See a Grown Man Cry - Corroded Prop !

[DotDun]

Originally Posted by Pete7

I am wondering (because I am not a electrician) how you could measure any leakage from the shore power circuit.

Simple, open the shore power ground and measure the potential voltage (both AC & DC) and current flow (again, both AC & DC) between the boat and the shore power.

High AC voltage/current is not good. That means the commercial power ground is missing and your boat is now the ground rod for all 'leaky' devices on the whole circuit. (aka, a 'hot' marina)

DC is caused from the battery formed by metal on the bottom of each boat with the water acting as the battery acid. Then the 2 boats connect together through the shore ground - it makes a circuit. Definition of a battery is 2 dissimilar metals electrically connected together submersed in acid. Since the metals and acid in this case are not ideal battery material, the voltage/current isn't the values you'll find in a real battery. Typical voltage is 200-400 millivolts and maybe ~50 milliamps of current.

[Dockhead]

Originally Posted by Pete7

Thats got to be worth a 10 minutes check with a meter to see if there is a good electrical connection between the anode stud on the inside of the boat and the prop shaft.

You are assuming that it's shore power doing the damage, but could it be a 12 volt circuit? I am wondering (because I am not a electrician) how you could measure any leakage from the shore power circuit.

Pete

Oh, you bet I am. Unfortunately I am about 1000 miles from the boat at the moment but I will check the shaft brush as soon as I get back. It must be non-functional; otherwise the hull anode would have protected the prop.

I have consulted with electricians and know how to check -- Ohmeter between some part of the bond circuit, and the shaft. I assume that the brush is dirty and needs to be cleaned.

I will also put on a shaft anode for a bit of belt-and-suspenders.


As to current leakage -- remember, you don't need to have current leakage on your own boat. The culprit is the ground (earth) wire. When you hook up to shore power, you bond your boat to the shore ground (earth), which is shared with the other boats in the marina. So any current leakage might come aboard your boat, if your bonding and ground (earth) is lower resistance, and eat up your zincs instead of your neighbor's.

As far as I understand, alternating current (like shore power) cannot cause electrolytic corrosion. That because (again, if I understand correctly) the current, in an AC circuit, switches direction 50 times a second, which means anodes become cathodes and vice versa 50 times a second. So metal can't be removed, since nothing remains an anode long enough (any material removed would actually be replaced during the moment when the metal object becomes a cathode (you electrical geniuses feel free to pipe up and correct me; this is an amateur's possibly naive understanding of how it works). It is direct current causes damage, and it's usually a very small current like a fraction of a volt. It doesn't have to be current leakage at all; it can be current resulting from galvanic reaction between unlike metals immersed in an electrolyte (sea water).

There must be some way of checking the integrity of your on-board circuits. Certainly, I would like to do that, if anyone can tell me how.

[Pete7]

Originally Posted by Dockhead

Oh, you bet I am. Unfortunately I am about 1000 miles from the boat at the moment .

Suggest you stay for a while, horizontal rain across the Solent this morning, wind F7 this afternoon, it's horrible

[s/v Jedi]

So, what can I do? From the questions I deduct that you're not an EE and don't know how an galvanic isolator works. I'll try to keep it as simple as possible and call the galvanic isolator "GI" from here on:

The GI is flawed by design. It's intended function is to break the ground wire between ship and shore while that wire is needed for safety. Now, someone has come up with the idea to decide for you how safe is safe enough, i.e. trading some safety for galvanic isolation in certain circumstances. An isolation transformer doesn't have that compromise: it provides increased safety while giving 100% isolation under all circumstances. The reason it can do that is because it galvanically breaks the hot and neutral circuits, coupling them by magnetic flux. The GI does not provide galvanic isolation of hot and neutral lines (only the ground line).

