G234 HAAS TCPC Help

[StrikeQ]

We're a small shop that does Impeller work. We got a new VF2 TR160 machine that came with TCPC/DWO. None here are familiar with it and we're in the process of getting our post finalized. There's been some back and forth as we're learning we have to discard some of our old practices (G93 etc).

Essentially we're on one last hang up. We like to position XYAC before doing anything with Z. Here's a snippet of some GCode

G00 G17 G20 G40 G49 G80 G90 G94
G91 G28 Z0. M05
N1
(ROUGH ALL)
(OPERATION NO - 2)
T13 M06 (5/32 TIP TAPER MILL)
S12000 M03
M11
M13
G187 P1
G00 G90 G55 X-1.7192 Y2.1279 A85.615 C33.991
/M08
G234 H13 Z3.8327
G94 G17 G01 X-1.423 Y1.6886 Z3.7921 F400.
X-1.2 Y1.3579 Z3.7615 F30.
X-1.177 Y1.3778 Z3.6665 A84.862 C32.462
X-1.1716 Y1.3826 Z3.6494 A84.575 C32.152
X-1.1639 Y1.3892 Z3.6323 A84.269 C31.823
X-1.1558 Y1.3962 Z3.6186 A84.003 C31.536
X-1.1429 Y1.4073 Z3.6014 A83.631 C31.135
X-1.0925 Y1.4491 Z3.5496 A82.341 C29.743
X-1.0279 Y1.5003 Z3.4874 A80.742 C28.017

I hand modified the G234 H13 to be below the /M08 (We also want coolant on before Z moves as there is a delay and if not it will start cutting before coolant actually flows). By default in the current post it's on the same line as the G00 G90 G55.

The problem is the X-1.7192 Y2.1279 are not accurate. They are based off the G234 but since it hasn't been called yet the machine moves the tool somewhere over the trunnion (for this program). Then when the G234 is called it does a 3 axis positioning to the proper place. Bringing Z down close to the trunnion/part while XY are mvoing. Which is what we want to avoid, only Z should move after everything else is initially positioned. If I leave the post default with the G234 on the same line as the other 4, Z comes down near the tool setter, trunnion, and part as it wraps around. In this case it's all avoided, but I'm not sure how it'll be with a smaller part that requires Z to move down more etc. It's something we'd just like to not have to worry about.

If I try to move the G234 H13 (leaving Z where it's at) above the G00 G90 G55, or even on the same line just with Z by itself still I get a Y Axis over travel alarm.

I have the post people working on this too but it's been slow and the few responses I've got are leading me to believe they can't do what we want.

[crazy^millman]

Ref points or point toolpaths are you friend here. No need for a post edit if approached differently with regards to G234. Create a travel limits shape and use it for your reference when programming. If any approach move is inside that travel limits then you know you are not setting your ref point high enough. The module works toolpaths can do this, but what a pain it is to do so. When not made by the creator you get different things. In those cases you will want to use a point toolpath to establish a safe distance to give the output code for the G234. 

Other things are easy to adjust. Use the post debugger and go from there.

 

HTH(Hope that Helps)

[Brian@PC]

You need to rotate to your desired angle, activate G254 and then call  your X/Y location. Then call G255 to deactivate DWO. Now you are ready to call up G234 to do your simultaneous operation. On larger machines with smaller trunnions, it is not as big of a problem. It is a safer practice to have the program call your initial rotation up, activate DWO for your preposition and then cancel DWO before TCPC. 

"The problem is the X-1.7192 Y2.1279 are not accurate."

This is not true, it is because the position is from A0/C0 and you are not at A0/C0, you must activate DWO so the position is reflected to the rotated point. 

 

 

 

[crazy^millman]

6 minutes ago, Brian@PhillipsCorp said:

You need to rotate to your desired angle, activate G254 and then call  your X/Y location. Then call G255 to deactivate DWO. Now you are ready to call up G234 to do your simultaneous operation. On larger machines with smaller trunnions, it is not as big of a problem. It is a safer practice to have the program call your initial rotation up, activate DWO for your preposition and then cancel DWO before TCPC. 

"The problem is the X-1.7192 Y2.1279 are not accurate."

This is not true, it is because the position is from A0/C0 and you are not at A0/C0, you must activate DWO so the position is reflected to the rotated point. 

 

 

 

Why wouldn’t a safe height using ref points not work from you experience on the HAAS? I have not run one in years and curious what is the better approach here?

[Brian@PC]

This is from a UMC with the IHS post, different rotary configuration but the general idea. 

