INVMAT MULMAT

[Robo Guru]

Anyone have experience using these 2 Motoman commands?  Any programming examples or situations that can be shared??

RoboGuru

[Fabian Munoz]

I used MULMAT (I inherit the program)
I think is useless or at least it is on that particular program
I sent you a copy of the program

[95devils]

I've used those instructions both in vision jobs and in arc welding as a 3D shift, instead of using relative job.  There is a parameter for the 3D shift instruction.

Check your email, I have included two links to the inet with a write up from TAT. 

[Thescoe]

I seem to recall a long time ago using INVMAT to change a positive variable value to a negative one or vice versa. I was experimenting with rolling my own weld weave macro. The experiment didn't fulfill expectations but the INVMAT instruction worked as planned.

[jejbr84]

You can use it to convert positions between User Frames.
It works the same as with homogeneous transformation matrices (a H-matrix).
INVMAT inverts the H-matrix and MULMAT multiplies by a H-matrix.

[Robo Guru]

Jejbra84,

I am having a hard time understanding these commands and I believe it is because it gets into "robot math" with matrix calculations.  Do you have any resources you may be able to share that might help me get a better understanding but without getting completely involved?

RoboGuru

[jejbr84]

Everything I write below is just from my own test experience with the inform commands, so I don't know if everything I write is 100% true, but I think it is like this:

The functions just do a standard coordinate transformation that you should be able to find on the internet in many places.

INVMAT:
Say e.g. that you have the coordinates of a position P in user frame 1. The INVMAT instruction will turn this around, so it gives you the coordinates of the ORG of user frame 1 in the position P frame (remember: the position P is 6-dimensional, so it actually also is a frame).

The instruction would now be:
INVMAT P000 P001, where
P000 = coordinates of ORG of user frame 1 in position P frame (the result)
P001 = coordinates of position P in user frame 1

MULMAT:
This will convert a position P from one user frame to the other. Say e.g. that we have a user frame 1 with the position P in it, so P is given in 1. If we want to know the coordinates of that same position P in user frame 2 we can use the MULMAT instruction. For this to work, the offset beteen the user frames is needed. That is: we need the coordinates of the ORG of user frame 1 expressed in user frame 2 (you need to calculate that yourself).

The instruction would now be:
MULMAT P000 P001 P002, where:
P000 = position P in user frame 2 (the result)
P001 = ORG of user frame 1 in user frame 2
P002 = position P in user frame 1

So to combine things:
If you have e.g. the ORG of user frame 2 in user frame 1 you could INVMAT it first and then use it in MULMAT as P001

Hope this helps

[Robo Guru]

Thanks Jejbr84!

I'm a bit busy this week but when it gets quiet again I will do some tests for myself.  I was thinking of using MotoSIM offline to visually see the affects of what MULMAT / INVMAT does.  I'll post the results if your interested.

RoboGuru