Now, let's look how the GI works: It started with two diodes connected anti-parallel. This means one is connected in the other direction from the other one. A diode is the basic semi-conductor. It always blocks current flow in one direction. In the other direction, it can either block or conduct, depending on the bias voltage over the diode. It starts conducting when this voltage is 0.6V (I used 0.7V before because that is the value to bias it when you use it in a different role but that goes too deep into the matter).
When you have an AC current, the polarity changes 50-60 times a second so the diode would only conduct half the time. That's where the 2nd diode connected in anti-parallel comes in as it will conduct the other half of the time.

So, this was the first GI. You basically have no ground connection until the voltage difference becomes 0.6V or more. It was deemed "safe enough" to be exposed to AC wiring as long as that difference stayed below 0.6V and the risk of a diode failing to work was deemed low enough to allow this (how many deaths vs commercial interests... tough call!).

In practice this GI failed because too often the ground faults between shore and boats was more than 0.6V meaning the diodes "open" and galvanic protection was lost. Manufacturers really wished to sell thise $1 diodes for a couple of hundred dollars so they came up with a solution: let's put two diodes in series to double that bias voltage to 1.2V. Of course, this also halves the safety because there is no ground protection until the voltage reaches 1.2V. Again, they decided that the risk of getting sued by the family of a person who died of this loss in protection was nothing compared to the profits to be made so they called it safe. Also, with twice the number of diodes, the risk of one failing doubles but WTH let's sell & become rich.

So here we had our 2nd generation of GI. At 1.2V galvanic protection, it was better than about 90% of DC offsets (which are mostly limited to 1.0V). But for some reason, many reports came up where these GI didn't work... with a negative impact on sales and profit as the result. Something must be done about that! It turned out that when you add DC and AC voltages are present on the same conductor, the AC voltage goes on top of the DC voltage (something every 1st year EE is taught in school). This means that if a 1.0V AC voltage is present in addition to just a 0.2V DC voltage, the 1.2V bias for the diodes is reached and thus all galvanic isolation is lost. Now, AC voltage made it's entrance in an ugly way.

Unfortunately for the manufacturers, by this time governments and organizations had homed in on what they were doing and considered the safety aspects. They told the manufacturers that safety had been reduced more than enough and they forbid them to add a 3rd diode into the chain. The only option left for the manufacturers was trying to keep that AC voltage away from the diodes, something that can't be done without loosing ground protection. But they came up with a solution anyway: when we connect a capacitor in parallel to these diodes, the capacitor will act as a short for AC while the diodes block DC and so we get to the current 3rd generation of GI. The problems:
1. Why didn't they do this before? Surely the EE's employed would have thought about this option before? Yes they did, but they found that an AC potential did lead to damage on metal parts (to the point of shedding paint off and producing gasses). This is not true galvanic corrosion but damaging anyway. The big bosses decided that as it was not galvanic corrosion, their GI do not need to protect against this damage. Nice huh?!!!
2. A capacitor is limited to a maximum current and will fail (explosively) if that is exceeded. But capacitors that can deal with the high currents involved are too expensive (profit!). So they decided that a small capacitor will do. When it fails/explodes it isn't there anymore and the diodes will be biased and start conducting the fault current instead.

So these units are sold. And who ever thought this: again reports came in that boats still got galvanic corrosion with these GI installed. Upon examination it turned out that the capacitors were blown up, putting them back to the 2nd generation flaws. Apparently, sometimes users of the boat made a short or the capacitor failed (so many do fail). Hmmm.... they might have thought of that before. Time to improve the 3rd generation: add an alarm system that sounds when the unit has stopped functioning. This costs more but also leads to more profit, so all is good.

I can't really understand why the governments have allowed these GI at all. And in their current design, they still lead to damage like shown by StillRaining in this thread.

When you read all above and know that there is this perfect isolation transformer option, how can you decide to go for a GI??!!! Do you think it's worth some money or bulk or weight to reduce ground safety for you and your family/crew or that the damage by AC potential differences is okay for your boat?!

