How to wire and test an isolation transformer

[RaymondR]

I suspect the neutral to ground connection came about when they used star connected, multi phase, generators where the sum of the voltages at the common node is zero for the phase voltages but could be above ground level so they grounded all neutrals to achieve a common base voltage in a multi input system.

[barryglewis]

Jedi, please continue to post your diagrams etc on the forum!

Barry Lewis

[Colin Stone]

Waking up this thread.
My Jemelelc UK transformer has the following connections:

L, N and E on the input connector. E is connected to the core shield

L, N and E on the output connecter. N and E linked together.

Then case E connection joining main case and lid.

Case is mounted on insulating strips so no part of the case touches boat metal.

So, shield to shore E. And case E to where??? Shore or boat???

[s/v Jedi]

Originally Posted by Colin Stone

Waking up this thread.
My Jemelelc UK transformer has the following connections:

L, N and E on the input connector. E is connected to the core shield

L, N and E on the output connecter. N and E linked together.

Then case E connection joining main case and lid.

Case is mounted on insulating strips so no part of the case touches boat metal.

So, shield to shore E. And case E to where??? Shore or boat???

Measure between de input E and the output E with the meter in resistance mode. There should be absolutely no reading.

This means that every conductor coming from shore terminates at the transformer, i.e. there is ni galvanic connection for any of them whatsoever. Measure all that before connecting.

The shore power inlet probably requires a breaker between itself and the transformer input terminal.

The output E is connected to boat ground. This is where the earth pins of your outlets connect as well as the ground output of the inverters.

Search for my basic reference diagram on this forum

[Colin Stone]

Thanks. But you haven't answered the question?
There is a leakage/over current etc breaker before the transformer. And as I wrote N and E are linked on the output.
Your diagram at #1 shows a Victron box. It doesn't show how/if the case is earthed? Is there a clearer diagram elsewhere??
500v insulation resistance IT case to ship ground is infinity.

[s/v Jedi]

Originally Posted by Colin Stone

Thanks. But you haven't answered the question?
There is a leakage/over current etc breaker before the transformer. And as I wrote N and E are linked on the output.
Your diagram at #1 shows a Victron box. It doesn't show how/if the case is earthed? Is there a clearer diagram elsewhere??
500v insulation resistance IT case to ship ground is infinity.

I did write that the output E terminal is connected to boat ground. I also explained what that means. Seems like a complete answer to me?

It’s fine that N and E are linked. I don’t do that but it’s fine.

In my diagrams it shows the gound of the transformer output terminal is connected to boat ground, like I said again and again.

If you connect output ground to shore ground, you defeat the isolation and risk underwater metals as well as safety of crew and swimmers around the boat. It’s better to not install it at all.

So I stress once more: ground at the input terminal connects to shore ground and this must not be connected to the output ground nor at anything else. This is crucial. Same for the other conductors from shore.

[Colin Stone]

I did write that the output E terminal is connected to boat ground. I also explained what that means. Seems like a complete answer to me? Yes, I know. And the output E is. The metal transformer case is not connected to anything and is isolated from both the core inside and ship structure. Charles Wing book diagrams 9.32 - 9.34.

It’s fine that N and E are linked. I don’t do that but it’s fine. ABYC standard is to bond output N to E, creating a polarised supply. Nigel Calder book.

In my diagrams it shows the gound of the transformer output terminal is connected to boat ground, like I said again and again. Yes, but case?? What have Victron done with the case??

If you connect output ground to shore ground, you defeat the isolation and risk underwater metals as well as safety of crew and swimmers around the boat. It’s better to not install it at all. Yes, I know that.

So I stress once more: ground at the input terminal connects to shore ground and this must not be connected to the output ground nor at anything else. This is crucial. Same for the other conductors from shore. Yes, I know!!

Still no actual answer to my original question : "So, shield to shore E. And case E to where??? Shore or boat???" I was hoping for the pros and cons of to shore or to ship for the case as Charles W doesn't discuss.

[ullar]

Originally Posted by Off Trail

This statement was important for me, thank you! The label "ship's ground" in the diagram had me thinking this was the same as ship's DC ground, which on my (and many) boats includes engine block/shaft. That this is an isolated ac ground network is an important part of the puzzle. (I assume this is correct?)

