Galvanic Isolators

[WotEver]

6 minutes ago, alan_fincher said:

I thought most site transfomers were not isolation transformers, but were auto-transformers where the primary and secondary were actually the same winding.

Unless I'm wrong this isn't going to work.

Besides, once you've bought two of those you might as well spend just a little more and buy a single Airlink IT which is the same size as one of them and does the job properly. 

Also, what would you do with the two centre-tapped earths on the two 110V trannies? It's a non-starter.  

[Gardner]

Check out www.airlinktransformers.com. They make these to order. Not cheap but a better option than a GI.

[WotEver]

10 minutes ago, Gardner said:

Check out www.airlinktransformers.com. They make these to order. Not cheap but a better option than a GI.

It's been mentioned a few times now in this 10 year old thread:

http://www.airlinktransformers.com/boating_transformers/BT3231/

[RD1]

2 hours ago, alan_fincher said:

 

I thought most site transfomers were not isolation transformers, but were auto-transformers where the primary and secondary were actually the same winding.

Unless I'm wrong this isn't going to work.

Most CE approved 230/240 volt to 110/115 volt isolation transformers should be isolated, in to out, in fact the output winding is centre tapped to the incomming earth, so the output is still say 110 volts, but +/- 55 volts to earth. So if you stood in a bucket of water, you are at risk from 55 volts ac not 110, and not 230 volts.

Combining the outputs from two of these transformers might give you 230/240 volts but you will have to remove the centre tap of each transformer from earth, and thus affecting the insulation breakdown voltage across each output winding. Any of the output windings would normally only have to sustain 55 volts to earth, but now, one of the windings could see 230/240 volts on the output to the disconnected earth. You can virtually guarantee that these transformers have not been tested or even designed for that. You might want to do an insulation "flash test" on them, but all adds to the cost.

So do not combine the outputs of two industrial isolation transformers. 

I hope I have made it clear.

Richard

[alan_fincher]

14 minutes ago, RD1 said:

Most CE approved 230/240 volt to 110/115 volt isolation transformers should be isolated, in to out, in fact the output winding is centre tapped to the incomming earth, so the output is still say 110 volts, but +/- 55 volts to earth. So if you stood in a bucket of water, you are at risk from 55 volts ac not 110, and not 230 volts.

Combining the outputs from two of these transformers might give you 230/240 volts but you will have to remove the centre tap of each transformer from earth, and thus affecting the insulation breakdown voltage across each output winding. Any of the output windings would normally only have to sustain 55 volts to earth, but now, one of the windings could see 230/240 volts on the output to the disconnected earth. You can virtually guarantee that these transformers have not been tested or even designed for that. You might want to do an insulation "flash test" on them, but all adds to the cost.

So do not combine the outputs of two industrial isolation transformers. 

I hope I have made it clear.

Richard

Yes, very clear, thank you.

In fact as soon as I looked for more info, I could see tat this transformer must be isolate betwen primary and secondary, so my comment doesn't, of course, apply.

However I think we are all agreeing that any attempt to combine the outputs from two would be highly unwise!

[matty40s]

And I do like the fact that the search engine is bringing threads like this back to life, the previous search engine was awfully difficult to use beyond a few months back.

Fascinating stuff, I remember being enthralled by this several years before I actually plunged in and joined.

[ditchcrawler]

How about if you disconnect the centre earth but then connect them in series 180 deg out of phase

[Mikexx]

5 hours ago, ditchcrawler said:

How about if you disconnect the centre earth but then connect them in series 180 deg out of phase

 

That was my thought. And continuously rated at 3.6kVA for less than £100.

Most transformers of this type have a lid priding access to the connections so it may be possible to remove the centre earth tap. Ideally you would like to maintain a land earth to the laminations to catch the consequences of most fault conditions.

[Chewbacka]

9 hours ago, Mikexx said:

 

That was my thought. And continuously rated at 3.6kVA for less than £100.

Most transformers of this type have a lid priding access to the connections so it may be possible to remove the centre earth tap. Ideally you would like to maintain a land earth to the laminations to catch the consequences of most fault conditions.

If you should overheat the transformer causing a short from the primary to the laminations and then continue to run hot such that the secondary overheats to short onto the laminations, then without an earth the shore mains is now connected to the boat electrical system without an earth. This is very bad.

In my opinion, to run a transformer without an earth is not unwise is it reckless.

 

Added - Also if you bond the neutral to earth at the transformer output and the boat RCD trips then the boat hull will become live via the neutral side of the transformer.  You are now relying on the shore supply bollard to trip to avoid killing anyone that touches the boat whilst stood on the bank.

Never remove an earth if you do not fully understand the consequences of faults that may happen.

Edited March 3, 2017 by Chewbacka

[WotEver]

15 hours ago, ditchcrawler said:

How about if you disconnect the centre earth but then connect them in series 180 deg out of phase

 

10 hours ago, Mikexx said:

That was my thought. 