Also, two examples about shore power: In Trinidad we were on about the best and most expensive yard: Peake's. Their outlets were just the 13A like in every house. We got about 80-92V from it depending on time of day. This results in increased currents to provide the same power so all equipment running at much higher temperatures. Our transformer boosted the voltage up to limit this. A GI can't do that.
We're now in this marina in Panama. The docks and everything are just like the finest marina's in the US or EU. When I stick my Fluke in an outlet, I see up to 80V AC between ground and neutral. This should be 0V. That is how good shore power is. A transformer deals with all that too because it is actually the power source for your boat and you can feed it crap and still get perfectly safe power inside your boat with full (increased actually) safety and full protection for the boat. There are no polarities, no ground connection ever etc.

Also, keep in mind that my description of the GI above is for ideal components and be sure that they are not ideal at all. Diodes can fail and start conducting "the wrong way" and capacitors fail often. And their biggest problem is that they can fail in such a way that the user is unaware of it. When a transformer fails you have no power: 100% safety by design).

cheers,
Nick.

[DotDun]

Nick,

No doubt that an isolation transformer is a better safety device than GI. One safety you failed to mention is the fault isolation breaker in the AC panel. My main breaker is 63a and trips at 30ma of fault, so from a safety standpoint, that is a fallback. (Note, it is not a GFI)

-Marc-

[Dockhead]

Originally Posted by s/v Jedi

So, what can I do? From the questions I deduct that you're not an EE .

How did you ever guess?!

Originally Posted by s/v Jedi

and don't know how an galvanic isolator works.
.

I wouldn't go that far. I have a pretty good idea.

Originally Posted by s/v Jedi

I'll try to keep it as simple as possible and call the galvanic isolator "GI" from here on:

The GI is flawed by design. It's intended function is to break the ground wire between ship and shore while that wire is needed for safety. Now, someone has come up with the idea to decide for you how safe is safe enough, i.e. trading some safety for galvanic isolation in certain circumstances. An isolation transformer doesn't have that compromise: it provides increased safety while giving 100% isolation under all circumstances. The reason it can do that is because it galvanically breaks the hot and neutral circuits, coupling them by magnetic flux. The GI does not provide galvanic isolation of hot and neutral lines (only the ground line).

Now, let's look how the GI works: It started with two diodes connected anti-parallel. This means one is connected in the other direction from the other one. A diode is the basic semi-conductor. It always blocks current flow in one direction. In the other direction, it can either block or conduct, depending on the bias voltage over the diode. It starts conducting when this voltage is 0.6V (I used 0.7V before because that is the value to bias it when you use it in a different role but that goes too deep into the matter).
When you have an AC current, the polarity changes 50-60 times a second so the diode would only conduct half the time. That's where the 2nd diode connected in anti-parallel comes in as it will conduct the other half of the time.

So, this was the first GI. You basically have no ground connection until the voltage difference becomes 0.6V or more. It was deemed "safe enough" to be exposed to AC wiring as long as that difference stayed below 0.6V and the risk of a diode failing to work was deemed low enough to allow this (how many deaths vs commercial interests... tough call!).

In practice this GI failed because too often the ground faults between shore and boats was more than 0.6V meaning the diodes "open" and galvanic protection was lost. Manufacturers really wished to sell thise $1 diodes for a couple of hundred dollars so they came up with a solution: let's put two diodes in series to double that bias voltage to 1.2V. Of course, this also halves the safety because there is no ground protection until the voltage reaches 1.2V. Again, they decided that the risk of getting sued by the family of a person who died of this loss in protection was nothing compared to the profits to be made so they called it safe. Also, with twice the number of diodes, the risk of one failing doubles but WTH let's sell & become rich.

So here we had our 2nd generation of GI. At 1.2V galvanic protection, it was better than about 90% of DC offsets (which are mostly limited to 1.0V). But for some reason, many reports came up where these GI didn't work... with a negative impact on sales and profit as the result. Something must be done about that! It turned out that when you add DC and AC voltages are present on the same conductor, the AC voltage goes on top of the DC voltage (something every 1st year EE is taught in school). This means that if a 1.0V AC voltage is present in addition to just a 0.2V DC voltage, the 1.2V bias for the diodes is reached and thus all galvanic isolation is lost. Now, AC voltage made it's entrance in an ugly way.

