Normally the GFCI contains a set of two power contacts, switching the Line and Neutral. The mechanism is so, the contacts are held closed by a kind of latch, while a spring pushes them open. Once a fault is detected, the latch releases, so that spring opens the contacts.
Now if there is the "diagnostic", the contact carrier just carries another small contact and this contact controls the power to the indicator: If it is in the position the main contacts are closed, the auxiliary is open, so there is no power going to the indicator lamp. Once the thing get triggered and the contract carrier makes the main powr contacts open, the auxiliary contact closes, so turns the power for the indicator ON, so the indicator lights.

In other words it is nothing else than another contact switching just the indicator lamp, but that contact is mechanically coupled to the main GFCI mechanism, so it just responds to it's mechanical position, nothing else. So a contact in series with a small lamp...

Thats one option, but i think the aux contact is not necessary



The lamp current must be << alowed diff current. For a Neon at <1mA this would work

Quote from: Medved on June 04, 2016, 09:37:51 AM

Normally the GFCI contains a set of two power contacts, switching the Line and Neutral. The mechanism is so, the contacts are held closed by a kind of latch, while a spring pushes them open. Once a fault is detected, the latch releases, so that spring opens the contacts.
Now if there is the "diagnostic", the contact carrier just carries another small contact and this contact controls the power to the indicator: If it is in the position the main contacts are closed, the auxiliary is open, so there is no power going to the indicator lamp. Once the thing get triggered and the contract carrier makes the main powr contacts open, the auxiliary contact closes, so turns the power for the indicator ON, so the indicator lights.

In other words it is nothing else than another contact switching just the indicator lamp, but that contact is mechanically coupled to the main GFCI mechanism, so it just responds to it's mechanical position, nothing else. So a contact in series with a small lamp...

Thats one option, but i think the aux contact is not necessary



The lamp current must be << alowed diff current. For a Neon at <1mA this would work

Now I get it. And you just replace the neon indicator with an LED.
I guess I was thinking of if there happened to be a short a fault in the indicator lamp, But I suppose there are resistors to limit the current.

In Medved's option that can be a LED

In my option it cant :

LED takes atleast few mA to light up, that is allready about the same current as what it takes to trip the GFCI. In my design, when the GFCI is on it is shorting the Neon and one resistor, so the current in the other is more than before. The GFCI would trip all the time from the current in that resistor alone

For a <1mA Neon its about 1mA leakage current, so acceptable for a GFCI which trip current is 10mA (and is not precise to 10.0mA, so add or take 1mA does not make big difference)

It is doable with LED and some more electronics though

In Medved's option that can be a LED

In my option it cant :

LED takes atleast few mA to light up, that is allready about the same current as what it takes to trip the GFCI. In my design, when the GFCI is on it is shorting the Neon and one resistor, so the current in the other is more than before. The GFCI would trip all the time from the current in that resistor alone

For a <1mA Neon its about 1mA leakage current, so acceptable for a GFCI which trip current is 10mA (and is not precise to 10.0mA, so add or take 1mA does not make big difference)

It is doable with LED and some more electronics though

I have never heard of a GFCI outlet that doesn't use LED.

Actually thinking some more, there are significant problems with my circuit

 - It is not reliable. If there is a leakage between the output side and Earth, then the Neon won't light up when it is tripped

 - It lets through leakage. When it is tripped, the same 1mA of the Neon can go to the output too, that is supposed to be disconnected. less than 1mA won't harm anyone, but it may well not be allowed by standards to leak so much in tripped position (1mA is quite a lot)

 - It may not be protected well against surges. If ordinary 1/4W resistors are used, then a surge on the input may flash across the resistor so the separation between input and output may not be sufficient

So its really only the way Medved describes it

Actually if such indication would be on the line side, the ground leakage (which was, what likely triggered the device) won't affect the indicator functionality.
But even when such connection may indicate it is open in a cheap way (well, sometimes, as written before about the actual faulty leakages; I have assumed you mean something like that), definitely such leakage path is not acceptable at all.

