implementing gantrykins
23 Nov 2012 18:38 #26876
by cncbasher
Replied by cncbasher on topic implementing gantrykins
The coordinates= parameter specifies the initial gantry joint mapping. Each axis letter is mapped to a joint, starting from 0. So coordinates=XYYZ maps the X axis to joint 0, the Y axis to joint 1 and 2, and the Z axis to joint 3. If not specified, the default mapping is coordinates=XYZABC
so changing coordinates in your ini file from XYZA to XXZA should be all you need to do
for speed changes just reduce the acceleration and or velocity values in the ini
so changing coordinates in your ini file from XYZA to XXZA should be all you need to do
for speed changes just reduce the acceleration and or velocity values in the ini
Please Log in or Create an account to join the conversation.
27 Nov 2012 22:25 #26995
by decoyable
Replied by decoyable on topic implementing gantrykins
originally there was XYZA and when i change it it stays the y axis that is joined it doesn't matter if i put XXYZ or anything else its it stays the y axis
in the screen it does change from 4 joints to 3 but nothing else changes
in the screen it does change from 4 joints to 3 but nothing else changes
Please Log in or Create an account to join the conversation.
29 Nov 2012 18:53 #27073
by andypugh
Some confusion here.
The COORDINATES= setting in the INI file is used to determine what axis words are valid. It might, or might not, affect the settings of gantrykins depending on the HAL file.
Gantrykins does not automatically look in the INI file for anything.
To set up gantrykins in a particular way then the HAL file needs to contain either
loadrt gantrykins coordinates=XXYZ
or, you can do it later in the HAL file
loadrt gantrykins
...
setp gantrykins.joint-0 0
setp gantrykins.joint-1 0
setp gantrykins.joint-2 1
setp gantrykins.joint-3 2
The method of using the coordinates= parameter is a little nicer, as is mentioned in the (brief) ganrtykins manpage:
www.linuxcnc.org/docs/html/man/man9/gantrykins.9.html
If you want, you can make the HAL file read the coordinates= parameter from the INI file:
loadrt gantrykins coordinates=[TRAJ]COORDINATES
The docs do mention this: "COORDINATES = X Y Z - The names of the axes being controlled. Only X, Y, Z, A, B, C, U, V, W are valid. Only axes named in COORDINATES are accepted in g-code. This has no effect on the mapping from G-code axis names (X- Y- Z-) to joint numbers—for trivial kinematics, X is always joint 0, A is always joint 3, and U is always joint 6, and so on. It is permitted to write an axis name twice (e.g., X Y Y Z for a gantry machine) but this has no effect."
The reason it is allowable to repeat an axis entry is, I assume, so that you can use the entry for Gantrykins. Note that if the gantrykins HAL line references the INI as I showed above, then there will not be "no effect"
Replied by andypugh on topic implementing gantrykins
originally there was XYZA and when i change it it stays the y axis that is joined it doesn't matter if i put XXYZ or anything else its it stays the y axis
in the screen it does change from 4 joints to 3 but nothing else changes
Some confusion here.
The COORDINATES= setting in the INI file is used to determine what axis words are valid. It might, or might not, affect the settings of gantrykins depending on the HAL file.
Gantrykins does not automatically look in the INI file for anything.
To set up gantrykins in a particular way then the HAL file needs to contain either
loadrt gantrykins coordinates=XXYZ
or, you can do it later in the HAL file
loadrt gantrykins
...
setp gantrykins.joint-0 0
setp gantrykins.joint-1 0
setp gantrykins.joint-2 1
setp gantrykins.joint-3 2
The method of using the coordinates= parameter is a little nicer, as is mentioned in the (brief) ganrtykins manpage:
www.linuxcnc.org/docs/html/man/man9/gantrykins.9.html
If you want, you can make the HAL file read the coordinates= parameter from the INI file:
loadrt gantrykins coordinates=[TRAJ]COORDINATES
The docs do mention this: "COORDINATES = X Y Z - The names of the axes being controlled. Only X, Y, Z, A, B, C, U, V, W are valid. Only axes named in COORDINATES are accepted in g-code. This has no effect on the mapping from G-code axis names (X- Y- Z-) to joint numbers—for trivial kinematics, X is always joint 0, A is always joint 3, and U is always joint 6, and so on. It is permitted to write an axis name twice (e.g., X Y Y Z for a gantry machine) but this has no effect."
