Getting correct offset output in Transform toolpath

[JB7280]

I have a tombstone with 4 parts on it.  OP2 is on B0 and B180, and OP3 is on B90 and B270.

 

I'm trying to use transform, to output code that will cut both parts, identically.  I'm able to get the toolpaths in the right place, but the I haven't been able to get it to output the right work coordinates.  OP3 has 5 offsets, numbered 10-14.  I would like for the offsets for the second side to be a continuation of these offsets (15-19)

 

Here are my transform settings right now.

 

 

 

 

[AHarrison1]

What happens when you change 'work offset numbering'?

[JB7280]

3 minutes ago, AHarrison1 said:

What happens when you change 'work offset numbering'?

Closer, depending on the settings I pick, however, then the offsets are jumping back and forth between parts, if that makes sense.  B0/B90 on part 1 are 5, and 6, then B0/B90 on part 2 are 7 and 8.  so part one will end up being something like P5, P6, P9, P11, P13.  I was just trying to keep them grouped together, to make life simpler for the operator.  

 

Am I asking for too much of Transform?  Or am I just doing it wrong?

[AHarrison1]

Might be a good time to reach out to your re-seller,

those questions are slightly above my pay grade.

[JB7280]

Just now, AHarrison1 said:

Might be a good time to reach out to your re-seller,

those questions are slightly above my pay grade.

Thanks for the effort!

[JParis]

I always,, a;ways, always create my offsets now via the Create new offset setting, then I go through and name them as I want them...and then I go through and do my transforms.

Can you share your file?

 

[JB7280]

23 minutes ago, JParis said:

I always,, a;ways, always create my offsets now via the Create new offset setting, then I go through and name them as I want them...and then I go through and do my transforms.

Can you share your file?

 

Check your PM's

[AHarrison1]

32 minutes ago, JB7280 said:

Thanks for the effort!

 Might be worth a read

 

[Jake L]

Hard to fully understand what you've currently got without a file. 

I understand what you're looking for but there are some holes:

are op3 offset 10-14 5 separate parts all at B90? or are those 5 different angles on one part?

Using a generic post or something custom? Got any options on the toolpaths misc. values page?

What does the output for B270 OP3 look like with your settings above? Do they get assigned new work offsets? Or do the B90 OP3 offsets get copied for B270?

One of these days someone is going to ask a transform question and I'm gonna take some time to do some research and make a long post explaining each option and common pitfalls. Unfortunately today is not that day, but some day.

 

[JParis]

I can generally get anything I want out of transform..looking through his file, digging thru the nci, his issue, I believe goes much deeper into the database matrix...it looks like a slight skew, several digits out, is off by .XXXXXXX amd that is enough, that the post can't calculate an exact offset at 180°....I have seen the issue previously on a handful of occasions..even Karlo was able to see it and even he could not correct for it...

Essentially his file is trying to create a move in an axis that his machine does not have...

I hope it is not his issue as the file will have to be completely restarted if it is.

I directed him to his reseller in hopes that they had a better solution 

[Kyle F]

50 minutes ago, Jake L said:

One of these days someone is going to ask a transform question and I'm gonna take some time to do some research and make a long post explaining each option and common pitfalls. Unfortunately today is not that day, but some day.

I will be patiently awaiting

I have used transform to rotate toolpaths around a tombstone with B in 90 deg increments, but that was using dynamic work offsets so everything is G54 / G254

My workflow is usually a single part at a time so it's rare I transform toolpaths, but I should be doing it much more often in the future.

[JB7280]

1 hour ago, Jake L said:

Hard to fully understand what you've currently got without a file. 

I understand what you're looking for but there are some holes:

are op3 offset 10-14 5 separate parts all at B90? or are those 5 different angles on one part?

Using a generic post or something custom? Got any options on the toolpaths misc. values page?

What does the output for B270 OP3 look like with your settings above? Do they get assigned new work offsets? Or do the B90 OP3 offsets get copied for B270?

One of these days someone is going to ask a transform question and I'm gonna take some time to do some research and make a long post explaining each option and common pitfalls. Unfortunately today is not that day, but some day.

 

Sorry, I didn't want to share a file with EVERYONE.   People I've dealt with before, not as big of a deal.   To explain some of the "holes".   

 

Part #1 has 5 B rotations.   10-14.   

 

MPMaster, with some modifications.   

 

Only options I'm using on the misc. page is G05.1.   

