Posts by WattUp

Quote from T STUDS

Sorry I missed this before replying,,

"Press and hold the forward key doesn't seem to work, I had to drive axis to 0 positions."

Edit, helps if if you check movement speed before saying forward key doesn't work ^^^^^

So I have the Robot in 2nd home position, all axis are at 0, and all markers as far as I can tell are all lined up, the wrist T axis doesn't have the same markers as the rest of the machine, the only thing I can see is two punch marks that do kind of line up. The torch is still twisted about 10 degrees out of square, could this be that something has twisted inside the torch mount? Is it also possible that something on the motor could have moved? I guess without being 100% sure that the 0 markings are correct there is no way telling.

Display More

T-Axis marks are probably covered by your EOAT mounting.

Are you sure the robot is 100% perpendicular to the laser at the start? It would only have to be off by a miniscule fraction of a degree for that spiral cut to be uneven as shown there.

You will need to push the offset value over some network connection, then use SETE to push those values into a position variable (P000) then use IMOV P000 to move in that direction.

Quote from RobotWelder

HI

Anybody use offline programming on here for welding robots? if so what software? how have you found it? we have octopuz but have never been able to properly align the realworld into the software.

Cheers

When i spoke to the guys at octopuz, they mentioned that they would come onsite to dial in that realworld/sim match up.

Might be worth reaching out to them to see what they can do.

But i would never trust any sim job to just run 100% no matter how close they say it will be.

I have run into this same situation before when setting up my first few FSU areas. I always programmed them to be my desired limit only to find out that in AUTO, the robot stopped far short of those limits.

Its important to remember / understand that the FSU operates separately from the robot program. FSU is monitoring current position and velocity and doing predictive math on if a hazard / restricted area is going to be violated, and will stop to prevent that condition.

It does not know that the next waypoint is actually clear of that area / restriction. So in order to reduce those false stops you need to reduce speed or add waypoints to pull the robot away from the restricted zone.

SOLUTION:

The tech enabled "SPEED LIMIT DURING PLAY" option on the FSU, but for a reason that stumped both me and motoman service, the input to disable that option does not exist on the YSU board that was installed.

Motoman was able to disable the option remotely from my CMOS.bin and send it back.

Tool weight & COG have been setup correctly.

I just pulled the CMOS from the bot and ran that in motosim... It runs 36sec also

but i did notice motosim does not populate the FSU menus so I still suspect that might play a part.

I have a project where I am utilizing an old DX100 / HP20D robot.

I have programmed the job using MOVJ 90 for all air moves and MOVL 1200 for all in process moves.

When I run the program on robot i get a cycle time of ~60sec, but when I run that job file in motosim i get 36sec.

Is there some parameter that is overriding default system max speeds? There is an FSU installed, but I did not see any speed limiting function listed with that.

Quote from Kate Harris

I know that the second home position must be done after error 4107 and I know that this position can be set manually, i.e. modified. What is its purpose and benefit? we have to have a point in space or can it commute anywhere while checking?

Second home is 0,0,0,0,0,0 position for the arm by default. it is the calibrated position at motoman when the servos were set up.

Each axis has a small arrow & scale to align the joints. this is useful when replacing motors/ gearbox. And also to verify nothing broke during a 4107, I get this fault often when a power outage hits during robot motion.

Most robots it's fine to just leave the 2nd home position, but a few of mine have EOAT that would collide at a OEM 2nd home, so i had to teach my wrist joint 90* from Zero.

Check the following menut; Robot > Work Home Position

Compare that position information to where you want it to be home. Modify & Enter if you need to update it.

Quote from Woebot

Thank you I didn't know this and just tried it. super helpful! Also worked for extended language PVars in my program (ie. P[I000]

It's a super handy trick that a lot of programmers I've worked with didn't know.

Also works for opening Call : Job statements

You can quickly edit a PVAR by highlighting the line, then hit OPEN DIRECT.

It will bring up the PVAR setting directly.

Hitting again will bring you back to the job.

RoboDk will also provide a trial license if you contact them.

I was able to test a few of the advanced features out during that time.

I would suggest to try and think of use cases before starting the trial, so that you can learn with a target goal in mind rather then just playing around.

For example, i used my trial to test the 'Path Follow' feature to test a dispensing application.

I store offset values from a baseline position per typeset.

Then use a sete to load the values into a posvar.

I have also done GETS, Invmat, Mulmat combo to get new posvar based on current position

It wasn't the best way to do it; but i have seen this done by teaching ~a dozen points along the bend. the suction cups allowed for enough flexibility to make up for exactness.

I think they taught it with vacuum off, and just kept touching up the points as it moved.

Quote from Woebot

Hello All,

Obligatory first post.

I am in the planning stages for a cell with two machine tending robots, each with a YRC1000 controller. Because there is a pretty large number of workpieces that these robots will eventually handle, I want to have a large number of positions available (maybe eventually 500+ as the customer continues to add new workpieces to their catalogue, more than I could get by just expanding the memory allotment). Because of the shape of the workpiece’s modifying the p-variables by calculations/shifts are probably not going to work. The customer and I are going to have to teach the pick/place positions for each workpiece.

I made a similar cell using a FANUC controller and my strategy was:

-Robot gets part number from PLC

-Robot runs a program specific to each part with the local positions for that particular workpiece which overwrites the global positions to be the same as the local positions taught for that part

-Robot runs does it motion using the newly overwritten global positions. The customer can copy and paste new programs and update the local positions whenever they developed a new part.

I was going to do something similar but Yaskawa local teach points are not numbered/nameable like the ones in a FANUC controller and I don’t think they can be worked on by commands like `GETE` etc.. I searched around the forum and didn’t see the same issue, I'm also pretty new to robotics so maybe I'm overlooking something obvious? Should I possibly be using a FTP system?

Display More

What about storing position values in the PLC and feeding the robot X,Y,Z,DX,DY,DZ over network?

or if parts are similar, One general job with typeset specific offsets fed from PLC data? I use this method on several cells.

I did the Octopuz trial a few years back. I was a really impressed by the quality of the program, but for my needs as a non-integrator, it was too rich for my budget.

If you changed the job name offline, ie saved as new with N++, make sure you changed the job name in the header informaton to match. (second line of JBI file)

So that is a safety system then. You need to review risk assessment and safety regulations to ensure you are setting up a safe environment.

This barrier, it is protecting a part load area? a shared space with robot operations?

Is the open/close contact a human interaction like a door or something, or is it just a sensor / status from the machine?