2015-01-05 17:01 GMT+02:00 Sebastian Kuzminsky <[email protected]>: > On 01/05/2015 04:46 AM, Andrew wrote: > > Hello Sebastian, thank you! > > > > I actually enjoyed the process and I'd like to continue (very slowly) > > contributing to genhexkins and its configs. > > I could add optional screw lenght compensation for cardanic joints after > > it's throughly tested on my machine. > > Probably add some pins to adjust position online (could be useful for > > homing, but I have to figure it out yet). > > I could also share a few realtime configs, e.g. parport stepper and > > analog servo. > > Though I mostly use configs for joints_axes, they could be reverted. > > Please let me know if it's appropriate. > > That would be great. I'm glad you're giving nontrivial kinematics a > workout, that's an area of LinuxCNC that could sure us a good developer. > Any patches in that area would be much appreciated. >
Unfortunately I'm not a good developer, neither a developer at all. I know too little and only can do the simplest things. But sometimes that could be enough, e.g. for kinematics. Do you have a Stewart Platform machine? If so i'd also love to see a > video of it! > Here's a couple videos of my old machine https://www.youtube.com/watch?v=G_UmhUjZhNo https://www.youtube.com/watch?v=nebJ59TcYlQ I don't have it any more, but I built a new machine with better servos and Mesa 5i25 + 7i77. Now I'm slowly working on it to perform an actual milling. Recently I managed to get it homing. I also have a toy hexapod https://www.youtube.com/watch?v=ouyx65C7L-o a delta robot https://www.youtube.com/watch?v=2STkBhUiKwg Actually the homing procedure is one of the weakest spots for non-trivial machines, particularly Stewart platforms... let me make an example. We usually want home to be something meaningful like 0 0 0 0 0 0 or 0 0 -50 0 0 0 (instead of 1.251 3.562 -52.276 0.646 0.252 0.156), but home switches generally can not be adjusted right to those positions. That means HOME and HOME_OFFSET are different. And that means the first joint to hit its home switch already goes to its HOME while others still search for their switches. That could easily break the machine. The question is could we synchronise the joints homing by introducing an optional pause until all home switches (or indexes) trip and then go to home? That should probably help with a gantry homing too. > Also, if you have experiences with both the mainline branches and > joints_axes, that's something i'd love to hear about. Since I tried joints_axes for the first time I could not use anything else for my non-trivial machines. I just dream of it to eventually go to master. I'm also retrofitting a lathe and a mill to LinuxCNC (but not using them much yet). So yes, I'm familiar with both mainline and joints_axes branches. Oh, I should mention that i changed your first commit a tiny bit. You > had removed the Copyright lines from genhexkins.[ch], i added them back in. > > Also, a small request.... Please try to avoid making spurious changes > unrelated to your intended change, it makes it harder to review the > patch. I highlighted a couple of the instances i stumbled over in your > genhexkins.c patch here: > > > https://github.com/jepler/linuxcnc-mirror/commit/0cc3445e1d7fc0a405039d07df6bef8d46b35c29 > Thanks for your notes. The Copyright gone is my mistake, thanks for correcting it. And I really should be more careful with unnesessary changes. I remember that I wanted to clean a little and used "replace tabs by spaces" function on the whole genhexkins.c. Now I see how bad it was. About joints_axes again, how can I help? -- Andrew ------------------------------------------------------------------------------ Dive into the World of Parallel Programming! The Go Parallel Website, sponsored by Intel and developed in partnership with Slashdot Media, is your hub for all things parallel software development, from weekly thought leadership blogs to news, videos, case studies, tutorials and more. Take a look and join the conversation now. http://goparallel.sourceforge.net _______________________________________________ Emc-developers mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-developers
