Rotational axis offset
327 views
Skip to first unread message
peter....@fau.de
Feb 15, 2016, 2:19:01 PM2/15/16
to Pocket NC
Hi... I recieved my machine last week and it's brilliant. You can see that it's designed by someone who put their heart in it..
There's two things I'm struggling with though:
Firstly, it seems the B axis (haven't gotten to measuring it on the A axis yet) is offset with respect to coordinate zero. I made a test piece and I get an offset of 0.39 mm in z and 0.72 mm in x. Can I adjust that in Linux CNC?
Secondly, I get weird behaviour in Linux cnc with some of the g code posted from fusion. At some point during the initial rapid in the bit flips it's direction 180 degrees in the visualisation and I get an error that the movement on line ... Exceeds joint 2s limit. Anyone else having that issue too?
Pocket NC
Feb 15, 2016, 3:37:35 PM2/15/16
to Pocket NC
Hi Peter,
Can you send a picture that would describe what you are talking about, I am having a hard time visualizing this B axis issue. Can you send a the Gcode output file you are working with or the fusion file. With such small Z travel, it is not uncommon to have a z axis over travel error as the setup only has to be off by a small amount.
Thanks,
Matt
peter....@fau.de
Feb 16, 2016, 1:53:08 AM2/16/16
to Pocket NC
Hi Matt,
I made a simple cube with recesses on the back and front. They are arranged such that one side is machined at B = 0 and the side at B = 180. This way I can measure the offset in x. The front and back recesses don't touch, wich allows for the z offset to be measured, by comparing the distance between the recesses to the dimensions in the drawing (assuming the TLO is correct).
I made a simple cube with recesses on the back and front. They are arranged such that one side is machined at B = 0 and the side at B = 180. This way I can measure the offset in x. The front and back recesses don't touch, wich allows for the z offset to be measured, by comparing the distance between the recesses to the dimensions in the drawing (assuming the TLO is correct).
I noticed that offset on all pieces i tried to machine from both sides, that's why I
Alex Richard
Feb 16, 2016, 2:05:47 PM2/16/16
to Pocket NC
Peter, where is the work coordinate in the CAM system compared to the work coordinate on the machine? Are they both on the center of rotation where the A and B axes intersect? If you program off of center of rotation, then you shouldn't have these issues. If you have your offset somewhere other then center of rotation, you need to make sure the CAM system isn't rotating the coordinate system when outputting nc code because to my knowledge the Pocket NC machine only supports a fixed global coordinate system.
From what I found, the default work coordinate position on the PocketNC is .885 in above the table (center of rotation of A), G53 X0, G53 Z0. G53 Z0 is 3.6 inches away from the center of the table in the Z direction.
The Pocket NC tool length offset tutorial says to subtract each tool stick out length from 3.6 inches to account for this default work coordinate position. In my opinion, a better way to do this is to shift the work coordinate position 3.6 in Z in the Axis interface so that it is at the center of the table and then just enter the tool stick out distance in the tool table. Either way, the Z0 position on the machine will be at the center of rotation of both A and B axes.
Hopefully this is helpful. I haven't used Fusion360 to know what settings need to be adjusted from that side of things.
Alex
peter....@fau.de
Feb 16, 2016, 3:32:28 PM2/16/16
to Pocket NC
I've used the supplied pocketNC fusion 360 reference file. The origin point is sketched in it and I've used it as the zero point. I sent the file over to Matt. Maybe the reason is just me doing something wrong. If you want to try on your machine, the file is:
Pocket NC
Feb 17, 2016, 2:21:59 PM2/17/16
to Pocket NC
Hi Peter,
It looks as if your machine may have an alignment issue. Your fusion file is correct. I can explain how to adjust that file if you like, or we can send you a shipping label and re align the machine here. Just let us know.
-Matt
peter....@fau.de
Feb 17, 2016, 2:27:21 PM2/17/16
to Pocket NC
Hi Matt,
i think we may have to work it out through remote assistance. I am sitting in Germany and had the machine passed on through a friend in Michigan.
i think we may have to work it out through remote assistance. I am sitting in Germany and had the machine passed on through a friend in Michigan.
Do you think it would only be the B axis or likely the A axis too?
You can email me directly too.
Best, Peter
peter....@fau.de
Mar 4, 2016, 2:37:10 AM3/4/16
to Pocket NC
I've found the issue. Every time I switch on linuxcnc, it has an offset on the WCS *facepalm*. This offset additionally is not the same every time, making it really confusing. This was the reason the rotational axes were offset with respect to the WCS. Once the offset was cleared, the rotation was good within abt. 8 mil. I've got it dialed in down to 4 now, but i guess 2 are doable.
I also switched to a Delrin block for the fine alignment, since stray wood fibres did get between the dial indicator and the block a number of times.
fra...@pinza.demon.co.uk
Mar 4, 2016, 6:10:19 AM3/4/16
to Pocket NC
"WCS" ?
peter....@fau.de
Mar 4, 2016, 6:37:54 AM3/4/16
to Pocket NC
Work coordinate system. G54 etc
rob.johns...@gmail.com
Jun 14, 2016, 8:05:03 PM6/14/16
to Pocket NC
What is the procedure for checking the alignment? Wondering if this might be the problem I'm having.
Tokyo Pav
Jun 15, 2016, 6:45:55 AM6/15/16
to Pocket NC
@Hi Peter
Out of curiosity how much did it cost to ship from Michigan to Germany?
Out of curiosity how much did it cost to ship from Michigan to Germany?
Reply all
Reply to author
Forward
0 new messages