TCP 5-axis kinematics

More
26 Dec 2019 11:31 #153225 by papagno-source
Thanks for the reply.
Summing up.
If I have a table where the rotation center of B is perfectly in axis with the rotation center of C the value of X.offset is 0,
the z.offset value will be equal to the distance of the table surface (perfectly horizontal) and the rotation axis of B.
I wonder would a y.offset make sense?
While the values of x.rot.point and y.rot.point are the distance values between the zero ref position and the table center of the respective axes.
While the value of Z.rot.point, would it be the distance with respect to the position of z ref, of the spindle nose with respect to the table surface or to the rotation axis of B?
Thanks you

Please Log in or Create an account to join the conversation.

More
26 Dec 2019 13:15 - 26 Dec 2019 14:21 #153232 by plopes9000
Replied by plopes9000 on topic TCP 5-axis kinematics

Thanks for the reply.
Summing up.
If I have a table where the rotation center of B is perfectly in axis with the rotation center of C the value of X.offset is 0,
the z.offset value will be equal to the distance of the table surface (perfectly horizontal) and the rotation axis of B.


If B and C rotation centers are perfectly aligned, Z and X offsets should be set to zero.

I wonder would a y.offset make sense?

You’re right it would not, I was referring by mistake to an AC configuration.

While the values of x.rot.point and y.rot.point are the distance values between the zero ref position and the table center of the respective axes.
While the value of Z.rot.point, would it be the distance with respect to the position of z ref, of the spindle nose with respect to the table surface or to the rotation axis of B?

X, Y and Z are the offsets from whatever position your LinuxCNC considers as zero machine coordinates and the center of rotation of your B axis assuming no offset from A to B. A to B offset is then set on X and Z offsets (corrected)

So the logic is A axis to B axis offset and then B axis to machine zero. (Corrected) The spindle is not part of the equation here.

Have not tested XYZBC formulas from the branch but just had a look and they look good - XYZAC are for sure good.
Last edit: 26 Dec 2019 14:21 by plopes9000.

Please Log in or Create an account to join the conversation.

More
26 Dec 2019 13:27 - 26 Dec 2019 13:29 #153236 by jsskangas
Replied by jsskangas on topic TCP 5-axis kinematics
There are not yet working kinematic for BC on branch.

But working ones That I have made for BC machine can be found and downloaded from this thread.
You can find these packets from Page 5, initial and Page 8, some update.

Please read pages 4-9 there are my posts that that has need to know info how to implement those.

I will so update to BC-kins as soon as I get bit free time...
Last edit: 26 Dec 2019 13:29 by jsskangas.

Please Log in or Create an account to join the conversation.

More
26 Dec 2019 13:58 #153237 by papagno-source
Thanks for reply.
I stay use in Dgarret/s-kins the example xyzbc-switchkins configuration.

It implement 2 remapped fuction M428 and M429 for select TCP-on and TCP-off.

I thing kinematic module is OK.

I would like further clarification regarding the values of x.rot.point, y.rot.point, z.rot.point.
the my machine when have ref ok executed, have axis X at 300mm left the B center rotation, Y 350mm in positive direction the Y (door machine direction) and Z axis at 700 mm (nose spindle) up B center rotation.

in this case X.rot.point have value 300
Y.rot.point have value 350 mm
and Z.rot.point have 700 mm?

Thanks

Please Log in or Create an account to join the conversation.

More
26 Dec 2019 14:22 #153238 by jsskangas
Replied by jsskangas on topic TCP 5-axis kinematics
Hello

"I stay use in Dgarret/s-kins the example xyzbc-switchkins configuration."


Last time I took a look those, Kinematics where not complete, rotary point implementation has not done to BC kins.
ln AC kins rot point is ok.

Please Log in or Create an account to join the conversation.

More
26 Dec 2019 14:24 #153239 by plopes9000
Replied by plopes9000 on topic TCP 5-axis kinematics
I’ll reply to your last post later, unless someone gets ahead of me sooner.

Just wanted to point out that I updated my last post ... please read it again.

It’s been a while and I’m speaking from memory.

Please Log in or Create an account to join the conversation.

More
28 Dec 2019 17:48 #153371 by plopes9000
Replied by plopes9000 on topic TCP 5-axis kinematics

I would like further clarification regarding the values of x.rot.point, y.rot.point, z.rot.point.
the my machine when have ref ok executed, have axis X at 300mm left the B center rotation, Y 350mm in positive direction the Y (door machine direction) and Z axis at 700 mm (nose spindle) up B center rotation.

When you say "have ref ok executed" do you mean that you homed the machine?

When you say "axis X at 300mm left the B center rotation, Y 350mm in positive direction the Y (door machine direction) and Z axis at 700 mm (nose spindle) up B center rotation", do you mean that your G53 offset values at B Center of rotation are X 300, Y 350 and Z 700? if yes then yes to the below :

in this case X.rot.point have value 300
Y.rot.point have value 350 mm
and Z.rot.point have 700 mm?

Please Log in or Create an account to join the conversation.

More
29 Dec 2019 10:44 #153423 by papagno-source
Hi.

Thanks for reply.
Sorry my English. I meant when the machine is homed on every axis.

Ok for now i wait the fix XYZBC switchkin , because JSSKANGAS said that that is not complete, it not have x,y,z,rot.point implementation.

But i Thing for some test : if i write in ini file in HOME_OFFSET the value the distance the every axis restect B centre (Home_offset on Joint0 = 300,Home_offset on Joint 1 = 350, Home offset joint 2 = 700), the machine when executed home procedure, after have read the zero mark the linear scale , positionig axis on Home offset value and machine move on centre the B axis)

I will do some tests, I hope that JSSKANGAS can help us to implement the x.y.z rot.point in the XYZBC switchkin soon.

Thanks for your advice

Please Log in or Create an account to join the conversation.

More
29 Dec 2019 12:07 #153428 by jsskangas
Replied by jsskangas on topic TCP 5-axis kinematics
Hello

Did you read my earlier post?

"But working ones That I have made for BC machine can be found and downloaded from this thread.
You can find these packets from Page 5, initial and Page 8, some update.

Please read pages 4-9 there are my posts that that has need to know info how to implement those."

I have BC machine running with TCP.
Page 5 in this topic you will find attachment where is working BC kinematic .c file.

Please Log in or Create an account to join the conversation.

More
29 Dec 2019 22:53 #153520 by papagno-source
Ok thanks have read it and will test this switchkin file and your configuration.

One question .

at present, when I have to activate or deactivate TCP, do I have to do it with the axis positions of B = 0 and c = 0?

Thanks

Please Log in or Create an account to join the conversation.

Time to create page: 0.116 seconds
Powered by Kunena Forum