Many tools can be used either for good or for evil — it just depends on the person flipping the switch. (And their current level of mischievousness.) We’re giving [Callan] the benefit of the doubt here and assuming that he built his remote-controlled Residual Current Device (RDC) tripper for the purpose of testing the safety of the wiring in his own home. On the other hand, he does mention using it to shut off all the power in his house during an “unrelated countdown at a party”. See? Good and evil.
An RCD (or GFCI in the States) is a kind of circuit breaker that trips when the amount of current in the hot and neutral mains power lines aren’t equal and opposite, which would suggest that the juice was leaking out somewhere, hopefully not through someone. They only take a few milliamps of imbalance to blow so that nobody gets hurt. Making a device to test an RCD is easy; a resistor between hot and the protective ground circuit would do.
[Callan] over-engineers. He used a 50 W resistor where 30 W would do under the worst circumstances. A stealthy solid-state relay switches the resistor in, driven by an Uno and a Bluetooth module, so he can trip his circuit breakers from his smartphone, naturally.
He did find one circuit in his house with an unconnected protective earth line, so the device fulfilled its safety function. But it points out a weakness with this project as a prank device: he can only punk people who have properly installed circuit breakers, after all. Relative to the way we’ve tested our circuit breakers, this is a relatively safe project. But still, have a read through our nice guide on working with mains voltage if you’re insecure in your mains safety practices.
In my experience involving a faulty toaster, some GCFIs seem to have a very limited number of cycles.
They are usually marked with a number of cycles before you need to replace them.
OTOH if you never trip them, they might get stuck and not trip when supposed to, some has a “test” button for that reason.
Or they trip early… e.g. one we had fitted to the house here when the house was built… was supposed to trip at 20mA but was tripping at 15mA.
The upshot of this was I had to be very specific in the order in which I turned things on (e.g. CRT monitors… which are still dominant) or else the power would be cut. The faulty RCD was replaced with a 40mA model.
I had one that always triped when a USB phone charger was plugged in.
Was it one of those fifty cent ones handmade literally with a soldering iron in a country with extremely low wages that omitted almost every important safety element possible yet took the time to stamp it with a fake UL listing and other fake logos? If so, that might be a relevant reason why it happened because it was actually tripping because of the condition you were giving it. Or the GFCI could have been faulty.
I’d be throwing out that charger…
Nope, really old GCFI outlet.
They are designed to fail safe.
Tripping sooner and sooner.
At my HOA four GCFI were used at the swimming pool to protect against the underwater light failure.
Best we could figure maintenance dept never checked them.
When we finally did we found only two of the four working correctly.
The other two failed in the “always on” state. Not good!
(All four were replaced based on age alone.)
All four were different designs from different manufacturers.
One that failed had a main trigger SCR (used to drive the solenoid to open the circuit). Somewhere in the past the SCR had exploded. So this GCFI could never be trip.
The other that failed had the solennoid melted. So the plunger was fused in the plastic coil mess. Consequently, even if driven, the thing couldn’t trip.
These were fascinating failure mechanisms. Each was identified as “failed” based simply on using that “TEST” button.
My apartment complex did an upgrade and installed nearly 200 GCFI at the bathroom sinks. When I tripped mine with the TEST button it clicked, and the LED in the GCFI lit showing it was tripped. However, the desk lamp I had plugged into the outlet was still lit. Oh no!
An electrician friend pointed out that the install was a rookie mistake.
Sure enough, the apartment complex (in order to save money) had an electrician instuct the maintenance man on how to do the wiring, but never checked the results. (Should have lost his license.)
To his credit, the maintenance man had the electrician go over the instructions with him again, and re-visited each apartment. He reported that abotu 140 of the 200 had been installed backwards.
This “rookie” mistake is that the GCFI can be installed in a daisy-chain fashing so the one GCFI can provide trip protection to several other downstream regular outlets. However, if you don’t follow the instructions, and wire the thing backwards (thru the downstream), the GCFI outlet is always wired to the main wiring, and the GCFI, once tripped, doesn’t isolate the outlet.
GFCIs don’t handle reactive loads very well. Fluorescent lightbulbs, switching PSUs etc. make such a noise on the power lines that they either trip the relays without a fault, or mask a small fault for a longer time. Usually it’s the first way around, so higher breaking point interrupters have to be installed and then your life depends on whether specific appliances happen to be on at the time of your electrocution.
Iirc these things HAVE TO trip at the mentionned current (30mA here), but they CAN already trip at 50% of this value.
Be careful while testing RCDs, if there is no earth somewhere there will be voltage on some appliances! If you want to check your installation don’t touch metal devices that are class 1 and be careful what your wife/husband/kids/animals do.
Very true. Once I experienced, that at first it did not trip with the test button. Then I switched it manually and it did not want to stay on any more. I decided to omit any experiments and went straight to a hardware store and replaced it. Had to notice, that 1 phase types were much more expensive than 3 phase types. But on the package of the 3 phase type it was described, how to use it in a 1ph circuit. Basically you have to ensure, that you use the 2 current paths which are connected to the test button, otherwise the button would not work.