Unfortunately for the manufacturers, by this time governments and organizations had homed in on what they were doing and considered the safety aspects. They told the manufacturers that safety had been reduced more than enough and they forbid them to add a 3rd diode into the chain. The only option left for the manufacturers was trying to keep that AC voltage away from the diodes, something that can't be done without loosing ground protection. But they came up with a solution anyway: when we connect a capacitor in parallel to these diodes, the capacitor will act as a short for AC while the diodes block DC and so we get to the current 3rd generation of GI. The problems:
1. Why didn't they do this before? Surely the EE's employed would have thought about this option before? Yes they did, but they found that an AC potential did lead to damage on metal parts (to the point of shedding paint off and producing gasses). This is not true galvanic corrosion but damaging anyway. The big bosses decided that as it was not galvanic corrosion, their GI do not need to protect against this damage. Nice huh?!!!
2. A capacitor is limited to a maximum current and will fail (explosively) if that is exceeded. But capacitors that can deal with the high currents involved are too expensive (profit!). So they decided that a small capacitor will do. When it fails/explodes it isn't there anymore and the diodes will be biased and start conducting the fault current instead.

So these units are sold. And who ever thought this: again reports came in that boats still got galvanic corrosion with these GI installed. Upon examination it turned out that the capacitors were blown up, putting them back to the 2nd generation flaws. Apparently, sometimes users of the boat made a short or the capacitor failed (so many do fail). Hmmm.... they might have thought of that before. Time to improve the 3rd generation: add an alarm system that sounds when the unit has stopped functioning. This costs more but also leads to more profit, so all is good.

I can't really understand why the governments have allowed these GI at all. And in their current design, they still lead to damage like shown by StillRaining in this thread.

When you read all above and know that there is this perfect isolation transformer option, how can you decide to go for a GI??!!! Do you think it's worth some money or bulk or weight to reduce ground safety for you and your family/crew or that the damage by AC potential differences is okay for your boat?!

Also, two examples about shore power: In Trinidad we were on about the best and most expensive yard: Peake's. Their outlets were just the 13A like in every house. We got about 80-92V from it depending on time of day. This results in increased currents to provide the same power so all equipment running at much higher temperatures. Our transformer boosted the voltage up to limit this. A GI can't do that.
We're now in this marina in Panama. The docks and everything are just like the finest marina's in the US or EU. When I stick my Fluke in an outlet, I see up to 80V AC between ground and neutral. This should be 0V. That is how good shore power is. A transformer deals with all that too because it is actually the power source for your boat and you can feed it crap and still get perfectly safe power inside your boat with full (increased actually) safety and full protection for the boat. There are no polarities, no ground connection ever etc.

Also, keep in mind that my description of the GI above is for ideal components and be sure that they are not ideal at all. Diodes can fail and start conducting "the wrong way" and capacitors fail often. And their biggest problem is that they can fail in such a way that the user is unaware of it. When a transformer fails you have no power: 100% safety by design).

cheers,
Nick.

An incredibly eloquent defense of the isolation transformer; this thread (or at least Nick's post) should be a sticky.

But there's not actually so much here I didn't know already. I'm not an EE, but I have a good grasp of how diodes and capacitors work. I operate from the principle that the perfect is often the enemy of the optimum, especially where boats are concerned. If you spend an extra $1600 every time there is somewhat better theoretical solution to every problem, as opposed to a practical and decent one, our boats would all cost $10 million.

