TwinCat / PlcOpen / IEC1131 6-axes Robotic coordinate transformation ?

Is there anybody who can help me with an example of the coordinate transformations for my 6-axes Manutec R3 robots ?

I want to control 2 robots with Beckhoff Twincat & Ethercat drives.

That problem is too difficult to explain in a forum. Get a text book and prepare to do some hard core math.
There is a reason people buy robotic controllers. They are much more than just devices. There is a lot of software knowledge built into the controllers. You will need to re-invent the wheel and this wheel is not simple.

Hallo Peter,

of coarse I know it takes some real math.
I can also write all matrix calculations myself but it would be easier if someone has an example in plcopen or any basic language.
I plan to calculate all motor positions myself and then force the set-position to the drives.
I have done this before with some 3-axes manipulators but it would be nice if someone has a fault-free kinematic calculation.
All mechanical parameters are known for this robot.

If I program this directly in TwinCAT it is not necessary to have a special robot controller because my NC-cycle time will be 1 msec.

First, I don't use matrices when doing this kind of math. The indexing is slow. I work it out symbolically where the resulting formulas can be simplified. I use Mathcad or more recently Mathematica to do these symbolic simplifications. The code may be a little longer but it executes much faster.
The hard part is the inverse kinematics where you know the end points but must iteratively calculate angles and extensions using Jacobians. If these calculations were done using matrices there would be a lot of 6 x 6 matrix math. I just can't see doing this without a library and then one needs to write the inverse kinematics code.

1 ms cycle time is very aggressive if running on PC platform where Ethernet must be handled too.

i have some pdf on this subject, and probably the factory has this done before.
give an email to [email protected]

all these calcs are very consuming on a plc as it stays a language and a machine to do some ladder not complex arithmetics.
yes i am dutch.

i forgot to mention the book from stryk

Will the Beckhoff package work?
http://www.beckhoff.com/TF5110/

CharlesM said:

Will the Beckhoff package work?
http://www.beckhoff.com/TF5110/

Click to expand...

I don't think so. TF5113 should.

I just bought the 2 robots for the fun to study the possibilities.
( after I sold the original ancient control cabinets I still have the 2 robot-arms for free )
Therefore I am not planning to buy the option-package from Beckhoff.
With Level 4 you can configure a 6-axes robot arm.
Processor network capacity is no problem with TwinCat & EtherCat drives.

So I only need an error-free example of these calculations ( if possible in plc-open basic )

You mention an "option package" from Beckhoff. I understand that they do have such kinematic transformation software and that the PLC can do the kinematic calulations on-the-fly. They have an demo they took to many shows some years ago and you can find a video on the web.

A 3-arm gantry robot positions a "hand" in space using rotary positioners (3?). The "hand" rotates a "finger" w/ more rotary positioners (2?). They move to stick the finger in and out of a hole in a rotating platter. They start real slow so you can see the motions, then speed it so fast it becomes a blur. A Beckhoff salesman told me that the "schedule" sent to the PLC is just the desired x,y,z coordinates for the finger-tip vs time (with finger pointing straight down). The PLC code calculates in real-time the necessary rotary positions to realize those, i.e. real-time kinematic calculations probably at <1 ms cycle (some Beckhoff PLC's can run 0.05 ms cycles). I have no idea how they do that, given Peter's discussion of matrix calculation efficiency. I know that in the somewhat similar world of satellite trajectories that "quaternians" simplify calculations. I vaguely recall that is a 4-D "tensor" that includes time as the 4th dimension, which is already into grad-school Physics beyond my education.