I test mine with a (grounded) moist dowel
How do you determine the correct resistance of said dowel?
With fingers
Just inch closer and closer to the ground point. It should tip long before you risk a short
So the next version works with and without GFCI but relies on the breaker correctly functioning to trip instead? Would need a pretty hefty dummy load and possibly even some cooling. Plus, what if the breaker is defective? Seems like not the best idea.
Cooling? Pull enough current, that the breaker trips magnetically. :-)
Friend of mine took an industrial safety class in the late 80s.. The way the instructor demonstrated a GFI was to stick the bare, stripped ends of a plugged-in AC line cord against his tongue.
Oh, I do that all the time…
(Only with semi-dead 9v batteries… And no GFI…But the tongue part at least!)
Me too! Well I stick my tongue in my Partners bare end.
Future Darwin Awards candidate… or did he have kids already?
He just knew what he was doing :)
RIP, haha.
Ouch.
Finger contact with a 240VAC line with working ELCB protection will trip the ELCB, but it will still give you a very noticable (non-lethal) jolt as it does so. Tongue contact would be an electrifying experience.
Most common are 30 mA here… but detection is from 15 mA to 30 mA max.
To test this correctly it would be nice to draw only 30 mA… prank with deeper sense…
I did a dumb prank device 2 years ago called blackout for a startup themed hackathon. IoShC (Internet of Short Circuits), afaik the first one in the world.
Basically an esp8266 (back in the day when you had to get the VM toolchain), a relay across the mains plugs and an iphone charger for esp. Made a cool looking red enclosure for it with a big ass lighting bolt across it
See when I read “Remote-controlled” for some odd reason I initially thought he made something equivalent to an EMP gun that he could point at a GFCI across the room and get it to trip because he caused enough confusion on the part of the current sensor coil….
This brings back bad memories. I really wish I could tell the story but it would probably put me in jail.
The only thing I will say is Management is dangerous. Especially in Heath Care. They really dont care about you or me or anyone just the DOLLAR. I know bin threw it.
Play with GFIs and breakers is not a good thing or safe thing. Electricity KILLS.
Even though they say to test GFI’s they should be tested at the min. They will go bad threw time.
And never keep blowing breakers they will stop working when you need them the most.
Think safe.
That is why you use a pseudonym and change nouns to other things…. lets see…
call yourself say:
BrokenVax. with email bookspace@myfacemadeup.net
and say:
My colleague Margret was using a hoover and she was almost killed by a faulty GFI in the mainsbox that my manager David Bryte demanded I to bypass.
Change it to:
My colleague was using a hoover and was almost killed by a faulty GFI in the mainsbox that my manager demanded I to bypass.
Do this over Tor and possibly over a free VPN.
Oh,and ignore anyone who call you a tinfoil-hatter for staying anomynous (As you DO have something to HIDE!!!)
I’m not quit sure were you are coming from?
As far as I know I’m not hiding.
And yes I do. I can goto jail for saying some things because I have been ordered to by the courts.
I would Luv to say a lot. But I have a family to think about.
I know I don’t wright properly, I have a lot of issues with that. And I really hate it. But you have to live with what you have.
I think that is what you mean?
But on the bright side. We can all have a good day. ;-)
Well if you had a court restraint, then fair enough. Telling those stories can
sometimes be traced back to whom wrote it, especially when the story can
only be tied to one incident. Then no amount of
anonymizing can stop such backlash.
Also some people have gotten into trouble just because someone else guessed the
likely incident an wrote their rumor to the internet. Showing the seriousnes of such situations.
As you say, better safe than sorry.
Keep well/safe.
>”They only take a few milliamps of imbalance to blow so that nobody gets hurt. ”
That doesn’t mean they won’t pass more than the set number of mA. It may take up to 300 ms for the circuit to trigger, and that’s not always quick enough. Ideally they open under 50 ms, but that’s not a given.
Getting shocked still hurts, a lot, and you can injure yourself by the shock. The point of the GFCI is that once you’re laying on the floor unconscious, you aren’t also being slowly roasted to death by the current from the faulty appliance. That applies also to other faults where a device would catch fire because of a weak short to ground.
Given that you don’t have to be near an RCD aren’t they all remote controlled?
You can buy mains socket testers that have a RCD test button on them *shrug*
The prank part is that this is remote controlled via your cellphone, so you can trip it anytime covertly..
I’d hardly call a factor of safety of 1.6 “over-engineering” in this case.
Try this in US households. We don’t need no stinking ground fault protection.
So true – why would one need protection from electric shocks if(/when?) you kill each other with firearms on a regular basis anyway? ;)
Bathroom, near sinks, and outdoors require them by code where I live.
