Isolating Transformers, The Earth wire connection conundrum

[transmitterdan]

Originally Posted by goboatingnow

(3) one live faulted to appliance A and the other live faulted to appliance B. a person touching both appliances will be inserted into the circuit and receive a shock. No protection will be extended. However this is a very unlikely fault situation.

If a contentious owner has multiple such crummy appliances and they are still extremely worried about safety then individual GFI circuits for each appliance will solve this problem.

[Oceanride007]

Dave, I think pbijim and chala are talking about your scenario (3) above.
You dismiss it as unlikely as does s/v jedi, but I see that a single fault could exist for years (cose as you say there is no current flow but there is potential, and as discussed no one uses detection equipment to find it) then another fault occurs thats either opens the circuit or chala and pbi concern is that it is also isolated until a hapless person makes contact with both items and bang.
You say they use these circuits in hospital, in a way its used on ships as well but phase to phase on the LV circuit, but we monitor breakdown of insulation as a priority. You are not proposing monitoring faults are you?, I don't blame you for not liking your regs, I'd be concerned about electrolysis issues as well, and I guess if I don't get comment about my artificial earth idea, I'll do my own experiment on RCD protection as I discussed above.

[goboatingnow]

Originally Posted by Oceanride007

Dave, I think pbijim and chala are talking about your scenario (3) above.
You dismiss it as unlikely as does s/v jedi, but I see that a single fault could exist for years (cose as you say there is no current flow but there is potential, and as discussed no one uses detection equipment to find it) then another fault occurs thats either opens the circuit or chala and pbi concern is that it is also isolated until a hapless person makes contact with both items and bang.
You say they use these circuits in hospital, in a way its used on ships as well but phase to phase on the LV circuit, but we monitor breakdown of insulation as a priority. You are not proposing monitoring faults are you?, I don't blame you for not liking your regs, I'd be concerned about electrolysis issues as well, and I guess if I don't get comment about my artificial earth idea, I'll do my own experiment on RCD protection as I discussed above.

As transmitterdan has said. If you want to protect From two appliance faults or sceanario. (3) then fit RCD sockets


I don't like ABYC regs in this regard as I think its outdated ( and they are not actually regs) in European boats fully floating is a acceptable under the RCD ( cE ) directive which is a regulation
Dave

[s/v Jedi]

I have been writing to use gfci outlets to prevent any of the scenarios incl. the very unlikely, since way back in the thread. The only ones recognizing it are Dave and Dan, who obviously are engineers too. The rest still needs the quarter to fall, or are ignoring it because it would kill their argument they love to repeat

So I repeat, a gfci outlet will interrupt the circuit when current leaks away to anywhere else than back into that same outlet. This includes boat grounds, metal boat parts, alien conductive species etc.

[pbiJim]

Originally Posted by goboatingnow

I don't like ABYC regs in this regard as I think its outdated

The ABYC standards seem to be getting better, but I agree that they are not yet where they could/should be. I think that they are now being held back by NFPA standards that regard any ground fault equipment that trips above 5ma as unsuitable for protecting people. I am not supporting this position as being correct, I am only stating that I think this particular paperwork issue is one of the things that now holds ABYC back from further improving their standards. If ABYC had the political courage to stand on their own two feet without riding the coat tails of NFPA, they might move forward a little faster. Unfortunately, the US is a particularly litigious society these days & that would put ABYC at risk in the event that a 30ma ground fault device failed to prevent a death or serious injury.

As has been noted elsewhere, the 5ma devices have real-world functional issues when used to protect much more than 1 or 2 individual outlets.

Dave, Jedi & others have posted compelling opinions that were well supported with logical data in favor of using the 30ma devices on main feeders to boats. What I am writing here is not a rebuke of that position. I am just stating what I perceive to be the reason that ABYC is still dragging their feet on certain issues such as this & grounding.

[sailorchic34]

Originally Posted by pbiJim

Apparently, I am still not making myself clear.

In my example, Appliance #1 has a fault from L1 of the ac system to it's case, which happens to be touching a metal part of the boat. This causes an unintentional grounding of L1 via a fault.

Appliance #2 then has a fault from L2 to an exposed surface.

Together, these two faulty appliances cause a dangerous condition, despite the added protection of the isolation transformer.

I was not talking about shore ground. I was only talking about local ground on the boat itself, such as an engine block, keel, or other part that happens to be metal.

I understand your thought. But where is the circuit path to get back to L2. There is no ground path. Yes your can touch leg 1 to a metal part, but there is no conductive path back to Leg 2, so no current flow.

Without an isolated transformer you have a problem, agreed. But that's why its called an isolated transformer. The impedance is too high to get voltage back to earth from either leg of the of the transformer coil. There is no current flow unless both legs are in circuit.