So the fact that an isolation transformer may be a perfect solution to the problem does not necessarily prove that it is the optimum one. If 99% of the problems are solved by breaking continuity of the ground wire, and if safety can be ensured to a reasonable degree by self-testing regimes and alarms (nothing is perfectly safe on a boat; for God's sake we keep propand on board), and if a GI is five times cheaper and much less heavy and bulky, that would make it an optimum solution in my book. The fact that diodes and capacitors can fail is not, by itself, persuasive. Through-hulls can fail, hulls can fail, rigging can fail, engines can fail. At a certain level of reliability and with certain backup plans in case of failure, these risks may be acceptable. The same with the diodes and capacitors in a GI -- we don't demand perfection from any system on our boats.

If, if, if. If 99% of problems are NOT solved by breaking continuity of the ground wire, then the logical outcome may be different. If safety is NOT assured to a reasonable degree with self-testing and alarms, then the logical outcome may be different. So I'm not saying that GI's are the optimum solution, as compared to isolation transformers, I'm just saying that it hasn't been conclusively proven in this thread -- so far -- that they are not.

I did learn a few new things here: (a) benefit of isolation transformers in boosting low voltage; (b) benefit of isolation transformers for dealing with shore power which has simply f***ed grounding (wouldn't want to swim in that marina); (c) possibility of problems from AC current leakage, which isolation transformers protect from. Hmmm. All that does add even more weight to the isolation transformer argument, which in the end I will probably be unable to resist.

[DotDun]

Originally Posted by Dockhead

I did learn a few new things here: (a) benefit of isolation transformers in boosting low voltage; (b) benefit of isolation transformers for dealing with shore power which has simply f***ed grounding (wouldn't want to swim in that marina); (c) possibility of problems from AC current leakage, which isolation transformers protect from. Hmmm. All that does add even more weight to the isolation transformer argument, which in the end I will probably be unable to resist.

Boosting voltage is not inherent in isolation transformers, it's an added cost, ~70% more than a non-boosting model.

[s/v Jedi]

GFI are just some letters but what you describe is equal to what is called GFI. Let me explain it in simple terms:

Current that flows through the hot wire, flows through the device that is powered on and back out the neutral wire. The current through the neutral wire is the same value as the current through the hot wire.
Now, what happens if the current is not equal, let's say the current through the neutral wire is less than the current through the hot wire? Well, the neutral wire is connected to ground at the transformer that feeds the power to your outlets. So, if some of the current coming through the hot wire finds another way back to ground, it can close the circuit that way without going through the neutral wire. This part of the current could have flowed through the ground wire but also through your body to the floor to... ground.

What these ground fault protection devices do is compare the current through the hot wire with the current through the neutral wire. If the difference is more than the rated value (30 mA in your case), the hot wire is disconnected (often the neutral too). So this device actually doesn't need the ground wire to work. It was designed to provide safety for all the old installations that didn't have a ground wire but it is so good that it is now used for any installation.

The reason that it is so good is that is measures exactly the thing that can hurt us: the current that potentially flows through our body. And we know that a value like 30 mA can't kill us.
A ground wire works very bluntly in comparison. I tries to make a short so that the main fuse will blow. But devices with metal housings are naturally safer when that metal is connected to ground because that ground wire has less resistance than our body (we hope the ground wire and it's connections are well done and corrosion-free). The ground fault detector is much smarter and elegant (the EU way ;-) but it could fail too. That is why that test button is there... this actually shunts a little current from hot output to neutral input to see if the device triggers.

Internally the device is simple. It is like a transformer with windings around both the hot and neutral wires (which act as the primary windings). The windings around the hot are clockwise and around the neutral anti-clockwise (or the other way around). This means that the magnetic fields cancel each other out as long as current through the wires is the same. The test button completes a circuit between hot output and neutral input with a resistor to limit the current to a safe level that is higher than the rated value (like 40mA for a 30 mA rating). The current by-passes the winding in neutral that way and the safety should trigger with the detected imbalance. Smart and elegant like the whole device itself.

ciao!
Nick.

[Pete7]

Nick, how could you 48 hours ago I set my mind on fitting a GI this Spring. Okay so GI is another $150 but I was resigned to the cost on the grounds of safety.

