Galvanic isolator

[Graham_Wright]

Why are they rated at 70A and more? I thought they were to protect against leakage currents.

 

[pvb]

Because if there's a fault, they have to be able to handle the high fault current.

 

[David2452]

Why are they rated at 70A and more? I thought they were to protect against leakage currents.

All of the ones I've fitted have been rated either 16a or 32a depending on the shorepower installation and I've fitted over 50, the same with isolation transformers of which I have done around twenty. Of course this assumes 240v AC, for 110v AC it would be double.

 

[pvb]

All of the ones I've fitted have been rated either 16a or 32a depending on the shorepower installation and I've fitted over 50, the same with isolation transformers of which I have done around twenty. Of course this assumes 240v AC, for 110v AC it would be double.

Lots of DIY installers are likely to look to eBay for the galvanic isolator, and most listed there seem to be 70A or more.

 

[Graham_Wright]

To further expose my ignorance, how do they work?

Is it because the diodes only conduct above their threshold voltage and that is greater than the galvanic voltages produced in sea water?

 

[maby]

In the event of a wiring fault, the isolator may need to pass several thousand amps for a few milliseconds until the circuit breakers have tripped. Obviously, none of them can manage that for any prolonged period - the current rating is just some sort of indication of how likely they are to survive until the circuit breakers are flipped. If you ever actually blow a fuse or trip the breakers, you should probably assume that the galvanic isolator is toast - even if it still seems to be conducting.

 

[maby]

To further expose my ignorance, how do they work?

Is it because the diodes only conduct above their threshold voltage and that is greater than the galvanic voltages produced in sea water?

exactly. They usually contain two diodes in series which results in a threashold voltage of around 1.2v - more than the usual galvanic voltage.

 

[Sans Bateau]

In the event of a wiring fault, the isolator may need to pass several thousand amps for a few milliseconds until the circuit breakers have tripped. Obviously, none of them can manage that for any prolonged period - the current rating is just some sort of indication of how likely they are to survive until the circuit breakers are flipped. If you ever actually blow a fuse or trip the breakers, you should probably assume that the galvanic isolator is toast - even if it still seems to be conducting.

So what happens when you press the self test on the RCD?

 

[VicS]

exactly. They usually contain two diodes in series which results in a threashold voltage of around 1.2v - more than the usual galvanic voltage.

Two in series and another pair in parallel with the first pair but facing the opposite way.

The unit may also contain a capacitor . I believe American specs also call for monitoring indicators.

Some useful reading at http://www.yandina.com/GalvInfo.htm

 

[maby]

So what happens when you press the self test on the RCD?

I have to say I'm not sure about that - don't know the mechanism of the self test button. I do know that a genuine short circuit from live to earth will pusu a short spike that can reach thousands of amps through your earth line and that is a level that will strain any galvanic isolator - you should assume that it is useless since, even if it still appears to conduct,the junctions in the diodes may be degraded to the point where it will fail on the next short circuit before your fuse or RCD trips.

 

[VicS]

So what happens when you press the self test on the RCD?

I have to say I'm not sure about that - don't know the mechanism of the self test button. I do know that a genuine short circuit from live to earth will pusu a short spike that can reach thousands of amps through your earth line and that is a level that will strain any galvanic isolator - you should assume that it is useless since, even if it still appears to conduct,the junctions in the diodes may be degraded to the point where it will fail on the next short circuit before your fuse or RCD trips.

ITYWF that it just mimics a small live to earth leakage. Just sufficient current to test that the RCD will trip when it should

 

[VicMallows]

ITYWF that it just mimics a small live to earth leakage. Just sufficient current to test that the RCD will trip when it should

It's actually a small leakage from L on the protected side to N on the input side. Earth is not connected to the RCD.

 

[superheat6k]

The GI blocks low voltage DC galvanic currents as described, but presents virtually nil resistance to high voltage AC current flow.

An RCD works like a balance scale. A current flow in any electrical circuit has to balance. So if you balance what's flowing in against what's flowing out at the circuit supply point on the Live and Neutral feeds, then and assuming there is no leakage to earth, current on the live must equal current on the neutral.

So if a fault occurs anywhere on the protected circuit leading to current leakage to earth then some of the current flowing in will find its way back to the neutral grounding point via the Protective Earth ( which is always on the supply side of the RCD, and normally at the supply source). As soon as this happens the RCD detects the imbalance and if above its trip threshold (typically 30ma) it will trip. The reaction speed has to be very fast.

 

[maby]

The GI blocks low voltage DC galvanic currents as described, but presents virtually nil resistance to high voltage AC current flow.

An RCD works like a balance scale. A current flow in any electrical circuit has to balance. So if you balance what's flowing in against what's flowing out at the circuit supply point on the Live and Neutral feeds, then and assuming there is no leakage to earth, current on the live must equal current on the neutral.

So if a fault occurs anywhere on the protected circuit leading to current leakage to earth then some of the current flowing in will find its way back to the neutral grounding point via the Protective Earth ( which is always on the supply side of the RCD, and normally at the supply source). As soon as this happens the RCD detects the imbalance and if above its trip threshold (typically 30ma) it will trip. The reaction speed has to be very fast.

The reaction speed is certainly very fast but it is still measured in miliseconds and, during that period of time, diodes in the GI that are rated at tens of amps may be carrying thousands of amps, so there is always a risk that they will become degraded. No company manufactures diodes designed to carry current at that level continuously. To keep prices down to acceptable levels, manufacturers of galvanic isolators simply use relatively high current commercial diodes and rely on them surviving long enough for the circuit breakers to trip.

 

[nigelmercier]

... Earth is not connected to the RCD.

An RCD works like a balance scale...

Which is why the earth does not need to be connected to the RCD.

You can however connect L to E via a suitable resistor for testing an RCD, this is how I made a big STOP button for my "workshop". Problem was, it would sometimes trip the main RCD in the house.