…and additionally the NEC now requires AFCI protection pretty much everywhere else.
Well, you only have 110V (at best, it varies a lot) in your outlet, we have 230V
And BTW, the standard of most installations in the US would be illegal in europe.
Still, I like your stinking, unconfirming, mile-mesuring, farenheit backwords country with jiggly outlets where the plug falls out….
Especialy CA and other sunny states
Is there a safer way to do this?
In particular, is there a way to put only 30mA between Neutral and ground without involving the live wire? Such as a transformer in-line with an inductor that will return out of phase?
What exactly do you mean? Introducing an external (electrically separated – eg. transformer or battery) power source solely between neutral and protective earth (PE) to trigger the RCD?
If the tester is battery powered it should even be possible to test RCDs which are not ‘live’ – or do RCDs require mains to work?
In other words one can test a ‘offline’ setup.
Often RCDs also trip, if you touch the Neutral to PE
That may be because the neutral wire has a voltage in it due to I^2R losses in the cabling since other devices are pulling current through it.
That’s why “grounding” to the neutral is depreceated. Two devices with different amounts of current draw will have different voltages present at their ground terminals and a person/device in between such as an audio cable from a computer to a PA system will then pass current through its chassis between those two voltages.
Another reason is that in many installations neutral is NOT connected to ground locally. It is connected to ground at the power plant and maybe somewhere else along the way to your home, but not inside or anywhere near your home/house. https://en.wikipedia.org/wiki/Earthing_system#TT_network
You could apply a voltage between neutral and ground the neutral/earth link in the meter box will complete the circuit.
With couple of ohms circuit resistance you woukd only need a fraction of a volt to trip the circuit.
Might try that later…
Neutral and PE are not linked in the meter box in many installations. https://en.wikipedia.org/wiki/Earthing_system#TT_network
“current level of mischievousness”
I’m “shocked” nobody started a pun train, there!
Ha ha – in high school I made one in a normal transparent plastic mains plug, with a neon lamp inside, and a button that inserted a resistor between neutral and earth. We got off several lessons that way. Our science teacher though the overhead projector (what’s that!@!) was faulty after I tripped the earth-leakage breaker (aka RCD / GFCI), and he sent is back for a new one several times.
Actually the one with the resistor was my second model (after the first model, without a resistor, blew up in my hand because the earth-leakage breaker was faulty). Still got off a maths class with it though.
Brings to mind a stunt from 25 or so years ago.
My school had this one misguided kid. Was always working to one up everyone in some insulting fashion. Once changed the word during a telephone game to “necrophiliac” and laughed at everyone when they couldn’t figure it out. Basically, if you weren’t “smart”, you were just a worthless shit in his eyes. That basically amounted to all but maybe a handful of kids at the school. Made sense, I figure at least 8 of the top 10 students were assholes of a premium grade. I only ranked in the 30s IIRC. Proof that laziness got me nowhere.
But I digress, he once brought a small box he hacked together from an old walkman, power cord and a pot (probably a light dimmer) to school. I think around the 8th grade perhaps…. maybe 9th. Plug the cord in then turn the pot until the circuit breaker blew. I asked him how it worked, but the guy was a real dick about it and blew me off. Apparently, I wasn’t the only one he pissed off. By mid-day, he got caught with the device, tripping the circuit breakers in gym and was expelled. Never saw him again.
Interestingly, that stunt was what got me interested in electronics in the first place because I was so embarrassed that I couldn’t figure out what he did.
UL standards for common gfci protection-
Every GFCI must pass the following end-of- line manufacturing tests:
1. no trip below 4mA (no load)
2. must trip at 6mA (no load)
3. no trip below 4mA (with load) at rated voltage
4. must trip at 6mA (with load) at rated voltage
5. must trip with 2 ohm grounded neutral
6. must trip within 25 ms with a 500 ohm fault
7. must trip with test button at +10/-15% rated voltage
8. must not trip with noise test of GFCI Standard
9. calibration test at +10/-15% rated voltage
In America you also get to use gfpe or equipment t ground fault protection, usually set around 30 mA, used for hard wired heat cable and such things.
In regards to the pool light situation a gfci receptacle might be the wrong application for these. They are usually set up as a gfci class a device which is set up as described above, however Pool lighting installed pre 1965 can be set up on a gfci class b device which allows 20 mA of leak correct before tripping.
As of last summer all gfci receptacles in the us are now self testing, and, in theory at least, should fail safe before they no longer function properly.
If you read your gfci outlet it should sternly tell you to test monthly. Arc faults should also be tested monthly. Breakers should be cycled yearly as well to maintain proper lubricant distribution. And above all things don’t forget to change your mole detectors batteries and replace your detectors every 7 or so years. And grease your garage door opener gears once a decade, clean out your bath fan blades a couple times a year, vacuum out your dryer vents at least once a year, more if you have a family, and buy your wife flowers every couple of months- if your dating- buy her mother flowers once every 3 months.