 

(MILL ANGLE ON OUTSIDE)
M31
(T2     - 1/2 FLAT ENDMILL     - H2     - D2     - DIA .5")
N201 T2 M06
G00 G17 G90 G54
S6500 M03
M11 (C-AXIS UNLOCK)
M13 (B-AXIS UNLOCK)
C0. B-12.
G254 <-Activate DWO
X3.3023 Y2.1369 <- Preposition 
G255 <- Cancel DWO
G234 H2 X3.1969 Y2.1369 Z.8427 <-TCPC call
X3.5088 Z-.6246
G94 G01 X3.592 Z-1.0158 F80.
X3.5038 Y2.1813 Z-1.0538 F30.
X3.412 Y2.2163 Z-1.0933
X3.3175 Y2.2415 Z-1.1341
X3.2211 Y2.2568 Z-1.1756
X3.1237 Y2.2619 Z-1.2176
Y2.2568
Y2.2516
Y2.2465
Y2.2414
Y2.2362

[StrikeQ]

I appreciate the responses so far. I'll admit I haven't tried too much on the programming side for a solution, but the few things I did try with regards to where I approach from have resulted in Over Travel alarms.

As for Brian's advice I've tried both the following:

 

G00 G17 G20 G40 G49 G80 G90 G94
G91 G28 Z0. M05
N1
T13 M06
S12000 M03
M11
M13
G00 G90 G55 A14.95 C344.777
G254
X1.504 Y3.1845
G255
/M08
G234 H13 Z2.5729 
G94 G17 G01 X1.4852 Y3.1153 Z2.3045 F400.
X1.4581 Y3.0158 Z1.9181 F30.

G00 G17 G20 G40 G49 G80 G90 G94
G91 G28 Z0. M05
N1
T13 M06
S12000 M03
M11
M13
G254
G00 G90 G55 X1.504 Y3.1845 A14.95 C344.777
G255
/M08
G234 H13 Z2.5729 
G94 G17 G01 X1.4852 Y3.1153 Z2.3045 F400.
X1.4581 Y3.0158 Z1.9181 F30.

Both still do XYZ movements when the G234 is called

[Brian@PC]

Quote

Why wouldn’t a safe height using ref points not work from you experience on the HAAS? I have not run one in years and curious what is the better approach here?

It is fine if you are programming Center Line of Rotation. But, you lose the associativity when programming from any other point, as the ref point will be in X/Y coordinates in the non rotated view. If a standard retract is used with DWO it will follow the part, if your fixture height changes. 

One note, Haas TCPC and DWO does have differences in the rules and commands to that of a FANUC.

Also, with the new NGC software release Rapid to feed moves is now an option on all 5 axis rapid moves. 

[crazy^millman]

19 minutes ago, Brian@PhillipsCorp said:

It is fine if you are programming Center Line of Rotation. But, you lose the associativity when programming from any other point, as the ref point will be in X/Y coordinates in the non rotated view. If a standard retract is used with DWO it will follow the part, if your fixture height changes. 

One note, Haas TCPC and DWO does have differences in the rules and commands to that of a FANUC.

Also, with the new NGC software release Rapid to feed moves is now an option on all 5 axis rapid moves. 

Sorry lost me here with this comment “In the non rotated view”. If you have a toolpath done in a different view and make the ref points associated to that view then your coordinates should be rotated in relation to that. Where I said have travel limits in relation to your part in its relationship to the machine as a level. Then you you can always make sure you never have a move outside of that travel limit. Thing about Mastercam limits with 5 Axis programming is non kinematic awareness for any toolpath. That responsibility falls back on the programmer. Putting the correct things in places to make up for this lack of awareness is part of the job. Ideally the post builder can connect to the machine definition and check some things, but with DWO and TCPC this is really difficult since in all reality a setup person could setup a part in a place in all reality doesn’t even fit? That is the grey area of 5 Axis programming keeping it all under our control and make it work on the machine. 

You don’t just setup a part up without some kind of idea where it is going it be on the machine, you are using a Lang vice, Raptor Dovetail, 5th Axis, Jegens or even a custom fixture you always have an idea where you are going to make the workoffset. You always have an idea where that will be relationship to the machine. DWO and TCPC are great tools and give the programmer and setup person tons of freedom, but yes there are things oh have to go about as part of the programming to make them work correctly. Glad HAAS has finally brought their controls up to date. 

Thank you for the answer I appreciate it. 

[Brian@PC]

I guess I can only respond a few times per day, with a new account. As the reply window was missing for 24hrs.

Quote

Sorry lost me here with this comment “In the non rotated view”. If you have a toolpath done in a different view and make the ref points associated to that view then your coordinates should be rotated in relation to that. 

I could have been more clear and used better phrasing. I was referring to being on the machine, not in Mastercam.  If you call up a X/Y position, when rotated, without DWO active, the X/Y position will not be using the calculated/adjusted work offset on the machine at Axxx.xxx Cxxx.xxx . It will be from your work offset at A0/C0. If you rotate, call up DWO and then your ref point, it will be what you see in Mastercam (retaliative to your T/C plane).  

In Mastercam using a Haas DWO/TCPC post, if you program from the top of your part with a single offset and multiple planes, the program will be the same if you have a 2" or 20" tall fixture. The simulation will be different because of your change in fixture height, just as the machine motion will be different. But, the actual code will be the same as DWO/TCPC is handling the relationship change. The problem occurs when the initial position does not use DWO, there is no way to account for the relationship between the COR and the work offset on the initial move before TCPC. 

