Fanuc Oi Keep relay tool change bit changing automatically

I had posted about this in PM many years ago and no one seemed to have a working answer so thought I would bring it back up again.
I have a buddy I worked at Continental with message me last night just catching up and he ask me if I ever came up with a way to fix the You-ji VTL keep relay problem.

What happens is every tool change on the cross ram after it completes the relay sets back to (0). In the post we had a M0 after each tool so we could change to go back in and set it. It has been like this since the machine was installed and Absolute / Flint Machine Tool have never been able to fix it. Granted we got to where we could change it in about a minute going though all the parameter write enable process but its just a PITA. Hand changing the tools is not even in the cards. Not many people can heave a 100-150 tool over their head and many less can do it while standing on a ladder.

Its not a really big deal but I see why he would like to get it fixed. Most of their parts will use up all 12 tool slots and the changes are frequent so really no time to multi task (Flipping a 10K-15K pound part takes some time). Any suggestions I'll pas along. I'm just trying to help make his job a bit easier.

If I was troubleshooting this I would try to single block thru the tool change while observing the keep relay in question. It sounds like there may be something wrong with the macro program.

If the tool change if controlled by the ladder logic I would search that keep relay address in ever rung where it is used. I can't believe the manufacturer can't figure this out. That is very odd.

If I understand you correctly, the KeepRelay bit is changing to 0 during a toolchange but needs to be 1 for the machine to operate?

Using a bit of a KeepRelay address is a somewhat common method for a machine builder's ladder logic to flag that a toolchange is in process. The purpose being, that if the power shuts off while a toolchange is in process the flag will be read when power is restored and an alarm raised or other actions can be performed to start the recovery of the toolchanger.

In the case I think you are describing, I'm wondering if the this might be the case. Though the toolchange appears to be completed, it actually has not and therefore the logic does not process the ladder rung that would change the KeepRelay bit state back to normal.

Vancbiker said:

If I understand you correctly, the KeepRelay bit is changing to 0 during a toolchange but needs to be 1 for the machine to operate?

Using a bit of a KeepRelay address is a somewhat common method for a machine builder's ladder logic to flag that a toolchange is in process. The purpose being, that if the power shuts off while a toolchange is in process the flag will be read when power is restored and an alarm raised or other actions can be performed to start the recovery of the toolchanger.

In the case I think you are describing, I'm wondering if the this might be the case. Though the toolchange appears to be completed, it actually has not and therefore the logic does not process the ladder rung that would change the KeepRelay bit state back to normal.

Click to expand...

I have seen this before when e-stop or some other reason the tool changer did not complete the cycle. I had to manually change a keep relay after recovering the tool changer. This only happened a few times, as most of the time the keep relay was okay. I think Kevin is on to something here.

If I understand you correctly, the KeepRelay bit is changing to 0 during a toolchange but needs to be 1 for the machine to operate?

Click to expand...

The Keep changes to (0) as the tool change macro ends from what I can tell. It doesn't stop the machine from running it just alarms out at the next tool change if not reset to (1). Problem about that is for some reason the machine tool builder in their infinite Taiwanese wisdom set it up that any alarm requires a full system reboot. So, If the system reboots and you forget to change the Keep before homing there comes another alarm then repeat all over again.

As far as watching ladders (tried altering on an A2100 and cost me about $2k to fix my mistake), I stay out of that. They have a full time tech guy that builds all their robotic welding systems and seems somewhat knowledgeable about the logic so I'll send the recommendations from the responses to him.

I put a mandatory M0 in the post just before any tool change before I left Continental but some of the new guys prefer fingercam and forget to add the M0 thus loosing quite a bit of time.This is a big ass machine, the X travel is just over 9' and the Z stroke is about 60" then another 6' of cross ram travel so homing that thing can take well over 5 minutes.

Sounds like a deep look into the ladder logic is called for. I've found that trying to create a "flowchart" of what's happening helps figure out what is going on.

I've never worked on a You-Ji so can't comment on how well written their ladders are or are not. IME, many low volume smaller builders have pretty poorly written/debugged ladder logic so would not be surprised by anything that was figured out.