Now I read about isolation transformers and its going to cost £600 ($1000) its a huge great box and god only knows were it is going to fit. Additionally I now need to work out how big an IT I need by adding up all the amps we might use, batery charger, heater, kettle


Pete

[Dockhead]

Originally Posted by s/v Jedi

GFI are just some letters but what you describe is equal to what is called GFI. Let me explain it in simple terms:

Current that flows through the hot wire, flows through the device that is powered on and back out the neutral wire. The current through the neutral wire is the same value as the current through the hot wire.
Now, what happens if the current is not equal, let's say the current through the neutral wire is less than the current through the hot wire? Well, the neutral wire is connected to ground at the transformer that feeds the power to your outlets. So, if some of the current coming through the hot wire finds another way back to ground, it can close the circuit that way without going through the neutral wire. This part of the current could have flowed through the ground wire but also through your body to the floor to... ground.

What these ground fault protection devices do is compare the current through the hot wire with the current through the neutral wire. If the difference is more than the rated value (30 mA in your case), the hot wire is disconnected (often the neutral too). So this device actually doesn't need the ground wire to work. It was designed to provide safety for all the old installations that didn't have a ground wire but it is so good that it is now used for any installation.

The reason that it is so good is that is measures exactly the thing that can hurt us: the current that potentially flows through our body. And we know that a value like 30 mA can't kill us.
A ground wire works very bluntly in comparison. I tries to make a short so that the main fuse will blow. But devices with metal housings are naturally safer when that metal is connected to ground because that ground wire has less resistance than our body (we hope the ground wire and it's connections are well done and corrosion-free). The ground fault detector is much smarter and elegant (the EU way ;-) but it could fail too. That is why that test button is there... this actually shunts a little current from hot output to neutral input to see if the device triggers.

Internally the device is simple. It is like a transformer with windings around both the hot and neutral wires (which act as the primary windings). The windings around the hot are clockwise and around the neutral anti-clockwise (or the other way around). This means that the magnetic fields cancel each other out as long as current through the wires is the same. The test button completes a circuit between hot output and neutral input with a resistor to limit the current to a safe level that is higher than the rated value (like 40mA for a 30 mA rating). The current by-passes the winding in neutral that way and the safety should trigger with the detected imbalance. Smart and elegant like the whole device itself.

ciao!
Nick.

Well, ok, the Ground Fault Interrupter is actually ANOTHER electrical device I vaguely understand, despite being a mere law professor (former), and no electrical engineer. Amazing, he understands! The Russians say about a dancing bear -- it's not how well he dances, but that he dances at all, which is amazing.

But I do have one question -- doesn't that add another layer of protection, quite independent of what kind of screwing around we do with the ground wire? So even if, say, we have the dreaded open circuit fault of a galvanic isolator, wouldn't the GFI step in to protect us in case of a dangerous fault? Like all European boats, we have a GFI covering the whole boat's electrical system.

[DotDun]

If money/weight/space were no problem, I'd own an isolation/boosting transformer. I have a problem with at least one of those criteria, so I live with my galvanic isolator (no capacitor) and fault isolation breaker(s).

I feel safe on my boat.

[s/v Jedi]

Keep in mind that a 7kW isolation transformer costs a lot while the 3.6kW models come out much much better (like at $700 or so for the Victron).

On costs for boosting voltage: the Victron does a fixed 4-5% boost at that standard cost. A little more voltage never hurts while less does.

On how much problems GI solve: the new ones do not solve the problem with AC potential differences and thise are prevalent everywhere. Measuring between ground pin of outlet and the water will show.

On safety: yes, thru-hulls are tricky too. This means that they should be made as safe as possible too. You are too easy to give in safety vs cost or, you hold a different value to life than me ;-))
Also, remember that all those safety risks add up. If you have 0.1% chance of injury from the electric system and 0.1% from a thru hull: the total risk is 0.2% not 0.1%. It's just very little money to keep safety as high as possible because we are talking not about the cost of the item but about the cost difference between the good and the bad. On the scale of the whole boat it is negligible. Also, much of the safety is related to the quality of installation work done and the knowledge and discipline of the installer. It isn't just money that buys safety.