[sailorchic34]

Just to clarify or maybe confuse the issue more. If the transformer had a center tap that was at 0V potential to either leg and that point was tied to ground / earth. Then you would have all sorts of ground loop potentials. This is how every dockbox and step down transformer on power poles are wired in the us btw. There is a ground loop because the center tap neutral on the secondary side is tied to ground as is the power plant alternator

But get rid of the center tap and ground and you have a floating circuit, without a low impedance ground path. A Isolation transformer only makes power when both ends of the coil, that is leg one and leg two are shorted together. Then the circuit path is only from leg 1 to leg 2 and never from either leg to ground.

A GFI on a isolated transformer does not protect as the current flow is always balanced between L1 and L2 without leakage.

Yes if you grab L1 and L2 on an isolated transformer you will complete the circuit and you will get shocked. A GFI will not trip as there is no leakage to ground.

But if you say grab L1 which is in circuit with L2 say with a lamp. There is no shock potential as there is no path to ground. Grab L1 and L2 and your in circuit.

[Krawdad]

This is a great informative thread and I have been following it with keen interest! I know you experts on the subject might get a little irritated explaining things over and over again.... but in doing so, it puts your points across from every possible angle, one of which will invariably sink in to those of us who are ignorant of the subject but eager to understand.

From my perspective, it's still as confusing as hell but your time has not been wasted. Thanks!

[sailorchic34]

Originally Posted by pbiJim

Apparently, I am still not making myself clear.

In my example, Appliance #1 has a fault from L1 of the ac system to it's case, which happens to be touching a metal part of the boat. This causes an unintentional grounding of L1 via a fault.

Appliance #2 then has a fault from L2 to an exposed surface.

Together, these two faulty appliances cause a dangerous condition, despite the added protection of the isolation transformer.

I was not talking about shore ground. I was only talking about local ground on the boat itself, such as an engine block, keel, or other part that happens to be metal.

BTW agree totally that per your example, you have a fault between L1 and L2 and it could be shocking.

I'm thinking that using a floating ground wire to a center tap of the IT would solve that and allow either breaker on either leg to trip. It also gives a current path for a GFI, RCP from either pole.

The original OP diagram with the "ground" tied to one leg, is not really a ground, but one side of the hot circuit path. Most of the time it will do the same task as a real ground wire. I'm thinking it could have some issues. Such as if the Normal leg that "ground" connects to for some reason fails you still have a direct current path between the opposite leg and the "Ground" wire that still closes the circuit and may not be protected.

Why I think using the center tap as a 0V reference ground might be handy. Still pondering that one.

[goboatingnow]

Originally Posted by sailorchic34

BTW agree totally that per your example, you have a fault between L1 and L2 and it could be shocking.

I'm thinking that using a floating ground wire to a center tap of the IT would solve that and allow either breaker on either leg to trip. It also gives a current path for a GFI, RCP from either pole.

The original OP diagram with the "ground" tied to one leg, is not really a ground, but one side of the hot circuit path. Most of the time it will do the same task as a real ground wire. I'm thinking it could have some issues. Such as if the Normal leg that "ground" connects to for some reason fails you still have a direct current path between the opposite leg and the "Ground" wire that still closes the circuit and may not be protected.

Why I think using the center tap as a 0V reference ground might be handy. Still pondering that one.

As pointed out L1 to L2 faults are easily addressed by appliance or socket based RCDs. And L1 L2 fault in a single electrically continuous appliance will merely short and trip a breaker.

The conundrum. Is that using a centre tap ( and a centre tap protective wire ) , you can establish a fault path that inadvertently adds ground/earth. Once earth is in the circuit , shock hazards multiply , and reliance on safety moves from the IT to the protection devices. However centre tapped earth referenced IT has the advantage of halving the shock voltage relative to earth.

Many people this side of the water , will have seen the 220vac to 110vac yellow " safety transformers" often assuming these are ITs. But in this case what they are are centre tapped 110 secondaries , with the centre tap connected to the incoming earth wire , this has the characteristic of effectively reducing the shock to earth , to 55vac , a voltage non considered lethal. These units and the associated 110vac power tools are mandatory on building sites. ( we have paradoxically a lot of 110 vac tools in high end tool shops ! )

Hence in this case reducing the shock voltage was considered more important that simply floating the 110 output , even though arguably , in that case no earth return shock path exists at all. I believe however ground referencing is done on this situation , because of the risk of common mode voltages or devices interconnected to multiple floating ITs. , this is why the vast majority of wall warts are earth referenced. ( and cause dammed earth loop that cause interference issues )

However a boats IT is a single item, so the interconnect issue ( common mode voltage) doesn't arise.

Dave

[sailorchic34]

Originally Posted by goboatingnow

As pointed out L1 to L2 faults are easily addressed by appliance or socket based RCDs. And L1 L2 fault in a single electrically continuous appliance will merely short and trip a breaker.