The reason it is allowable to repeat an axis entry is, I assume, so that you can use the entry for Gantrykins. Note that if the gantrykins HAL line references the INI as I showed above, then there will not be "no effect"
Please Log in or Create an account to join the conversation.
27 Dec 2012 06:38 #28066
by nicubila
Replied by nicubila on topic implementing gantrykins
@decoyable
saw you did quite a lot of trial (and error), so wondering if you found a solution that works for the 3 axes 4 motors that allows for independent homing of axes.
I am on the way to switch from mach3 with smoothstepper to linuxcnc but struggeling to find a workable solution for this confid. I want also to make a step by step doc that explains for a newbie the solution.
My view is that a cnc user doesn't need to learn all the insides of the linuxcnc in order to do its job. in fact, to attract more users, linuxcnc has to be easy to implement (read have ot running) - then you can learn later more.
saw you did quite a lot of trial (and error), so wondering if you found a solution that works for the 3 axes 4 motors that allows for independent homing of axes.
I am on the way to switch from mach3 with smoothstepper to linuxcnc but struggeling to find a workable solution for this confid. I want also to make a step by step doc that explains for a newbie the solution.
My view is that a cnc user doesn't need to learn all the insides of the linuxcnc in order to do its job. in fact, to attract more users, linuxcnc has to be easy to implement (read have ot running) - then you can learn later more.
Please Log in or Create an account to join the conversation.
27 Dec 2012 08:59 #28071
by decoyable
Replied by decoyable on topic implementing gantrykins
@nicubila
I did find the solution to map 2 joints to 1 axis with independant homing thanks to all the help I got here, I just need to do some more tinkering in the software to make the X axis doubled instead of the Y axis and adjusting the speeds but I didn't have the chance to do that the past couple of weeks with our exams and the fact that all our materials are in school so I won't be able to do further testing till after the 2 week christmas-break.
When I figure it out i'll make sure to let you know.
We've just started making the mechanics part so I'll post the result when we're done. I expect that to be somewhere in April-May.
I share your view that the learning curve for new users is pretty steep, especially when you don't use linux often and you're not as big of a fan/specialist of programming.
I think it's that steep because linuxcnc is freeware and - like linux - things like gantrykins are made by users of linuxcnc and with their experience and knowledge they sometimes forget about people that don't use linuxcnc so often and who are trying to use the software rather than improve it.
I did find the solution to map 2 joints to 1 axis with independant homing thanks to all the help I got here, I just need to do some more tinkering in the software to make the X axis doubled instead of the Y axis and adjusting the speeds but I didn't have the chance to do that the past couple of weeks with our exams and the fact that all our materials are in school so I won't be able to do further testing till after the 2 week christmas-break.
When I figure it out i'll make sure to let you know.
We've just started making the mechanics part so I'll post the result when we're done. I expect that to be somewhere in April-May.
I share your view that the learning curve for new users is pretty steep, especially when you don't use linux often and you're not as big of a fan/specialist of programming.
I think it's that steep because linuxcnc is freeware and - like linux - things like gantrykins are made by users of linuxcnc and with their experience and knowledge they sometimes forget about people that don't use linuxcnc so often and who are trying to use the software rather than improve it.
Please Log in or Create an account to join the conversation.
27 Dec 2012 14:37 #28074
by nicubila
Replied by nicubila on topic implementing gantrykins
@decoyable
thanks for the reply and willingness to help.
actually i have the Y axes doubled (A slave to Y) so this is me:). My request would be therefore if you can share the current config; I am talking abot a very short description of the staps to follow: i.e. start with stepper_gantrykins config, change in the .ini file this and that etc. I don't know is there is a lot to change and would take a lot of your time..
Alternatively it is the option to have the config files simply replaced and it is up to me to figure them out.
As for the motor speeds, I am confortable adjusting them - I use the router for 2 years now and this is not anymore a "do go there, it's dark" area.
thanks
thanks for the reply and willingness to help.
actually i have the Y axes doubled (A slave to Y) so this is me:). My request would be therefore if you can share the current config; I am talking abot a very short description of the staps to follow: i.e. start with stepper_gantrykins config, change in the .ini file this and that etc. I don't know is there is a lot to change and would take a lot of your time..
