POST help needed for Mitsubishi M80 A-axis

[pickinben]

Here's a more concise and accurate version:

---

Hello all,

I'm using a router with an M80 controller and a 4th axis. While it works well with MadCAM and Rhino, I have projects that would benefit from being done in Fusion. Currently, my post-processor handles XYZ fine but fails for Z values when the A axis is introduced.

I spent a whole day on Zoom without any progress and was told to seek help here. I'm unsure if the issue is with my post-processor, machine definition, or both. I was informed that using the "HAAS A axis Pre NGC" post-processor generates the correct Z values, but it failed when I tried it.

Any direction would be greatly appreciated.

[a.laasW8M6T]

Hi

Can you share a bit more information.

When you say it Fails for Z values do you just mean the are not correct?

If you could share your post, Machine Definition and a sample file you are trying to machine we can have a look to see where the problem lies.

You can export the machine definition and post from their respective libraries and to share the Fusion file Goto File>Export>save as .f3d/.f3z

[serge.quiblier]

Hi @pickinben 

a simple question that may solve your issue.

Where is the wcs in your actual programming?

In 3 axis you can set the wcs where ever you want, as long as the same spot is used on the machine.

But with 4th axis programming, on most machine, the wcs MUST be at the centerline of rotation, not the top or bottom part of the stock.

Regards

[pickinben]

Hi @a.laasW8M6T and @serge.quiblier ,

Thanks for having a look at this and apologies for not having all of the files uploaded.  After a few more hours I have 2 main points I hope to clarify.

1. The post only completes when it is used with the machine model in the setup.  When complete though, it seems to only post the coordinates from the main WCS of the setup, even when there are A-axis rotations.  When A180 happens nearly all the Z values are negative which would be catastrophic.  The XYZ values are a little trickier as I cannot see if how they are transformed by the orientation.

2. Without the machine config the post will work until it hits a reorientation op, and then it fails.  I have attached the part file, but Fusion keeps crashing whenever I try and export the machine.  I'll try again on another computer.

Thank you very very much for having a look!

[a.laasW8M6T]

Hi, see video for explanation

I couldn't really do any more testing without the machine being correct and I don't have your post processor so I couldn't check that either

[serge.quiblier]

Hi @pickinben 

1- In the machine kinematic the rotary axis, what ever it is, MUST NOT be defined with TCP support.

Most 4th axis machines does not support TCP, whereas 5 axis machines will be supporting more frequently (cost option, ....)

So, untick the TCP checkbox in the rotary axis definition

2 - That is expected. In order to calculate one or two angular coordinates, for the rotary axes, the post MUST know the kinematic definition. By default a stand alone generic post does not contains such definition. Some may have properties to simply add a rotary axis, but not all posts support that functionality.

Using the machine definition file allow us to keep the post generic, while offering much more solutions, for the axis oridentation, limits.... using a machine definition file that is freely editable.

Regards

[pickinben]

Wow,  where to start I'd normally say, but that video gave many options that are less cryptic than banging my head against the postprocessor.  Thank you so very much for this.  I'll implement your machine model "standards" / hierarchy.  I had deleted my trunnion/rotary component as the last Autodesk Support person I had spoken to told me simple machine models were the way to go.  Lastly, somehow I didn't include the post I had thought was there.  I am attaching it now, though I am not sure at this point it is the most necessary problem to address.  I'll check back in when I have addressed the host of issues with the machine model in the woodshed.  Thanks again!

[pickinben]

1. Noted! My machine does support this function, but since I do not know what it is, how it works, or how to make the settings in the controller to unsure its safe usage, I'll leave it alone.
2. Gotcha. I'll clean up my machine definition and see to what extent this resolves my issues.

Many thanks!

[a.laasW8M6T]

Hi

"Autodesk Support person I had spoken to told me simple machine models were the way to go. "

That is correct , but you may have misinterpreted what they meant.

The machine model should be simple in the respect that there isn't too much detail in the model shapes, they should be simple as more complicated shapes take more processing power to collision check.

That doesn't mean you should remove machine parts entirely, especially functional ones like the Rotary.

You can turn of collision checking for components that will never interact and this will improve simulation performance too:

https://help.autodesk.com/view/fusion360/ENU/?guid=MFG-MACHINE-MANAGE-COLLISION-CHECKS

[pickinben]

A few things. First, I cleaned up the machine model and got the same results as before, as I am just not as facile with it as I should be regarding best practices for structure detail, etc. It seemed to only complicate my understanding of the post.

Then, an automatic update to the post appeared that, when paired with the kinematics in the machine definition (absent a 3D model of the machine), seemed to output code accurately during testing. The latest post handles a lot more input from the machine configuration outside of the post, which was a game changer, fixing a number of issues and allowing me to post a nearly correct program.

Today, while running test cuts, I saw that the A axis was turning backwards from the post, so I changed its orientation to custom and reversed the X vector orientation. This solved the A axis orientation issue and made everything work perfectly.

I would greatly appreciate getting a machine simulation model working at some point in the future, but for now, I’m very happy with the progress made thanks to your input.

Thank you all for your assistance. Your input has been invaluable, and I am deeply grateful for your help.