Safety relay - working principle

Safety relay (Emergency stop) is designed in such a way that any faults in the input circuit are detected, e.g. contact welding on an emergency stop pushbutton or on one of the safety contacts on the output relay.

How it's possible to differ normaly closed contact from welded contact?
On which way Safety relays "feels" this difference?

Dominic10 said:

Safety relay (Emergency stop) is designed in such a way that any faults in the input circuit are detected, e.g. contact welding on an emergency stop pushbutton or on one of the safety contacts on the output relay.

How it's possible to differ normaly closed contact from welded contact?
On which way Safety relays "feels" this difference?

Click to expand...

Redundancy. The timing of the changes of states of redundant pairs of contacts must be within a very tight tolerance to be considered free of fault. It is highly unlikely that a contact will weld on a push button carrying very little current, but if one contact does become mechanically "stuck closed" it is also highly unlikely that the other contact on that same switch will also fail at the same time. The safety relay will then detect that one input string changed states and the other did not, regard that as a fault, and disable its outputs.

Generally E-Stop circuits are n/c, so you mechanically separate them, so they can't weld shut.

The biggest danger is the contacts fall off the mushroom and nobody notices until they need to operate it.

janner_10 said:

The biggest danger is the contacts fall off the mushroom and nobody notices until they need to operate it.

Click to expand...

This is why the US automotive industry requires an extra contact that essentially reads true when the estop is correctly assembled.


Everywhere else, the best practice is to check all safety circuits once a shift/day/week to attempt to detect these sorts of faults.

janner_10 said:

Generally E-Stop circuits are n/c, so you mechanically separate them, so they can't weld shut.

The biggest danger is the contacts fall off the mushroom and nobody notices until they need to operate it.

Click to expand...

I always use a self monitoring block

As mentioned, dual redundancy and timing are how you tell if a contact is 'welded' or forced closed.

Additionally, some safety systems use a pulse wave to tell if all items in the circuit are 'ok'.

This might help explain your question


Safety Relay: How Does a Forcibly Guided Contact Work




http://www.omron-ap.com/service_support/FAQ/FAQ02481/index.asp

janner_10 said:

The biggest danger is the contacts fall off the mushroom and nobody notices until they need to operate it.

Click to expand...

I see this scenario only of the E-stop is the kind that is placed in an enclosure that has room enough so that the contacts may not always be held in place.
In the self-contained boxes with E-stops that are usually placed in the field, the design is such that everything is either in place, or is so much damaged or wrongly assembled that you instantly can see it from the outside.

rQx said:

I always use a self monitoring block

Click to expand...

How does that work ?
If the mushroom is not mechanically connected to the contacts, does the contact switch over by means of an internal spring force, or ... ?
Am genuinely interested. Could be something we can use.

OkiePC said:

Redundancy. The timing of the changes of states of redundant pairs of contacts must be within a very tight tolerance to be considered free of fault. It is highly unlikely that a contact will weld on a push button carrying very little current, but if one contact does become mechanically "stuck closed" it is also highly unlikely that the other contact on that same switch will also fail at the same time. The safety relay will then detect that one input string changed states and the other did not, regard that as a fault, and disable its outputs.

Click to expand...

Thanks for your reply, but I still can't imagine how it's possible. If we have NC EMstop button and if we press it, in the case when the contacts are welded there will be no opening of the contacts. So, in such situation how the safety relay will now that something is wrong?
Ohhh...in the case when both NC contact are welded, the relay can not "feel it"?

You have 2 of them (redundancy). Could be in an antivalent setup (NO+NC). The likelihood of them both welding is low. If you monitor the discrepancy time between them you'll know if something's wrong.

JesperMP said:

How does that work ?
If the mushroom is not mechanically connected to the contacts, does the contact switch over by means of an internal spring force, or ... ?
Am genuinely interested. Could be something we can use.

Click to expand...

Jesper, those "self-monitoring" NC contacts are held closed only when they are mounted to the mushroom. If they fall off OR if the button is pressed, the internal spring forces the contacts open.

rupej said:

Jesper, those "self-monitoring" NC contacts are held closed only when they are mounted to the mushroom. If they fall off OR if the button is pressed, the internal spring forces the contacts open.

Click to expand...

Exactly. It is like a 3-position switch.

It 'makes' in the middle when attached to a button, and unattached it breaks and pushed-in it breaks.

The only issue is that I have had some flake out and read open when attached. It is still safer than direct contacts, however, and those instances have been few and far between.

Dominic10 said:

Thanks for your reply, but I still can't imagine how it's possible. If we have NC EMstop button and if we press it, in the case when the contacts are welded there will be no opening of the contacts. So, in such situation how the safety relay will now that something is wrong?
Ohhh...in the case when both NC contact are welded, the relay can not "feel it"?

Click to expand...

You're correct if both sets of NC contacts weld, it won't work. But what is the probably of both them becoming welded at the same time?

The idea of a safety relay or safety switch is that while a contact may weld, it can be detected when the switch or relay is used by comparing it to the second contact. So, on the first failure it must be replaced.

This is also why there is a testing requirement on switches and relays as well. If a switch or relay is not normally used it must be tested at certain intervals (determined by the safety requirements) to see if one contact has welded (or has another failure). This reduces the likelihood of having two undetected failures before the next demand.

JesperMP said:

How does that work ?
If the mushroom is not mechanically connected to the contacts, does the contact switch over by means of an internal spring force, or ... ?
Am genuinely interested. Could be something we can use.

Click to expand...

We use these almost exclusively, because the safety validations we have done on most clients require them. Essentially, we need to be able to detect that the contact block has fallen off the back of the pushbutton head, or else the e/stop looks just fine from the outside but doesn't work. That reduces your diagnostic coverage significantly, because a failure like that doesn't just take out one channel and get detected next time someone tries to use the e/stop, like a welded contact might, it prevents the safety function from working and the fault from being detected.

They work similarly to a lanyard - on a safety lanyard, the contacts are only closed if there's exactly the right amount of tension on the rope. Too much (rope pulled) and they open, too little (rope broken) and they open.

In the out-of-the-box state, the contact is open. The little yellow "fingers" on the contact block are taller than the fingers on all the other pushbutton contacts, so when you put them on the back of an e/stop, they get partially depressed and close. If you press them further in again, they open again.