Posts by Nation

Yeah, they were trying to run style 35, but the robot would do this for any style you sent it.

If you right click on the group of the robot under the tree, you can output it as a .iges file. This file will be significantly reduced in quality however. More jaggy.

Yesterday, I got called out to a customer of mine with a weird issue on one of their old robots they use in their lab. An old R2000iA/165F running SpotTool+ 6.40.

They were trying to use the style method of starting the robot, but every time the PLC sent a start signal, the robot would reject the start (SYST-011) with the reason of SYST-073, which according to the error manual states:

Quote

SYST-073 Manual selection mismatch

Cause: The style input does not match the manually selected style.

Remedy: The PLC must send a style code that matches the manual selection.

Which is very descriptive and helpful.

What does it mean manually selected style? Where is this set? The style program exists, was defined in the style table, and was enabled. The bits for style selection were coming across from the PLC fine as well.

There are several ways to remotely write to a PR. Are you sure a PLC or other device isn't doing it?

That, or write a TP program that loads the PR with known good vaules before use.

Keep in mind that those variables will not update if the user jogs the robot. They only update when the robot is running a program.

Not to my knowledge. From what I have seen, the robot modifies the distance of it's wrist center to faceplate, defined by $MRR_GRP[1].$DH_D[6]. Without the flange, it is -175, with it, it is -198.

That variable is write protected.

Yes. That model was the one I had issues with in my original post.

Playing with insulated flange options in roboguide, you get a null user tool move of 23mm when changing from non insulated to insulated.

Quote from HawkME

I tested this in RoboGuide and it works fine, but haven't tested on a real robot to confirm.

I can confirm this works in the real. I implemented exactly this in a cell and it has been running fine for over a year.

Is that really an Arc Mate M20iA? That wrist looks a lot different from one.

How modern is your robot? You could just display the teach pendant on a web broswer if it is running V7.7 or higher.

Quote from SHIFT_Lock

This seems to be a frequently asked question on here and although I don't have any robots that require KFloppy maybe we could make this a sticky thread where the download is easily available? As long as it isn't against any rules of the forum that is.

Yeah, I'll probably move this thread to the tools subforum after a couple of days.

Here you go. Never used it myself. Good luck.

No way to directly grab karel vars from tp. You would have to write a karel "getter" routine that could grab the variable you want, and write it to a register or PR that TP can access.

I feel like this may be a case of Hanlon's Razor.

Quote from Wikipedia

"Never attribute to malice that which is adequately explained by stupidity."

With that said, Not right out of the box. You would have to write a karel routine to monitor the background variables, then run that in the background.

There is also the logbook.ls if your robot is setup with it. It captures screen shots and program changes, with time stamps. Just make sure the robot's time is set right, as they are set to Japan time when they ship.

Another option would be to take a ftp backup every 5 minutes or so, and then run a diff command on the previous vs the current, and see if anything changed.

You can disable group motion under the test menu. The robot will run as if it is moving, but no real motion will happen.

You can also enable 0% speed override in system config, and dropping the robot to 0% speed will stop it. You have to be on V9 or higher for this though.

It is possible to crash the ftp server on the robot without the rest of the robot going down. I've done it a few times. A reboot does fix it.

Another option may be to restart the service on robot, which is under host comm, then servers in the details tab. I haven't tried this though.

If you get them into ASCII format, and you have the ASCII upload option on your new robot, you should have minimal issues loading the programs in. The syntax of Fanuc programs hasn't changed much.

MotionPro should be able to record individual joint positions and export them as a .csv. I've only ever used it on real robots though, but I see no reason on why it couldn't be hooked up to a virtual robot. It is a separate license though.

Regarding another TCP, no way to do that, but would the tool trace work? It draws the tooling CAD at every point. Maybe replacing your tooling CAD with different colored small spheres at each point you want to monitor would work. DCS trace would be an option as well, placing small spheres (points in DCS) where you need them, but it would be harder to track what is what, since everything is green. The CAD replacement idea would only work for tooling cad attached to J6, but the DCS idea would work for points attached to any part of the robot.

Final question, why do you need to do this?

I use a .cm file to setup reference positions for touch up on new installs.

! Setup reference positions
KCL SET VAR $REFPOS1[1].$COMMENT = 'Home'
SETVAR $REFPOS1[1].$HOMEPOS 1
! Perch tolerance is in radians.
! Set to 1 deg.
SETVAR $REFPOS1[1].$PERCHTOL[1] 0.01745329
SETVAR $REFPOS1[1].$PERCHTOL[2] 0.01745329
SETVAR $REFPOS1[1].$PERCHTOL[3] 0.01745329
SETVAR $REFPOS1[1].$PERCHTOL[4] 0.01745329
SETVAR $REFPOS1[1].$PERCHTOL[5] 0.01745329
SETVAR $REFPOS1[1].$PERCHTOL[6] 0.01745329
KCL SET VAR $REFPOS1[2].$COMMENT = 'Service'
SETVAR $REFPOS1[2].$DOUT_INDX 51
SETVAR $REFPOS1[2].$ENABLED 1
SETVAR $REFPOS1[2].$PERCHTOL[1] 0.01745329
SETVAR $REFPOS1[2].$PERCHTOL[2] 0.01745329
SETVAR $REFPOS1[2].$PERCHTOL[3] 0.01745329
SETVAR $REFPOS1[2].$PERCHTOL[4] 0.01745329
SETVAR $REFPOS1[2].$PERCHTOL[5] 0.01745329
SETVAR $REFPOS1[2].$PERCHTOL[6] 0.01745329