Dan

That "Ship's ground" got me puzzled too. It could be better to change that to "AC ground" or "Ship's AC ground" to be clear?


Thanks Jedi for the diagram

[Colin Stone]

The AC and DC on a boat should share the same common ground point. Some boats use the engine block. I have a copper busbar 25mm x 6mm and 60cm long which is bolted to clean hull steel. All ground cables go to this bar - isolation transformer, inverter both of which have N and E bonded. Then ground wire from AC distribution board.
All the negatives also go to this ground bar - battery negatives, alternator negatives, solar controller. Then negatives from DC distribution boards.
Then anything else - eg prop shaft bond.

[s/v Jedi]

No, that’s all half truths and incomplete or dated info.

First, this is not a shoreside installation so code is irrelevant. Even for shoreside there are special cases for isolation transformers like in operating rooms, clean rooms etc.

For boats, the ABYC is lagging far behind but is working on new recommendations around isolation transformers and so called floating power networks.

If your boat has a bonding system and AC and DC are interconnected and everything is fine then let it be like it is. Just as long as nothing connects to any of the shore power conductors.

Colin, you keep repeating you know all this but at the same time you keep asking if you should connect the case of the transformer to shore ground. When I write -nothing else connects to shore ground- then I mean that literally.

Behind the isolation transformer, everything incl. metal surfaces are safe. Unless of course you start connecting AC power to them, like you do with the neutral to ground jumper on the transformer output. This is why I remove that.

You need to look at how a circuit completes. When the transformer output has no jumper and isn’t connected to anything, the output is just an unpolarized secondary winding. If you grab one terminal with one hand, you can touch any ground incl. shore ground with the other hand and you feel nothing because there is no return path via ground as the transformer isolates you from that. This is safe.

The only reason to polarize the output is when you have single pole breakers in your distribution panel. When you have double pole breakers and an isolation transformer, you can disconnect the neutral to ground jumper in the output of your transformer, genset and disable the grounding relay in the inverter/charger.

Now the term “ships ground”… I mean ships AC ground with that. It is in the diagram: a busbar that connects all the case ground terminals, all the ground conductors to outlets and metal parts of other AC electrical systems like the aluminum breaker panel. It is not connected to shore ground, not to a ground plate under water, nothing.

So how does it’s safety work: when an appliances develops a short between its metal case and one conductor, nothing happens. With a polarized system you do this on purpose, i.e. the neutral conductor is connected to all metal parts. Now when the second conductor shorts to the case, the breaker pops, just like a conventional installation. But you can never be electrocuted via water, plumbing etc. because none of that is a return path.

Some notes with that: if one appliance shorts a conductor to its case and another appliance shorts the other conductor to its case and you touch one appliance with one hand and the other with your other hand… now you can be electrocuted. This is a very unlikely scenario but this can be prevented as well.

First, when each appliance is connected to an outlet with GFCI then this will protect you. The test button won’t work because the ground conductor isn’t a return path, but the GFCI will detect return current is different and trigger.

Second is the traditional method: install an isolation monitor with alarm. This monitor will measure isolation of the ships ground from both AC conductors. This means that it triggers when the first short between a conductor and metal case occurs. Everything keeps working but the monitor detects the fault and you start pulling appliances from outlets until the alarm stops.

These kind of installations are often done to prevent power loss due to a breaker triggering when a short occurs. Now you keep power and can find the fault without loosing power, which is crucial for environments like operating rooms, clean rooms, war ships etc.

There are very few boats wired for a floating net. You recognize them as they have double pole breakers.

[Colin Stone]

>Colin, you keep repeating you know all this but at the same time you keep asking if you should connect the case of the transformer to shore ground. When I write -nothing else connects to shore ground- then I mean that literally.

So why does the Charles Wing book "Boatowner's Wiring Manual" in diagrams 9.23 and 9.32 - 9.34 have the various options??
9.23 - shield shore, case boat. 0
9.32 - shield and case to shore
9.33 - shield shore, case boat
9.34 - shield and case to boat

3 options, no explanation. That is all I am after. If I don't understand it, then there is a fair chance the author doesn't either.