Also, you downplay the GI vs galvanic isolator difference. It is not theoretical just because I can show the theory of why the GI is inferior: there is an ample supply of examples where the GI fails utterly. The isolation transformer is both the perfect and the optimum solution.

Also, I did this before in a long thread: http://www.cruisersforum.com/forums/...ers-21355.html
which is well worth the read... just ignore the posts about parallel vs series circuits.

I did an advanced search for keywords "isolation transformer" and user id "s/v Jedi" which shows I participated in 21 threads on the subject ;-) They provide a lot more info than what is in this thread, plus examples of other installations that went bad because of not having isolation transformers.

What I forgot to mention is that an isolation transformer like the Victron can also do a step up from 120 to 240V or a step down like from 230 to 115V. This comes in handy when you sail to other nations. The new switching types of isolation transformers have the potential to do frequency conversion too but I didn't dive into that so can't tell if they already do (but would think so).

cheers,
Nick.

[Dockhead]

Originally Posted by s/v Jedi

On safety: yes, thru-hulls are tricky too. This means that they should be made as safe as possible too. You are too easy to give in safety vs cost or, you hold a different value to life than me ;-))

Well, this is thread drift, but . . .

I am not happy with the through hulls on our boat. The bottom of our boat is like Swiss cheese. Through-hulls, sea cocks, and so forth are generally reliable but it's horrible to have a sinking hanging over your head. If I were designing a boat, I would definitely do it like the Sundeers -- through hulls only in separate compartments, not the passenger one. If that adds even, I don't know, 10% to the cost of the boat, it would be worth it for sound sleep at night.

[s/v Jedi]

Originally Posted by Dockhead

But I do have one question -- doesn't that add another layer of protection, quite independent of what kind of screwing around we do with the ground wire? So even if, say, we have the dreaded open circuit fault of a galvanic isolator, wouldn't the GFI step in to protect us in case of a dangerous fault? Like all European boats, we have a GFI covering the whole boat's electrical system.

I think that what you call the "dreaded open circuit fault of a galvanic isolator" is when one or more diodes have failed. This just shows that this isn't something that almost never happens. Diodes blow easily and quick (think of alternator diodes blowing the instant the battery switch is turned).

What you write is correct but it is wrong. When safety of an installation is considered it is not an allowed practice to come up with a device that reduces risk that is introduced by another device. The risk should be avoided and safety features are only to be counted on when the unthinkable happens, not to counter probable faults that can be avoided otherwise.

A weird example: a space rocket blows up because of a leak in a propellant tank. Upon examination they find a weakness in the tank at that spot. Engineers calculate that a piece of duct tape applied in that spot would have avoided the accident. What do you think: does the next rocket go up with a piece of duct tape or with the construction flaw corrected? (they used the tape unfortunately so my example is weird indeed).

Even when a manufacturer or even my own government decides that a safety risk is small enough to allow a device or practice that introduces that risk, this doesn't mean that it is acceptable to me. Always remember that those decisions aren't made with your safety in mind! They are made for money, profit & power while maintaining a charade like if they care and did make a wise decision. What they calculate is the loss of tax income if you die and if fingers will be pointed at them when many people die because of their decision, threatening their position. They don't think about how sad it would be for your family.

If you add a 2nd safety device it should be for added safety, not to compensate for a potentially dangerous device that isn't needed.
With my isolation transformer, I don't need more safeties other than fuses/breakers because it is safe already. If I install additional safeties like gfci outlets, it doesn't increase the safety anymore.

ciao!
Nick.

[Stillraining]

Nick...could you so eloquently describe the bonding issue...that's still and elusive one for me with conflicting thoughts...it does pertain to this thread as it is also galvanic in nature.

Your a venerable book of knowledge my friend and greatly valued.