The diagnostic contacts are one of the most common add-on features available with practically all better (so not the ones with the "being the cheapest on the market") makers of the electrical protection equipment (circuit breakers, GFCI's,...). Other common features are remote tripping (mainly for breakers; most often used with "emergency stop" buttons), sometimes even motorized arming (if the overloads and/or short circuits are normaly occuring there - e.g. open wires for some moving equipment with sliding collectors, crushing machinery motor overload protections,...).

Actually if such indication would be on the line side, the ground leakage (which was, what likely triggered the device) won't affect the indicator functionality.
But even when such connection may indicate it is open in a cheap way (well, sometimes, as written before about the actual faulty leakages; I have assumed you mean something like that), definitely such leakage path is not acceptable at all.

The diagnostic contacts are one of the most common add-on features available with practically all better (so not the ones with the "being the cheapest on the market") makers of the electrical protection equipment (circuit breakers, GFCI's,...). Other common features are remote tripping (mainly for breakers; most often used with "emergency stop" buttons), sometimes even motorized arming (if the overloads and/or short circuits are normaly occuring there - e.g. open wires for some moving equipment with sliding collectors, crushing machinery motor overload protections,...).

I have never seen anything like that...

The problem is, the way GFCI's are made in the US as simple sockets, do push them to use the cheapest possible solutions..



Those things dont appear in home grade equipment, but more in advanced (and usually bigger) breakers, that often have option to set the tripping currents/times, and other features

But even the ordinary DIN breakers have preovisions for such stuff. (dont know about US breakers). A 3 Phase breaker appear as 3 single breaker with connected handles. But inside there is a 2nd Mechanical interconnection of the tripping system. That is, when one trips, it does not take down the others by pushing their handles down, but by internally pulling the release of the spring in the others. The breakers will trip even if you hold the handle up to prevent it from going down

In a similar way, a "tripper" can be added to a breaker, which is a small electromagnet that moves that mechanical connector inside - Apply power, get trip..

Another option how to implement a "tripper" is - Thats how the emergency stop buttons in the labs at my school were done (built in 1980) - With the use of some resistor or inductor that takes >30mA and wiring it around an RCD, in a same way as a test button is

A very unsafe way to do that is to have a button which shorts the breaker out, But then you have the risk of causing a fire, Which defeats the reason for the breaker.

That oes not defeat the reasn for the breaker, but there is another problem..

Breakers wear when they trip at high currents, from the arcing as the contacts open, that burns them off, and contaminates the isolating surfaces with conductive coating. The breaker's life is rated in trips. Breakers are rated to trip ONCE at full rated current. That is, if a breaker was rated 6000A and it tripped at 6000A, you are supposed to replace it immediately, instead of resetting it. For relatively low current trips (overloads, ...) the arc is way smaller so the breaker can last for many trips

In reality breakers are chosen (when they are chosen properly) so that they are rated way higher than the expected short circuit current, so for example 6000A rated breaker on a circuit with ~1500A max current (typical figure for homes in 230V-world). So unless abused, they can last for few decades with a few short circuit events in that time

And the button, it would probably weld to permanently-on after the 1st such use...



With the RCD it is simple and safe - As RCD trips to leakage current, all you need is a button that switches on a load that takes >30mA (for reliability, lets say 300mA) between Phase and Earth. The current is small, not enough to damage any equipment, but appropriate to trip the RCD



Another version of the safety button does not rely on breakers at all. It is a self latching relay. The emergency buttons are all Normally Closed type and wired in series, and in series with the relay coil. Pushing either one cause the release of the relay. Switching it on can be done with an ON button in the panel, or by pressing the relay by hand in relays that have a built in button

A short circuit is nothing that should be there intentionally. By itseelf it means some risk the breaker won't respond and then you have a real safety problem. So shorting and/or overloading the branch in order to trip a breaker is really not acceptable.

