DeviceNet Interlock To PLC Input?

[I_Automation]

Working on a robot welding cell today that I have worked on a few of these machines before except this one the fixture air valves are DeviceNet, where all the others are 16 PLC outputs wired to a valve manifold. The PLC is a MicroLogix1500.



The machines have light curtains in the operator area - one vertical at the entrance and one horizontal covering the entire load area below the table height. The PLC checks both light curtain channels when controlling the valves & turntable, and sending the start signal to the robots.


One of the valves on each side is a part eject that when the welded part comes out and all the clamps open the eject piston pushes up 3 seconds then drops. These are dual coil air valves - one coil activates the eject, the second coil activates the retract after the 3 second push and a 0.75 second delay to let the other coil drop power.


This eject I just added last Friday to this machine and works in testing.


However during running the operator kept complaining the eject piston never went down. I checked the program, test cycled it, tested the valves and ran it fine.


After much digging I found out - even forcing the outputs on - that if the light curtain input is off none of the valves will do anything. If a valve is forced on, or on by the program, as soon as the light curtain clears the valve pops on. The reason the piston didn't go down was the operator walked in the light curtain during the 3.75 seconds.



Nothing in the PLC ladder shuts off the DeviceNet words for the valve manifolds or disables the DeviceNet scanner. There are also banks of input cards on the DeviceNet modules that always scan.


The only thing I can figure is there is an interlock programmed in the DeviceNet module where it itself is looking at the light curtain inputs on the ML1500 and refusing to power up the valves the PLC says to turn on.



Is this something possible in the setup of a DeviceNet scanner?


I don't have the DeviceNet software so I don't know for sure, but this seems to be the only reason.



I have only worked with DeviceNet on 2 other machines in the past and really not a fan, and this is one more reason to never use it. Plus this customer doesn't like it and has asked for a quote to replace the DeviceNet with regular IO.

 

[plvlce]

My experience with DeviceNet manifolds is that there are separate power feeds for 'control power' which runs the scanner and input modules and 'output power' which supplies any attached valves. Often the 'output power' is supplied through the safety circuit such that there is no power available when light curtains are broken, e-stops are tripped etc.

While I suppose it's theoretically possible that there's something in the scanner setup, I think it much more likely based on the description that the circuit is simply wired to drop output power to the manifold whenever the light curtain is broken.

 

[L33er]

This is not really about Device net.
You can either control the operator when to enter the cell. Typically, weld - index - open clamp - eject part - safe to enter cell green light. By adding your new sequence time frame to the 'safe to enter'.

Or you can simply replace the dual valve with spring return single valve.

 

[I_Automation]

L33er,


I would love to replace these dual coil valves with singles. Half the machines have dual coils and half singles.



Having to add latched bits, timers and logic to open the valves separately would be much easier just holding a single coil valve on.



As for trying to get an operator in a factory lately that follows instructions as directed is a problem in this and every factory and shop I get into. Things have to be made absolutely idiot-proof, except they are hiring better idiots.

 

[bill4807]

I was going to say what plvice said.
Output power to manifold is run through safety circuit