Alternatively it is the option to have the config files simply replaced and it is up to me to figure them out.
As for the motor speeds, I am confortable adjusting them - I use the router for 2 years now and this is not anymore a "do go there, it's dark" area.
thanks
Please Log in or Create an account to join the conversation.
27 Dec 2012 19:41 #28083
by andypugh
In the HAL file there are lines related to the gantrykins module that specify which joint (basically which motor) is driven by which axis letter in the G-code.
(It should be fairly clear which lines when you look in the right file)
Replied by andypugh on topic implementing gantrykins
It might help to realise that you don't have an "A" axis, you have X Y and Z and joints 0,1, 2 and 3.actually i have the Y axes doubled (A slave to Y)
In the HAL file there are lines related to the gantrykins module that specify which joint (basically which motor) is driven by which axis letter in the G-code.
(It should be fairly clear which lines when you look in the right file)
Please Log in or Create an account to join the conversation.
27 Mar 2013 23:06 #31976
by decoyable
Replied by decoyable on topic implementing gantrykins
Hello everybody it's been a while 
We've been focussing on the mechanical part for some time now and when we built the frame for the X axis (with the 2 stepper motors) we tried to test it but now the 2 x motors stopped working together.
When I jog axis 1 only the x motor runs and when i home joint 1 the A motor runs.
I have no idea how this happened but i took my gantry.hal and gantry-hs.ini files and put them here could you guys take a look?
Thanks, B
We've been focussing on the mechanical part for some time now and when we built the frame for the X axis (with the 2 stepper motors) we tried to test it but now the 2 x motors stopped working together.
When I jog axis 1 only the x motor runs and when i home joint 1 the A motor runs.
I have no idea how this happened but i took my gantry.hal and gantry-hs.ini files and put them here could you guys take a look?
Thanks, B
Please Log in or Create an account to join the conversation.
10 Apr 2013 22:33 #32531
by decoyable
Replied by decoyable on topic implementing gantrykins
Hey everybody,
We decided to start over with a new computer and a freshly formatted HDD we reinstalled everything and now it works like a charm.
We only have 1 other problem (again) when we home the Y axis motor normally it has to come to the right (to our sensor) when it reaches the sensor it has to go left until there's no more signal then go right until it gets signal again then go left until theres no more signal (from the sensor) right?
well ours doesn't go left at all it goes right towards the sensor then when it gets a signal from the sensor it stops for a split second and then it goes right again so it always goes right.
we tried everything to see wheres the problem (exchanged drivers, motors, everything) but its the signal that goes to the Y motor that doesn't invert the signal going to the Y motor.
I put the .stepconf the. ini and the .hal files with this message could anyboody take a look?
Thank you
B.
We decided to start over with a new computer and a freshly formatted HDD we reinstalled everything and now it works like a charm.
We only have 1 other problem (again) when we home the Y axis motor normally it has to come to the right (to our sensor) when it reaches the sensor it has to go left until there's no more signal then go right until it gets signal again then go left until theres no more signal (from the sensor) right?
well ours doesn't go left at all it goes right towards the sensor then when it gets a signal from the sensor it stops for a split second and then it goes right again so it always goes right.
we tried everything to see wheres the problem (exchanged drivers, motors, everything) but its the signal that goes to the Y motor that doesn't invert the signal going to the Y motor.
I put the .stepconf the. ini and the .hal files with this message could anyboody take a look?
Thank you
B.
Please Log in or Create an account to join the conversation.
- captain chaos
- Offline
- Senior Member
-
Less
More
- Posts: 62
- Thank you received: 3
23 May 2013 08:00 #34571
by captain chaos
Replied by captain chaos on topic implementing gantrykins
Only just seen this post, I've also been messing with gantrykins!! Try negative on the home search velocity ie. HOME_SEARCH_VEL = -0.0500
you may need to change HOME_LATCH_VEL = -0.0500 to negative as well.
hope this helps cheers
you may need to change HOME_LATCH_VEL = -0.0500 to negative as well.
hope this helps cheers
Please Log in or Create an account to join the conversation.
Time to create page: 0.084 seconds