Cell Interface I/O pulse length

Hi everyone,

I'm working with one of our PLC controls guys on sending a weld complete output to the PLC with a GO to keep track of welds expected and weld completes in a spot welding cell. The current weld position in the sequence (1, 2, 3, etc) will be sent in GO then once the robot receives the weld complete DI from the weld timer, DO135 One Spot welded (set in Cell Outputs line 16) will turn on then off when the robot receives the weld complete from the weld timer. Once the PLC receives the weld complete (DO135), it sends back a weld complete acknowledgement (DI135) to let the robot continue, otherwise it will wait. The problem we are having is that the weld complete input from the weld timer toggles so quickly that sometimes you can't even see it and PLC sometimes doesn't see it. We have other cells in our building that use this same logic, but on those robots, the one spot welded output seems to pulse for about a half of a second consistently, as opposed to pulsing for the same amount of time as the weld complete input from the weld timer.

I have to believe that somewhere there must be a variable that can adjust the amount of time that this output pulses, regardless of how long the input pulses for, but I can't find anything.

I will appreciate any input on this as I have been racking my brain for a few weeks on this. Sorry if some of my wording isn't very clear, trying to get this typed up quick.

Thanks!
Steven

Quote from BigFrank612

Have the Weld controller send its weld complete output to the Robot and then interconnect that input to an output going to the PLC. You can also look at the weld timer software and see if you can pulse its outputs for a period of time. 3rd option see if you have BGlogic option installed and check the robots that are working correctly to see how and if they are using it.

Sent from my SM-G900P using Tapatalk

We tried doing an interconnect but I can not find anywhere in the weld timer to adjust the weld complete output duration. It is a Bosch Rexroth weld timer if anyone is familiar with them. We don't have much experience with them here. All of our older stuff uses WTC/Medar. I have looked into BGLogic but on the cells where it works fine, the only BG logic is for the water savers and for Step Mode. As far as I can tell, I have everything set up identical to the working cells but the only difference is the One Spot Welded pulse duration. Even though the weld complete from the weld controller pulses so fast you barely see it, the One Spot Welded output in the cell outputs pulses for probably 250ms. In the new cells, the One Spot Welded output always pulses for the same amout of time as the wc input from the timer, just like any DI-DO interconnect would. That's why I think there may be a variable somewhere to set the pulse time of the output.

Quote from djogf412

Would you mind clarifying a couple things?

Robot model/ controller, and could you be a little more specific as to how you are passing weld done input to the plc? For example, you mentioned that the robot receives weld done signal, then turns an output ON then OFF to communicate to the PLC.. could you be a little more specific about how the controller handles this logic?

Big Frank is on the right track here, but I think we are missing some information.

How did you set up the interconnect?

The controller is the R-30iB and the robots are R-2000iB 210F with SpotTool+.

There are I/O sections called Cell Interface and Weld Interfrace which are part of the software to make communicating with a weld controller and PLC a little easier.
Specifically, in Weld Interface Inputs, there is one that you specifically assign as your Weld Complete input and in Cell Interface Outputs there is a One Spot Welded that you assign to pass the weld complete signal to the PLC.

We have cells that are programmed like this already from one engineering company that works great. These cells have been here for several years. We now have a handful of new cells that were engineered by a different company and all of the weld complete tracking is handled by the robot. We are trying to set up our new cells to handle weld complete data the same way as the other cells to ensure that parts do not make it through the line with out having all of their welds.

So here is the sequence of events:

1: weld position GO to PLC
2: fire weld and receive weld complete DI from weld controller.
3: send the One Spot Welded DO to the PLC to say that the weld was done.

In both the older cells and the newer cells, the weld complete DI pulses so fast that you don't ever actually see it turn on on the pendant. If you directly interconnect the DI to the DO, it pulses so fast that the PLC doesn't actually see it come on.

When looking at the older cells, the weld complete input turns on then off very quickly, but the One Spot Welded output always pulses for a good 1/4 of a second which is good for the PLC.

This is where something is different between the older cells and the newer ones.

In the new cells when I set everything up the same but the One Spot Welded output will only pulse for the same amount of time as the weld complete input.

What I am thinking is that there are variables to set how long you want the cell outputs to pulse to allow for more reliable communication with the PLC.

I'm pretty sure we have it figured out now. Fanuc's support confirmed that the One Spot Welded output will stay on until you run their HOME_IO program, or add an instruction to a program to turn it off.

After reading this, we tried a few more things and adding waits in the robot and timers in the logic to see exactly when each bit was turning on and off.

Everything was working as it was supposed to except for the One Spot Welded output still turning on then off really fast. We did a little more digging and found an interconnect that was forcing One Spot Welded to stay off. After getting rid of that interconnect, everything is working great.

I was searching for something I would never have found and it ended up being such a simple mistake haha.

Thanks everyone for the input though!