On 4/21/2018 at 12:41 PM, StrikeQ said:

Both still do XYZ movements when the G234 is called

The second program should not be used as it is calling a rotation when DWO is being used.  How much movement are you seeing on the first, are you seeing physical movement? 

 

 

[Greg Williams]

Brian,

 

Did you buy a post or create one? If you created one are the XYZ coordinates rotating with the part?

 

I use this format

N1 G21
G0 G17 G40 G80 G90 G94 G98
G255
G69
G1 G53 Z0.
G1 G53 X-1270.0 Y0.
G0 G91 G28 A0. C0.
(10. ENDMILL |TOOL - 1|DIA. OFF. - 1|LEN. - 1|TOOL DIA. - 10.)
M11
M13
T1 M6
G0 G90 G56 X0. Y-95.413
S6000 M3
A13.8 C130.758
G234 H1 X-36.294 Y19.514 Z50.
X-36.294 Y19.514 Z50.
G1 Z-9.92 F9000. M8
X-33.584 Y21.85 Z-24.487
X-32.68 Y22.629 Z-29.342 F2000.
X-32.63 Y22.534 Z-29.348 F5000.
X-32.533 Y22.359 Z-29.358

[StrikeQ]

3 hours ago, Brian@PC said:

I guess I can only respond a few times per day, with a new account. As the reply window was missing for 24hrs.

I could have been more clear and used better phrasing. I was referring to being on the machine, not in Mastercam.  If you call up a X/Y position, when rotated, without DWO active, the X/Y position will not be using the calculated/adjusted work offset on the machine at Axxx.xxx Cxxx.xxx . It will be from your work offset at A0/C0. If you rotate, call up DWO and then your ref point, it will be what you see in Mastercam (retaliative to your T/C plane).  

In Mastercam using a Haas DWO/TCPC post, if you program from the top of your part with a single offset and multiple planes, the program will be the same if you have a 2" or 20" tall fixture. The simulation will be different because of your change in fixture height, just as the machine motion will be different. But, the actual code will be the same as DWO/TCPC is handling the relationship change. The problem occurs when the initial position does not use DWO, there is no way to account for the relationship between the COR and the work offset on the initial move before TCPC. 

The second program should not be used as it is calling a rotation when DWO is being used.  How much movement are you seeing on the first, are you seeing physical movement? 

 

 

Physical movement? Yes both the X and Y are physically moving when the Z moves. It seems to position XY off of the G55 offset, then when G234 is called it repositions XY based off TCPC.

1 hour ago, Greg Williams said:

Brian,

 

Did you buy a post or create one? If you created one are the XYZ coordinates rotating with the part?

 

I use this format

N1 G21
G0 G17 G40 G80 G90 G94 G98
G255
G69
G1 G53 Z0.
G1 G53 X-1270.0 Y0.
G0 G91 G28 A0. C0.
(10. ENDMILL |TOOL - 1|DIA. OFF. - 1|LEN. - 1|TOOL DIA. - 10.)
M11
M13
T1 M6
G0 G90 G56 X0. Y-95.413
S6000 M3
A13.8 C130.758
G234 H1 X-36.294 Y19.514 Z50.
X-36.294 Y19.514 Z50.
G1 Z-9.92 F9000. M8
X-33.584 Y21.85 Z-24.487
X-32.68 Y22.629 Z-29.342 F2000.
X-32.63 Y22.534 Z-29.348 F5000.
X-32.533 Y22.359 Z-29.358

We purchased a post.

Outside of the initial positioning the rest of the tool path seems to run correctly.

[StrikeQ]

16 hours ago, StrikeQ said:

Physical movement? Yes both the X and Y are physically moving when the Z moves. It seems to position XY off of the G55 offset, then when G234 is called it repositions XY based off TCPC.

Don't see an edit button. But using the G254/G255 or leaving it default the XY go to the same location, then re-position when G234 is called. In the case of the first tool path used to Rough, it positions the tool over the trunnion approximately 6-7" away from where it starts cutting. The next program finishing the full blade doesn't have as big of a re position but still moves (~1" but it starts at the OD closer to A0 whereas the rough starts at the LE closer to A90). Haven't moved on to the other programs yet as I'm working on some surface finish issues so been running these 2 programs primarily. But I'm sure they'll have the same initial positioning issues.

[StrikeQ]

We ended up with the post guys modifying it so it puts an XY from the Top Plane in the initial positioning. Then a G234 XYZ from the TCPC plane. The G234 XY while numerically different then the top plane XY are still the same machine position so it accomplishes what we wanted.

Thanks to all who responded.

[crazy^millman]

23 minutes ago, StrikeQ said:

We ended up with the post guys modifying it so it puts an XY from the Top Plane in the initial positioning. Then a G234 XYZ from the TCPC plane. The G234 XY while numerically different then the top plane XY are still the same machine position so it accomplishes what we wanted.

Thanks to all who responded.

Perfect Solution. Thanks for the update.

[ikertx0]

Can this be implemented in the generic VF-TR?

[Greg Williams]

Yes, but you will need to be a post guy to get it all running, there are no on/off switches ready to go. If you are a scripting guy then Colin has some great videos on Youtube that will show you the way.