Trajectory Planner using Ruckig Lib
- tommylight
- Away
- Moderator
Less
More
- Posts: 18730
- Thank you received: 6297
11 Mar 2024 20:53 #295679
by tommylight
We do not like money, mainly as we have to think what to do with it...
Probably messed it up, but i hope the general idea is still there.
-
Kosovo is not the right place for that, banks are OK but way to much in fees and maintenance, there is an abundance of "fast money transfer" like Western Union that also work but again fees to high, Crypto is OK and usually can be changed for FIAT currency easily and i used it a lot so i do have wallets and stuff.
And that would also imply me sending all that to someone in USA that has to actually pay for server and stuff, so another fee or fee's.
One way of helping that i can think off that has no such issues and works properly is to buy Mesa stuff form JT (aka BigJT) from www.mesaus.com, or whatever he is selling on his website, so helping with no strings attached at all, especially for those in USA.
JT is responsible for this forum and does a great job of keeping it clean, something normal members do not see, but it does require a lot of time, and that in turn keeps him from tending to coops and chicks ... Chicken, chicken, not women!
Replied by tommylight on topic Trajectory Planner using Ruckig Lib
To quote Andy:I hope Tommylight can open a "non profit" bank account in Kosovo for forum donations.
It will certainly help to cover the forum costs, server costs, forum day's.
We do not like money, mainly as we have to think what to do with it...
Probably messed it up, but i hope the general idea is still there.
-
Kosovo is not the right place for that, banks are OK but way to much in fees and maintenance, there is an abundance of "fast money transfer" like Western Union that also work but again fees to high, Crypto is OK and usually can be changed for FIAT currency easily and i used it a lot so i do have wallets and stuff.
And that would also imply me sending all that to someone in USA that has to actually pay for server and stuff, so another fee or fee's.
One way of helping that i can think off that has no such issues and works properly is to buy Mesa stuff form JT (aka BigJT) from www.mesaus.com, or whatever he is selling on his website, so helping with no strings attached at all, especially for those in USA.
JT is responsible for this forum and does a great job of keeping it clean, something normal members do not see, but it does require a lot of time, and that in turn keeps him from tending to coops and chicks ... Chicken, chicken, not women!
The following user(s) said Thank You: Grotius
Please Log in or Create an account to join the conversation.
12 Mar 2024 15:02 - 12 Mar 2024 15:04 #295751
by Grotius
Replied by Grotius on topic Trajectory Planner using Ruckig Lib
Hi,
Today i coded the linuxcnc logic for the trajectory planner.
This means running trough gcode lines, cq waypoints in auto mode.
This part run's after the look ahead optimalisation that
gives us the vo, ve, vm values. After this code, the path interpolation for tp xyz uvw is done.
I coded this outside of linuxcnc source just to validate the output in the qt realtime plot.
Given waypoints for the test:
seg.pnt_s={-50,0,0};
seg.pnt_e={10,0,0};
seg.vo=0;
seg.ve=2;
seg.pnt_s={10,0,0};
seg.pnt_e={20,0,0};
seg.vo=2;
seg.ve=0;
seg.pnt_s={20,0,0};
seg.pnt_e={50,0,0};
seg.vo=2;
seg.ve=maxvel;
seg.pnt_s={50,0,0};
seg.pnt_e={75,0,0};
seg.vo=maxvel;
seg.ve=0;
seg.pnt_s={75,0,0};
seg.pnt_e={100,0,0};
seg.vo=0;
seg.ve=0;
Video's running the program, with some tests like :
- Running forward, reverse,
- Pause, resume.
- Feed override, adaptive feed.
Trajectory waypoints using high jerk:
Trajectory waypoints using low jerk.
Today i coded the linuxcnc logic for the trajectory planner.
This means running trough gcode lines, cq waypoints in auto mode.
This part run's after the look ahead optimalisation that
gives us the vo, ve, vm values. After this code, the path interpolation for tp xyz uvw is done.
I coded this outside of linuxcnc source just to validate the output in the qt realtime plot.
Given waypoints for the test:
Warning: Spoiler!
seg.pnt_s={-50,0,0};
seg.pnt_e={10,0,0};
seg.vo=0;
seg.ve=2;
seg.pnt_s={10,0,0};
seg.pnt_e={20,0,0};
seg.vo=2;
seg.ve=0;
seg.pnt_s={20,0,0};
seg.pnt_e={50,0,0};
seg.vo=2;
seg.ve=maxvel;
seg.pnt_s={50,0,0};
seg.pnt_e={75,0,0};
seg.vo=maxvel;
seg.ve=0;
seg.pnt_s={75,0,0};
seg.pnt_e={100,0,0};
seg.vo=0;
seg.ve=0;
Video's running the program, with some tests like :
- Running forward, reverse,
- Pause, resume.
- Feed override, adaptive feed.
Trajectory waypoints using high jerk:
Trajectory waypoints using low jerk.
Last edit: 12 Mar 2024 15:04 by Grotius.
The following user(s) said Thank You: Todd Zuercher, akb1212, tivoi, pommen, Joco, yrsiddhapura, tiagounderground, routerman22, zmrdko, Unlogic
Please Log in or Create an account to join the conversation.
- grandixximo
- Offline
- Premium Member
Less
More
- Posts: 139
- Thank you received: 5
30 Mar 2024 08:32 #297200
by grandixximo
Replied by grandixximo on topic Trajectory Planner using Ruckig Lib
This is really great, hope you guys succeed in your goal!
