|
Post by bLouChip on Jan 9, 2023 0:14:12 GMT
I just added the rotary A (4th) axis to my Mega V XL Tri-CAM machine. The assembly went well thanks to tips from this thread. Just a heads up that the latest version of UGS V2.0.12+ from GitHub has support and controls for the A axis now. winder.github.io/ugs_website/The 4 axis version of UGS on MillRight's Resources page is fairly stale now, last built in Feb 2020 from the then current version of UGS, so many improvements and bug fixes in UGS V2.0.12+ since then. I just noticed there is a V2.0.13 recently posted on GitHub, but I'm not using it yet so can't vouch for it. I've been using v2.0.12 for months now and its working great, including with my recently added A axis.
|
|
|
Post by bLouChip on Jan 24, 2023 19:24:16 GMT
I've recently added the MR 4th (rotary A) axis to my Mega V XL. Upon testing beyond the usual command line driven motion tests which all went as expected and fine, correct config and operation with distance movement and speed, an issue crops up on my first laser CAM generated gcode test yielding the infamous G94 G1 XA feedrate disparity issue in grbl, being X is driven in mm/min vs. A driven in deg/min.
I'm using grbl version: [VER:1.1i MegaV 4 Axis Router.20190120:] [AXS:4:XYZA] [OPT:VNML,35,255]
Another user has verified this behavior also with the Mega V. AND... grblHAL, which at one time shared a common source code fork with MR grbl, also has verified the existence of the bug. MR is aware of the bug and has this in the queue to verify and fix. In the meantime, grblHAL is about to finalize their fix, see details here. github.com/grblHAL/core/discussions/241
The good news is that grbl G93 Inverse Time motion mode works as expected and fine, so if you have the option to use this motion mode during gcode generation in the post processor, then you should do so.
Cheers, Lou
|
|
|
Post by bLouChip on Feb 15, 2023 4:51:03 GMT
I have a fix in the works for the MR-grbl G94 G1 XA (rotary) Feedrate bug. I've bench tested it on a separate atmega2560 board driving grbl with UGS and using the DRO and visualizer to time XA motion, so far so good, timings seem spot on with gcode program values. Tomorrow I'm flashing the Mega V for live CNC control testing. I'll continue to report results and progress here.
Update next day: The fix works well and as expected on the Mega V MR controller, per the gcode specification. I'm very happy with it.
I'm sharing it with MillRight Support. If anyone needs it, either of us can share it with you.
This gui reads and writes the MCU, so its very easy to read and save away the factory MR-grbl firmware before flashing to replace it. EEPROM storage, $$ grbl config info, can be saved also, however the flashing process can and does preserve it.
|
|
|
Post by bLouChip on Jun 13, 2023 23:32:42 GMT
A LightBurn user sent me a PM inquiring of more details re. the G94 G1 XA feedrate bug, he said he's seeing the same problem/bug in a Masso controller on an unspecified CNC. Of course with LightBurn, there is no option to use G93 motion mode to avoid the bug. I don't know if he's followed up with Masso re a fix, but I pointed him to the grblHAL fix for the purpose of passing it along as an algorithm or pseudo code to Masso.
I'm posting here as FYI in case this bug crops up in the Mega V Pro Masso controller.
Similarly, the workaround seems to be to use G93 motion mode if the CAM sw supports that.
Again, just posting here as FYI only.
|
|