Calder doesn’t even mention a case.

[Statistical]

Originally Posted by Colin Stone

>Colin, you keep repeating you know all this but at the same time you keep asking if you should connect the case of the transformer to shore ground. When I write -nothing else connects to shore ground- then I mean that literally.

So why does the Charles Wing book "Boatowner's Wiring Manual" in diagrams 9.23 and 9.32 - 9.34 have the various options??
9.23 - shield shore, case boat.
9.32 - shield and case to shore
9.33 - shield shore, case boat
9.34 - shield and case to boat

3 options, no explanation. That is all I am after. If I don't understand it, then there is a fair chance the author doesn't either.

Calder doesn’t even mention a case.

Because as long as they aren't connected to both then from an electrical standpoint any of them work. Personally I think 9.33 is best. This ensures the case remains grounded even when not on shorepower. ABYC requires 9.33 for all that is worth.

I would point out the shield ground and possibly case ground is done internally so you don't really have any control over that.

[Fuss]

Originally Posted by s/v Jedi

No, that’s all half truths and incomplete or dated info.

First, this is not a shoreside installation so code is irrelevant. Even for shoreside there are special cases for isolation transformers like in operating rooms, clean rooms etc.

For boats, the ABYC is lagging far behind but is working on new recommendations around isolation transformers and so called floating power networks.

If your boat has a bonding system and AC and DC are interconnected and everything is fine then let it be like it is. Just as long as nothing connects to any of the shore power conductors.

Colin, you keep repeating you know all this but at the same time you keep asking if you should connect the case of the transformer to shore ground. When I write -nothing else connects to shore ground- then I mean that literally.

Behind the isolation transformer, everything incl. metal surfaces are safe. Unless of course you start connecting AC power to them, like you do with the neutral to ground jumper on the transformer output. This is why I remove that.

You need to look at how a circuit completes. When the transformer output has no jumper and isn’t connected to anything, the output is just an unpolarized secondary winding. If you grab one terminal with one hand, you can touch any ground incl. shore ground with the other hand and you feel nothing because there is no return path via ground as the transformer isolates you from that. This is safe.

The only reason to polarize the output is when you have single pole breakers in your distribution panel. When you have double pole breakers and an isolation transformer, you can disconnect the neutral to ground jumper in the output of your transformer, genset and disable the grounding relay in the inverter/charger.

Now the term “ships ground”… I mean ships AC ground with that. It is in the diagram: a busbar that connects all the case ground terminals, all the ground conductors to outlets and metal parts of other AC electrical systems like the aluminum breaker panel. It is not connected to shore ground, not to a ground plate under water, nothing.

So how does it’s safety work: when an appliances develops a short between its metal case and one conductor, nothing happens. With a polarized system you do this on purpose, i.e. the neutral conductor is connected to all metal parts. Now when the second conductor shorts to the case, the breaker pops, just like a conventional installation. But you can never be electrocuted via water, plumbing etc. because none of that is a return path.

Some notes with that: if one appliance shorts a conductor to its case and another appliance shorts the other conductor to its case and you touch one appliance with one hand and the other with your other hand… now you can be electrocuted. This is a very unlikely scenario but this can be prevented as well.

First, when each appliance is connected to an outlet with GFCI then this will protect you. The test button won’t work because the ground conductor isn’t a return path, but the GFCI will detect return current is different and trigger.

Second is the traditional method: install an isolation monitor with alarm. This monitor will measure isolation of the ships ground from both AC conductors. This means that it triggers when the first short between a conductor and metal case occurs. Everything keeps working but the monitor detects the fault and you start pulling appliances from outlets until the alarm stops.

These kind of installations are often done to prevent power loss due to a breaker triggering when a short occurs. Now you keep power and can find the fault without loosing power, which is crucial for environments like operating rooms, clean rooms, war ships etc.

There are very few boats wired for a floating net. You recognize them as they have double pole breakers.

That’s very well written, we need this clarity more on these forums.

The ABYC is lagging behind is more likely due to a understanding problem more than anything else.