I was wondering if you had taken into consideration extending the Kinematics to provide information about the physical layout and geometry of the machine, such as the lengths of linkages, pivot points, and tool offsets. and then use this information for calculating the motion of each axis and adjusting speed, acceleration, and jerk dynamically based on the machine's configuration.
I find that large pivot lengths changes, or even certain movements that merge different joints with different limitations, require large variation of vel acc and jerk of the axis, would be nice if they were dynamically accounted for by the TP and Kinematics components.
I was wondering if you had taken into consideration extending the Kinematics to provide information about the physical layout and geometry of the machine, such as the lengths of linkages, pivot points, and tool offsets. and then use this information for calculating the motion of each axis and adjusting speed, acceleration, and jerk dynamically based on the machine's configuration.
I find that large pivot lengths changes, or even certain movements that merge different joints with different limitations, require large variation of vel acc and jerk of the axis, would be nice if they were dynamically accounted for by the TP and Kinematics components.
The following user(s) said Thank You: Grotius
Please Log in or Create an account to join the conversation.
31 Mar 2024 21:39 #297308
by Joco
Replied by Joco on topic Trajectory Planner using Ruckig Lib
Looking like things have moved along well. Watching with interest whenever I manage access to the forum. Unfortunately my IP is permanently blocked from the forums which makes participation in dicussions a bit problematic.
Cheers - J.
Cheers - J.
The following user(s) said Thank You: Grotius
Please Log in or Create an account to join the conversation.
03 Apr 2024 15:24 #297510
by Becksvill
Replied by Becksvill on topic Trajectory Planner using Ruckig Lib
looking good guys
i have a pretty big 4 and 5 axis horizontal cnc mill that i just moved into the shed ready for linuxcnc. its a mazak fh680 weighs 22tons. and i am keen as to try running the new jerk control on it. i really want to use linuxcnc as i can reuse the drives. but the fact that linuxcnc currently doesn't do jerk control almost makes me want to put a industrial control on it. I just don't want to spend 30k on it. so linuxcnc it is
i understand new jerk control will only do 3axis for now. haha
i think i am one of many people looking on with a lot of interest.
in my opinion this is the most important thing that linuxcnc needs currently
you guys are doing awesome thanks for not giving up.
i have a pretty big 4 and 5 axis horizontal cnc mill that i just moved into the shed ready for linuxcnc. its a mazak fh680 weighs 22tons. and i am keen as to try running the new jerk control on it. i really want to use linuxcnc as i can reuse the drives. but the fact that linuxcnc currently doesn't do jerk control almost makes me want to put a industrial control on it. I just don't want to spend 30k on it. so linuxcnc it is
i understand new jerk control will only do 3axis for now. haha
i think i am one of many people looking on with a lot of interest.
in my opinion this is the most important thing that linuxcnc needs currently
you guys are doing awesome thanks for not giving up.
The following user(s) said Thank You: tommylight, Grotius
Please Log in or Create an account to join the conversation.
- routerman22
- Offline
- Senior Member
Less
More
- Posts: 49
- Thank you received: 8
16 Apr 2024 13:11 #298363
by routerman22
Replied by routerman22 on topic Trajectory Planner using Ruckig Lib
centroid might be a solution for you. It is more expensive than lcnc. But it is relatively affodable.
The following user(s) said Thank You: Grotius
Please Log in or Create an account to join the conversation.
16 Apr 2024 16:17 #298383
by Lcvette
Replied by Lcvette on topic Trajectory Planner using Ruckig Lib
such awesome work!! this is a seriously great project and will be a very welcomed addition to the linuxcnc arsenal! keep up the fantastic work Grotius!
Chris
Chris
Please Log in or Create an account to join the conversation.
17 Apr 2024 09:13 #298424
by zmrdko
Replied by zmrdko on topic Trajectory Planner using Ruckig Lib
We should do some crowdfunding to push things forward.
I would be definitely willing to participate.
I would be definitely willing to participate.
Please Log in or Create an account to join the conversation.
17 Apr 2024 11:14 #298428
by RDA
Replied by RDA on topic Trajectory Planner using Ruckig Lib
This was some what discussed a couple messages before. Many of use probably would probably pitch in to donate, me included, but seems that's not "desired".
I personally probably would not take donations either for something that I might not finish. I know this might sound "wrong", "discouraging" and all that but if I would be working on something that is quite complex and something that other people have already failed before, getting "paid" in advance would add another layer of stress that I would not want.
Many of us have fingers and toes crossed, hoping this project sees light and I am confident that Grotius and Beef will reach the finish line somewhere in the future.
If funding is something that is needed, I hope the guys will let us know and like said before, probably there are quite a few people willing to pitch in even if this never gets finished.
I personally probably would not take donations either for something that I might not finish. I know this might sound "wrong", "discouraging" and all that but if I would be working on something that is quite complex and something that other people have already failed before, getting "paid" in advance would add another layer of stress that I would not want.
Many of us have fingers and toes crossed, hoping this project sees light and I am confident that Grotius and Beef will reach the finish line somewhere in the future.
If funding is something that is needed, I hope the guys will let us know and like said before, probably there are quite a few people willing to pitch in even if this never gets finished.
The following user(s) said Thank You: Grotius
Please Log in or Create an account to join the conversation.
17 Apr 2024 11:22 #298429
by zmrdko
Replied by zmrdko on topic Trajectory Planner using Ruckig Lib
How about collect $ upon delivery then?
The following user(s) said Thank You: Grotius
Please Log in or Create an account to join the conversation.
Time to create page: 0.209 seconds