Cascading E-Stop and Light Curtain Safety Relays

Happy Holidays Folks!


I am working on a robot cell and as we are adding PLC control to the system, I have some questions about implementing an upgraded safety circuit.


Currently (pre-PLC), the robot controller has 2 sets of safety inputs, an E-stop input and a “Protective Stop” input. There are two estop relays (Euchner ESM-BA301, datasheet here). The E-Stop pushbuttons are wired in series to one E-Stop relay, whose outputs (13,14;23,24) go into the robot E-Stop inputs. The light curtain output is OSSD and goes into the second relay, whose contacts (13,14;23,24) go into the “protective stop” inputs.



We want to re-wire this system with a Cat 4 setup using redundant contactors, such that both the light curtain and the E-Stop perform the same function (uncontrolled power cut of both robot and all actuators). The contactors would cut power to a few sets of designated terminal blocks powering PLC outputs, a couple servo motors, open the robot controller safety, etc. I’ve done this numerous times using a single relay with multiple E-stop buttons, but am not exactly sure how to wire both safety relays together to accomplish this, maintaining a Cat 4 system. A single reset button would need to reset one or the other relay, and there is NC contactor feedback from both the robot controller contactors as well as the planned redundant contactors (at safety relay outputs (14;24)) so I’m guessing the safety relays need to be cascaded somehow? Does anyone by chance have a schematic they would be willing to share showing how to do this?



As an aside, the way this system is currently set up has the light curtain safety relay auto-restarting on removal of the light curtain obstruction. Something about this doesn’t sit well with me; I don’t like the idea that there is no conscious re-powering of the actuators required. Is allowing auto-restart common practice for a light-curtain protected robot cell?


Thanks!

It was a thread about Categories 3 and 4. Probably, will help you a bit;
http://www.plctalk.net/qanda/showthread.php?t=107581

Rather than wiring individual relays to coordinate functions I would suggest something programmable like the Guardmaster 44C-CR30. This would make the system more flexible and easier to modify later if needed and these are very affordable.

https://ab.rockwellautomation.com/Relays-and-Timers/Safety-Relays/Guardmaster-440C-CR30

For you estop channels if you don't already have aux inputs to the plc to identify which estop in the chain was activate then guardlink might be a nice upgrade if it's within your budget.

https://literature.rockwellautomation.com/idc/groups/literature/documents/br/glink-br001_-en-p.pdf

I also like these for control and pilots for my larger safety rated contactors in dual contactor setups.

https://ab.rockwellautomation.com/Relays-and-Timers/PCB-Pin-Style-Safety-Control-Relays

Depending what PLC system you're using, getting a fail safe variant of the PLC would make the wiring design quite a bit simpler. Wire it all into safety inputs, control the robot with safety outputs. That does all the wiring checks, and the logic can be LAD or whatever.

Modern systems allow the robot to do safety comms over Ethernet as well, making things even easier. I've seen many systems where the PLC did safety and standard, with all safety comms over Ethernet to several robots and drives, plus IO blocks. I'm assuming this isn't an option for the retrofit.

That's the latest and greatest tech, not sure how much it makes sense in a retrofit, esp in a smaller cell. The smaller programmable safety relays can make sense too, although IMO they tend to be a little clunky.

Thanks everyone for the responses. Interesting discussion about safety devices (or at least E-Stops in series) not being Cat 4.



Also, thanks for the recommendations/insight on going for a programmable solution. Budget-wise, initially we'd like to try to work with what we've got as best possible, and if this turns out to not be satisfactory, then we'll consider a programmable solution.


Thinking about this a bit, what if we took the existing safety relays, and simply ran the light curtain safety relay output contacts in series with the E-Stops? The light curtain safety relay would be wired for automatic reset, but then the E-stop relay would control the main contactors, and each device would have its own safety relay (although you still have the series E-Stops & curtain). This would require an operator reset regardless of which was tripped, and only one pair of contactors and safety relays are necessary. This would be considered a Cat 3 circuit I suppose, based on what I've read in the discussion? Would this be a satisfactory solution, assuming Cat 3 were the requirement?



The example schematic I'm attaching describes what I mean, but may not correspond 100% with the actual correct terminal markings on the relay, since i just adapted this from an old schematic I found using a completely different safety relay.

I see nothing wrong with it. This is how our company does it for a quarter century. No problems with safety inspections, no accidents.

CAT4 --> PL e

I would not wire emergency switches and light curtains -relay contacts in series to this one safety relay for different reasons. I thinks you are pushing it over the limit (specially regarding to PL e) and this for different reasons. If only because of this reason:

https://literature.rockwellautomation.com/idc/groups/literature/documents/wp/safety-wp019_-en-p.pdf
--------------------
Up to now it has been commonly accepted that interlocks with simple volt-free contacts connected in series can achieve 60% DC and up to PLd but not PLe. The TR is expected to confirm the restriction for PLe and also provide guidance for determining the maximum DC that can be achieved.
Depending on application characteristics this could mean changes for some systems. In some cases it could be possible that systems that were formerly
considered to achieve 60% DC and PLd could, in the future, be considered to achieve only zero% DC and PLc.
--------------

I would suggest using Phil Buchanan's solution.