Posts by emielke12

Quote from 95devils

Inline is not something on a Yaskawa controller.

To generate a job and dump it in, with meaningful positions, not pulse counts, the positions would need to be in XYZ. This leaves your choices with using position variables or the relative job software option.

Relative job will take a job written in pulse counts and convert the positions to XYZ, relative to something, the robot or a user frame. It will also convert a XYZ job back to pulse.

Pxxx is directly addressed. P[n] is indirectly addressed. The [n] could be a constant value or a variable.

Example.

P000 would be using the data in P000, whether that is a position to move to, or a shift amount, or an increment move. This position would be stored in the job's header. If opening or generating the job off-line there would need to be data for this address in the header.

P[0] would still be using the data in P000. The data would not be stored in the job's header. This is convenient for an integrator who dumps standard jobs into all the cells they do controllers. The programmer would have to either set up the data manually or programmatically.

P[B000] would use the data in the position variable's address that is what ever is in the contents box of B000. If B000 contents box is 0, this would be the same as P000. If If B000 contents box is 1, this would be the same as P001.

Display More

I've seen examples of what I'm trying to do:

/JOB

//NAME TEST_JOB

//POS

///NPOS 1,0,0,0,0,0

///TOOL 0

///POSTYPE BASE

///RECTAN

///RCONF 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0

C00000=500,0,100,0,0,0

//INST

///DATE 2023/12/06 10:43

///ATTR SC,RW

///GROUP1 RB1

NOP

MOVL C00000 V=200.0

END

I get an error 3190[4] when I try to load this. From what I gather, this means I need the relative job software option?

Quote

Either need to use position variables

How do I define these inline though? It's too tedious to go in to each position register and enter the values needed, and I'm also leery of doing SETE commands, since this is going to be SETE 6 times per position, so if I have 100 positions to go to, my program is now 6x as long. If this is the only option, then I guess I have to use it.... But wondering if there's an alternative way. I see that you can set either a position register Pxxx or P[n]. Are these different? Where can I specify what P[n] is?

Quote

have relative job software option turned on

This is not something I'm familiar with. Could you elaborate?

Hey,

New to motoman, but I'm trying to compare the same spline path across different brands of similar geometry/payload robots so we can confirm what we've done with one brand will work well with another. As such, we have a list of relative positions that we are moving the robot through. This works well with say, Kuka robots, because you can easily define the positions with respect to a defined user frame and in-line describe the position you'd like to move to, i.e. with KUKA:

LIN {X 215, Y 400, Z -110, A 0.0, B 0.0, C 0.0}

Is there a way to do this with motoman on a YRC1000 controller? I have 100s of points that this needs to run through, so manually teaching is simply out of the question. I'd like to be able to generate a .JBI file given just the list of points, and then be able to load it onto the controller and be able to run it, as I can do with a KUKA.

Any ideas?

I'm sending cartesian positions. I've been sending anywhere from 1mm/s up to around 200 mm/s. If I haven't generated a smooth enough trajectory, I do get jerk faults, but I am able to achieve faster motion.

I'm fairly sure they are expressed in degrees, with the 2x,4x,6x multipliers being safety factors

I ended up using the HSPO option.

Thanks for your help!

According to my co-worker (and confirmed in the documentation) streammotion requires you to send an update packet every time you listen for a packet.

Have you done this before? I've tried just opening up the socket and sending the start packet and then waiting to receive data, but I am not able to get any data back.

Ah, ok. Thanks, I'll try it out.

But when you enable stream motion on the pendant program (IBGN Start), it waits for external inputs to command motion. I need to be able to command the motion from the pendant.

I am trying to compare motion achieved by fanuc's stream motion (J519) package to the same motion utilizing a teach pendant program. When I am running utilizing stream motion, I can easily check the position at a given time and record, so that after motion is done, I can check the reported achieved path. However, I am unsure if the same can be done with a teach pendant program.

This seems relatively simple, but I have yet to see anything about this in any of the fanuc programming manuals or training manuals that we have. Is there a way to record actual position throughout a LS program (either at a fixed time interval, or at a desired point) and then be able to dump this file to grab it via USB?

I am attempting to install a new sunrise project onto my lbr iiwa 14. The project appears correct in sunrise, and I can see the correct IP in the pop-up window when you install. However, part way through the install, I get a number of errors claiming "the network path was not found" and numerous "failed creating directory 172.31.1.147//KRC/...."

Any ideas would be helpful!