 

I'll repost tomorrow and let you know what it looks like with those options.  

1 hour ago, JParis said:

I can generally get anything I want out of transform..looking through his file, digging thru the nci, his issue, I believe goes much deeper into the database matrix...it looks like a slight skew, several digits out, is off by .XXXXXXX amd that is enough, that the post can't calculate an exact offset at 180°....I have seen the issue previously on a handful of occasions..even Karlo was able to see it and even he could not correct for it...

Essentially his file is trying to create a move in an axis that his machine does not have...

I hope it is not his issue as the file will have to be completely restarted if it is.

I directed him to his reseller in hopes that they had a better solution 

Is there a way to avoid a skew like this?   I move almost everything via Dynamic transform.   Should I be doing it differently?

[Jake L]

13 hours ago, JB7280 said:

Sorry, I didn't want to share a file with EVERYONE.   People I've dealt with before, not as big of a deal.   To explain some of the "holes".   

Part #1 has 5 B rotations.   10-14.  

MPMaster, with some modifications.   

Only options I'm using on the misc. page is G05.1.   

I'll repost tomorrow and let you know what it looks like with those options.  

No worries, JP knows a lot more than I do anyways. Don't think I would've caught the "slight skew", but good to know in case I run into something similar.

[JParis]

14 minutes ago, Jake L said:

No worries, JP knows a lot more than I do anyways. Don't think I would've caught the "slight skew", but good to know in case I run into something similar.

If it really acts suspiciously and you just can't get a simple offset to output....post the nci, you "should" only show XYZB info in the starting line, when you see that 5th axis has even the smallest value, something is toast....you can try to realign everything but you're working out to 8 decimal places and .00000001" can make it impossible...

I only see this crop up once in a great while and "generally" in my experience, it is related to trying to get rotational offsets and you are working on a WCS that is not the system origin...

I've said on numerous occasions, yes, it "should" work that way...the issues I see on occasion made my workflow become I ONLY work off of the system origin using TOP WCS....

And I know there are going to be people who say they work in space all the time and never see the issue....I will not dispute that they have not seen it yet...it only appears rarely but when it does, it can cause hours upon hours of lost work.

JM2C YMMV

[JB7280]

7 hours ago, JParis said:

If it really acts suspiciously and you just can't get a simple offset to output....post the nci, you "should" only show XYZB info in the starting line, when you see that 5th axis has even the smallest value, something is toast....you can try to realign everything but you're working out to 8 decimal places and .00000001" can make it impossible...

I only see this crop up once in a great while and "generally" in my experience, it is related to trying to get rotational offsets and you are working on a WCS that is not the system origin...

I've said on numerous occasions, yes, it "should" work that way...the issues I see on occasion made my workflow become I ONLY work off of the system origin using TOP WCS....

And I know there are going to be people who say they work in space all the time and never see the issue....I will not dispute that they have not seen it yet...it only appears rarely but when it does, it can cause hours upon hours of lost work.

JM2C YMMV

Just to update, I was able to get a pretty workable solution using these settings.  The only thing I don't LOVE is the first part will be machining on B0 and B90, and the second part will be machining on B180 and B270, with B180. in the work offsets for part number 2.  But, I think it'll work just fine. 

 

Part of the problem I was having, is that when it tried to machine part 1 at B0(relative to it's work offset) then tried to spin around to machine B180, and machine B0 again (relative to that work offset) the post, or mastercam wasn't recognizing a rotation, so it wasn't outputting the necessary codes.   At least that's what I think was happening.  

 

 

[Jake L]

54 minutes ago, JB7280 said:

The only thing I don't LOVE is the first part will be machining on B0 and B90, and the second part will be machining on B180 and B270, with B180. in the work offsets for part number 2.

If you check the "include WCS" box, the B180 side should post out as B0

[JB7280]

8 minutes ago, Jake L said:

If you check the "include WCS" box, the B180 side should post out as B0

It does, but that's when I get the behavior I mentioned above where it doesn't actually output the B rotations.  Maybe it's just a setting somewhere? 

 

This is what it outputs, when there SHOULD be a return to home, B axis unclamp, rotate, clamp, etc.  

 

G01 X-.0823 Y-6.8609 Z5.5807
G00 Z7.
(ROUGH TOP FACE)
G54.1 P4 X-2.8619 Y-7.8389 Z7.
Z6.1