Go for it

[nicknorman]

Foolishly trying to get away from "how to kill yourself with transformers" and vaguely back on topic, it's worth being aware that there is an organisation in the USA known as ABYC which has a spec for GIs that includes a need to be able to withstand a 5000A transient and to not go open circuit on overload. If you want to comply with the RCD / relevant ISO you need a GI that meets this spec. Plenty don't,

Edited March 3, 2017 by nicknorman

[mross]

Nick, does RCD recognise a US standard?  

[nicknorman]

15 minutes ago, mross said:

Nick, does RCD recognise a US standard?  

Not specifically, but the ISO and the AYBC standards are in a similar vein. Give me a moment and I'll pull out what the ISO says...

OK it says "failure of the GI shall not result in an open circuit" which is one of the requirements of the AYBC standard. GIs that don't mention this issue probably don't comply. For example Sterling sell GIs which don't mention it, and GIs that meet the AYBC requirements. The latter are more expensive, of course. I have to say that whilst Safeshore seems a popular choice, as far as I can see they don't meet the ISO or the AYBC and have very limited information - nothing about the max short circuit surge current which, even for a 16A supply, can be a hell of a lot more than you might think. And let's hope their website doesn't reflect the professionalism of their products!

Edited March 3, 2017 by nicknorman

[Mikexx]

17 minutes ago, nicknorman said:

Foolishly trying to get away from "how to kill yourself with transformers" and vaguely back on topic, it's worth being aware that there is an organisation in the USA known as ABYC which has a spec for GIs that includes a need to be able to withstand a 5000A transient and to not go open circuit on overload. If you want to comply with the RCD / relevant ISO you need a GI that meets this spec. Plenty don't,

Perhaps you can explain what a GI is? There is a US term of a GFCI ( Ground Fault Circuit Interrupter) certainly. Most UK MCBs, RCDs and RCBOs will have a fault breaking current capability of 5,000A. There are others for industrial environments that have a greater breaking current spec.

Sorry, but such a spec has no relevance to a transformer. Nor a US specification or requirement for a UK directive.

[mross]

GI stands for galvanic isolator.  Read all about it http://www.tb-training.co.uk/MarineE11.html#Galvanic Isolators (Zinc Savers)

[WotEver]

 

Just now, mross said:

GI stands for galvanic isolator.  Read all about it http://www.tb-training.co.uk/MarineE11.html#Galvanic Isolators (Zinc Savers)

And an isolation transformer (IT) is an alternative (many but not all would say superior) method of achieving the same galvanic isolation.

It has nothing whatsoever to do with circuit or personal protection and most certainly shouldn't add danger such as a pair of bodged site transformers might. 

[nicknorman]

3 minutes ago, Mikexx said:

Perhaps you can explain what a GI is? There is a US term of a GFCI ( Ground Fault Circuit Interrupter) certainly. Most UK MCBs, RCDs and RCBOs will have a fault breaking current capability of 5,000A. There are others for industrial environments that have a greater breaking current spec.

Sorry, but such a spec has no relevance to a transformer. Nor a US specification or requirement for a UK directive.

No, my post had nothing to do with transformers. But the thread topic is Galvanic Isolators which was the topic of discussion before some mad scientists wanted to connect a string of transformers in series to create life in a piece of meat. Or was it the other way round?

A GI is not anything to do with an RCD, GFCI etc. It is merely a means of reducing or hopefully eliminating galvanic corrosion caused  by the incoming shore earth being at a slightly different voltage from the boat's hull / surrounding water. But it is inserted in the shore earth lead so if it fails open circuit,  the boat is no longer earthed but nor is that obvious until there is a fault. So the rule with GIs is that if ever there is a breaker trip, it is strongly advised to check that the GI is still working properly and neither short circuited (making it useless) or open circuit (making it dangerous).

[WotEver]

9 minutes ago, Mikexx said:

Sorry, but such a spec has no relevance to a transformer. Nor a US specification or requirement for a UK directive.

But it has every relevance to this thread, which is titled "Galvanic Isolators"

[Mikexx]

17 minutes ago, nicknorman said:

No, my post had nothing to do with transformers. But the thread topic is Galvanic Isolators which was the topic of discussion before some mad scientists wanted to connect a string of transformers in series to create life in a piece of meat. Or was it the other way round?

A GI is not anything to do with an RCD, GFCI etc. It is merely a means of reducing or hopefully eliminating galvanic corrosion caused  by the incoming shore earth being at a slightly different voltage from the boat's hull / surrounding water. But it is inserted in the shore earth lead so if it fails open circuit,  the boat is no longer earthed but nor is that obvious until there is a fault. So the rule with GIs is that if ever there is a breaker trip, it is strongly advised to check that the GI is still working properly and neither short circuited (making it useless) or open circuit (making it dangerous).

 

It would help tremendously if you didn't need to resort to ridicule in an attempt to cover a lack of knowledge.

I was asking what a GI was in the context of a US specification. Sorry if that wasn't obvious.  GFCIs are still commonly used in the US which measure the voltage difference between a supply earth and a local earth. Their application to boats and supply disconnection in the case of a fault has been discussed here before.

