.

In G54.2, you use either your Common Offset or your Work Offset to set center of rotation position (for X and Z. Y isn't required) and you put the distance form COR to your part origin in the RTDFO table. In G68.2 you have your center of rotation position in the #19700 and #19702 parameters and just pick up an offset like you would on any machine. Control does all the calculation as you rotate around the part.

G54.4 you use to correct for small (preferrably) errors in position, orientation and rotation relative to the part's expected position, orientation and rotation. In an HMC you can only correct for theoretical axis (A and C on a typical HMC) when the face of the part is presented to the spindle. Rotating from that, because the head/table does not tilt you can't comp for it. But straght on, it will function much like coordinate rotation (G68).

G54.4 (like G68.2, G43.4, etc... require your COR to be set in #19700 and #19702 at the very least. Typically on an HMC, you'll set #19700, #19701 and #19702 (COR X, Y, and Z respectively)

As far as ease of integration into the post, G54.4 wins because you don't have to do any plane matrix stuff with it, whereas with G54.2 or G68.2 you could need to do some work there.

Hopefully that makes sense.

[Colin Gilchrist]

James, what do you think about just going with G68.2 on a Horizontal, versus the G54.4, since the machine doesn't have two rotaries to compensate?

Would you just go G54.4, and call it a day?

[SlaveCam]

Thank you for a detailed answer. I'm not 100% sure I fully got it yet, but I'm gonna make some tests with the machines to see how each option behave. Best way to learn is just do. Also, the control is MAZATROL so I have to figure out the corresponding macro numbers but "they" say both options should behave identically to Fanuc's

[cncappsjames]

I would go with G68.2 all day every day on a horizontal. I think it's easier than G54.2 personally... not that G54.2 is hard to use because it's not.

G54.4 on an HMC... case by case basis but G54.4 and G68.2 use the same kinematic parameters. If I was doing castings, or a lot of rework, I'd seriously consider it. On an HMC it's almost overkill because unless the part face is presenting to the spindle, you can't use the theoretical axes for much. 

[SlaveCam]

Shouldn't G54.2 be the easiest of all? "Many" use it with mcam without any post modifications (from what I've read on different forums), as the work offset simply rotates with every B-command and all you have to do in mcam is rotate the plane (origin stays in place).

Anyways, I've been experimenting G54.2 on SmoothG control and already losing my patience. Why does G54.2 perform axis motion?? Is it same on Fanuc or another Mazatrol idiocy?

 

[huskermcdoogle]

On 4/14/2023 at 4:23 PM, SlaveCam said:

Shouldn't G54.2 be the easiest of all? "Many" use it with mcam without any post modifications (from what I've read on different forums), as the work offset simply rotates with every B-command and all you have to do in mcam is rotate the plane (origin stays in place).

Anyways, I've been experimenting G54.2 on SmoothG control and already losing my patience. Why does G54.2 perform axis motion?? Is it same on Fanuc or another Mazatrol idiocy?

 

I believe G54.2 on a Mazatrol is more akin to G43.4 on a Fanuc.  That is old corrupted memory..., but as I recall that's the case.