With the RCD it may technically work, but bypassing an RCD by anything isn't allowed either. If the RCD trips due to real fault, such bypassing switch may make the once disconnected part back alive. That scenario isn't as hypothetic as it may sound, let's imagine some failure, where the phase reaches an unearthed conductive surface (it could be damaged tool,...). Then the user touches it and gets shock from it, probably the electricity "grab him" (the uncontrolled muscle tension makes it impossible for his will to release). After some time (a second or even less) the RCD trips and disconnects the power (that is it's purpose). But in the meantime he screams, so someone pushes the STOP button. And if that would bypass the RCD, it would mean that poor guy will get another jolt, this time from the tripping circuit from the button.
So no way to make such bypass except really the TEST button inside of the RCD (I think this has a separate contact in series with the test circuit, so once the RCD has tripped, even the TEST current can not flow; but I'm not 100% sure if that is mandatory for the internal RCD design, the motivation could well be to protect the test tripping resistor from overheating due to external current leakages - like miswired Neutrals or so)

The latching relay has one drawback: It consumes steady power (so heats up) and because it will be permanently energized, the core may get magnetized so, it may not release even when the current is interrupted. I've seen this approach on circuits, where the relay is used for normal power control of that device (start-stop control of a machine,...)
The breaker tripper consumes no power, while it is all the time ready to act, so it is used mainly on circuits, which are permanently powered.
Of course, when the button functionality is really exceptionally important, both ways are combined (have seen once, combined with a cover door interlock contact)

I think the opening contact that disconnect the resistor inside RCD indeed is there to save the resistor if the RCD is not tripping. And it does not een need to be a miswired circuit..

The test resitor is conenced across the RCD diagonally. The RCD can be wired with supply from below or supply from above - It works both ways, and in different countries there are different norms - In Israel and i think most of Europe it is from below, while in Russia (and maybe somewhere else too) it is from above. That means that in one setup the test connects the resistor to L-out N-in, in other setup to L-in N-out

Now imagine that the RCD is wired with the resistor on L-in, there is a fault in the circuit that the RCD tripped for (short of something to PE), then for whatever unknown reason, the user push Test. (nevermind that the RCD allready is down). Without an opening contact that would go up in smoke..



The version with the "test button"-like wiring is when it takes L-out to N-in. This way, pressing the button when the RCD is tripped only connects N-in to the circuit, so no voltage there, and no danger of that 2nd shock scenario (unless there is some other fault with upstream N being disconnected)

The other version is similar but with the button connecting a load between L-out and PE, not L-out and N-in



Last but not least, making a hard short circuit is legitimate way to break the circuit - if such act is justified. For example to disconnect the circuit if somebody is getting a shock, paperclip on plug pins and push the plug into another socket on the same panel (a good short circuit usually trips the main too if it is not set too high) may be the fastest thing you can do..

But even the ordinary DIN breakers have preovisions for such stuff. (dont know about US breakers). A 3 Phase breaker appear as 3 single breaker with connected handles. But inside there is a 2nd Mechanical interconnection of the tripping system. That is, when one trips, it does not take down the others by pushing their handles down, but by internally pulling the release of the spring in the others. The breakers will trip even if you hold the handle up to prevent it from going down

Some standard circuit breakers here function that way. Some even have a single handle but a double-width body and are internally mechanically interconnected. I have one here that is as you described only double pole. However, it is two distinct breakers that are only connected with the handles. The connecting rod is so loose that one can trip without the other (I miswired something plugged in a duplex split receptacle and found out).

Regarding tripping when the handle is held in the on position, here we have breaker "locks" that prevent the handle being moved to the off position. It is usually removable with a screwdriver so not a proper lock per se but prevents people from accidentally cutting power to emergency circuits (exit lighting, fire alarm panels, etc). A screw driver is necessary to reset them if they trip.

Yes the whole inside my house is led even the 12 bulb Chandler but I use 40w clear incandescent in the 2 front door lights and 3 40watt clear candle bulbs in light post seeing my neighbor pays the bill on those I leave on all night But my work shop I use 96" pgs and preheat I know it kind defeats the purpose of saving electricity but I am not in there to often.

Yes the whole inside my house is led even the 12 bulb Chandler but I use 40w clear incandescent in the 2 front door lights and 3 40watt clear candle bulbs in light post seeing my neighbor pays the bill on those I leave on all night But my work shop I use 96" pgs and preheat I know it kind defeats the purpose of saving electricity but I am not in there to often.

You use Power Grooves?
You should save them, There are only so many in the world, Unless they have a reproduction.