The conundrum. Is that using a centre tap ( and a centre tap protective wire ) , you can establish a fault path that inadvertently adds ground/earth.
Dave

I would NOT connect the center tap to earth (or to primary ground, sort of pointless there). It would be a floating reference to 0V or 55V as the case may be. This is what I think about when I wash dishes.... Silly girl I know.

Of course the breaker will trip when the current pulled through the loop exceeds the current rating of the breaker.

Hum. So with a isolated transformer and a RCD installed on the secondary, how would the device trip. The two legs will have balanced current flow as there is no ground loop to leak current away. I must be extra blonde today....

The RCD requires an unbalanced load between the two legs to trip. With no ground loop to conduct the 5 or 30 ma, how does the RCD trip. Not saying it can't happen, just want to know how you see it happening.

[goboatingnow]

Originally Posted by sailorchic34

I would NOT connect the center tap to earth (or to primary ground, sort of pointless there). It would be a floating reference to 0V or 55V as the case may be. This is what I think about when I wash dishes.... Silly girl I know.

Of course the breaker will trip when the current pulled through the loop exceeds the current rating of the breaker.

Hum. So with a isolated transformer and a RCD installed on the secondary, how would the device trip. The two legs will have balanced current flow as there is no ground loop to leak current away. I must be extra blonde today....

The RCD requires an unbalanced load between the two legs to trip. With no ground loop to conduct the 5 or 30 ma, how does the RCD trip. Not saying it can't happen, just want to know how you see it happening.

Yes but the problem in running a " neutral" to boat appliances is inevitably there is a risk of inadvertent earth referencing , by wet bilges , contact with machinery etc.

Socket based RCDs will specifically deal with the fault describes as a L1 fault to appliance A and L 2 fault to appliance B , note I said appliance of socket RCDs. Not a RCD on the secondary output point of the IT

Dave

[Oceanride007]

So I can see your point now with individual low trip RCD on each circuit, its clumsy but would work. I would still hesitate unless it can be done behind the panel. Its a little hard to swallow after accepting the simplicity of IT.
My neutral earth idea. (Matched resistors between the 2 hots) may be defeated because of Flying cloud suggestion that it would be earthed thru the battery charger or inverter. Yesterday I was working on a boat that had the common RCD after the inverter, I measured 2.5 Volts between ac and dc earth, therefore I think the AC earth is not connected, trouble is without a diagram, I don’t know what I got.

[FlyingCloud1937]

Originally Posted by Oceanride007

So I can see your point now with individual low trip RCD on each circuit, its clumsy but would work. I would still hesitate unless it can be done behind the panel. Its a little hard to swallow after accepting the simplicity of IT.
My neutral earth idea. (Matched resistors between the 2 hots) may be defeated because of Flying cloud suggestion that it would be earthed thru the battery charger or inverter. Yesterday I was working on a boat that had the common RCD after the inverter, I measured 2.5 Volts between ac and dc earth, therefore I think the AC earth is not connected, trouble is without a diagram, I don’t know what I got.

I still think most miss my point. In the USA we work against "potential fault" because potentials seem to happen and then someone loses a life.

I have a boat(customer) right now that has a floating IT/Step-up with a Euro 230v range. When plugged into shore all seems to work well, no tingler.

As soon as the shore side is disconnected, and the range is powered by the Diesel-Gen going thru the same IT/Step-Up, touching any part of the range/cooker it's a tingler.

So we have a fault, ie double fault. It's the only load on that circuit, it's wired correctly.

We have a potential that is an actual working fault. go figger

Lloyd

[sailorchic34]

Hum, I'm assuming someone has checked the AC/DC bonding of the generator and the range bonding. My first thought is the genny is got hot some how attached to the bonding of the range, bypassing the IT. The genny alternator AC ground is connected to DC ground and if its bonded, then also to the range.

Remove the bond wire to the range and see if the tingles go away. Actually check for voltage from the disconnected bond wire and the range with the genny running and shore power disconnected.

Second thought, I assume shore power is removed, IE plug disconnected from shore and not just the AC switch turned to gen.

If the AC wiring is correct, then those pesky electrons may be coming in on the bonding,with shore ground connected its pulled down to 0V. Disconnect shore and earth goes away. Or long shot, maybe from a ground loop from an adjacent vessel. Lord knows there's enough wonky wiring out there.

Originally Posted by FlyingCloud1937

I still think most miss my point. In the USA we work against "potential fault" because potentials seem to happen and then someone loses a life.

I have a boat(customer) right now that has a floating IT/Step-up with a Euro 230v range. When plugged into shore all seems to work well, no tingler.

As soon as the shore side is disconnected, and the range is powered by the Diesel-Gen going thru the same IT/Step-Up, touching any part of the range/cooker it's a tingler.

So we have a fault, ie double fault. It's the only load on that circuit, it's wired correctly.

We have a potential that is an actual working fault. go figger

Lloyd