I am very aware of what a Galvanic Isolator does and the current it must with withstand and for how long.

The issue I have with Galvanic Isolators is the earth currents that flow from suppression within equipment and the consequence they have on the potential between boat and shore. Nuisance tripping is quite common with RCDs, more so with computer and other electronic equipment, though heaters such as those in washing machines which employ a hygroscopic oxide for an insulator are often a culprit when left in damp environments for extended times.

If you want both safety and galvanic isolation then the isolating transformer and RCD boat side is the route to go as per:

http://www.smartgauge.co.uk/nb_ac_sys.html

What does surprise me is no one advocates a protection device where the boat and shore earths start to deviate. The ideal candidate for an ELCB (aka US GFCI) between the two 'earths' to isolate the transformer supply side in the unlikely event of a fault. Perhaps it indicates that isolation transformers rarely fail with dangerous consequences.

[WotEver]

12 minutes ago, Mikexx said:

I was asking what a GI was in the context of a US specification. Sorry if that wasn't obvious.

It wasn't at all obvious no, not when you started conflating GFCIs into your post and stating that it had no relevance to transformers. You were demonstrating your total ignorance of the subject.

A 'GI in the context of a US specification' is exactly the same thing as it is in the context of the European specification - namely a Galvanic Isolator as per the title of this thread. 

[nicknorman]

28 minutes ago, Mikexx said:

 

It would help tremendously if you didn't need to resort to ridicule in an attempt to cover a lack of knowledge.

I was asking what a GI was in the context of a US specification. Sorry if that wasn't obvious.  GFCIs are still commonly used in the US which measure the voltage difference between a supply earth and a local earth. Their application to boats and supply disconnection in the case of a fault has been discussed here before.

I am very aware of what a Galvanic Isolator does and the current it must with withstand and for how long.

The issue I have with Galvanic Isolators is the earth currents that flow from suppression within equipment and the consequence they have on the potential between boat and shore. Nuisance tripping is quite common with RCDs, more so with computer and other electronic equipment, though heaters such as those in washing machines which employ a hygroscopic oxide for an insulator are often a culprit when left in damp environments for extended times.

If you want both safety and galvanic isolation then the isolating transformer and RCD boat side is the route to go as per:

http://www.smartgauge.co.uk/nb_ac_sys.html

What does surprise me is no one advocates a protection device where the boat and shore earths start to deviate. The ideal candidate for an ELCB (aka US GFCI) between the two 'earths' to isolate the transformer supply side in the unlikely event of a fault. Perhaps it indicates that isolation transformers rarely fail with dangerous consequences.

Perhaps there is a fine line in your mind between humour and ridicule. In my mind, it is quite clear. Oh and could you be more precise as to the "lack of knowledge" I was trying to cover up? Or was it just a throwaway insult?

Anyway if you know what you are talking about, I don't understand why you are trying to link a discussion on GIs with ECLBs/GFCIs. They are completely different things with completely different purposes and aims. A GI is a GI whether it's in the USA or the UK, also sometimes known as "zinc savers". If you are "very aware of what a GI does" then why did you ask what it was? To be honest I'm struggling to understand what you are getting at.

Yes there can be some very small earth currents due to suppression equipment but it's AC and also, more recent GIs have a capacitor to sink those currents away to shore earth. Modern equipment such as phone chargers no longer have an Earth pin, therefore no longer dump supression currents to earth. I'm pretty sure that there are rules for anything CE marked that put a severe limit on any earth current. This talk of earth suppression currents being an issue stems from the early days of SMPSUs and is largely irrelevant these days.

i don't want a device that disconnects shore earth if there is a slight difference between it and local water "earth". That would be a complete pain in the arse. What I want is something that prevents a small potential difference from causing corrosive currents to flow. A GI does that unless the potential is large (and a GI with monitoring indicates when that happens). An IT of course does the same. Whether you want an IT or a GI is personal choice, the former is more expensive, heavy, bulky, can make some noise and is not 100% efficient. The latter is cheaper, small and light, noise-free and 100% efficient. However it can be argued that it is not as robust as an IT. As I said, a matter of personal choice.

Edited March 3, 2017 by nicknorman

[WotEver]

As for Gibbo's theoretical leakage currents, they remain just that; theoretical. 

Several folk on here have measured them in the past with all sorts of equipment connected and they have never exceeded the bias of a couple of diodes, namely 1.2V. Not even got close. 

Personally I believe that an IT is a superior solution if it is used correctly which includes its placement close to but remote from the boat and hence I usually recommend the Airlink IT which can be yours for £240. I would never recommend a bodge with butchered site transformers. 

A GI which meets the AYBC specification (and many do if you look for it) will be closer to the price of an IT anyway. 

[ditchcrawler]

I was under the impression that an Isolation Transformer should be mounted shoreside for maximum safety/protection so that all cables on the boat issolated from the mains supply.

[WotEver]

Just now, ditchcrawler said:

I was under the impression that an Isolation Transformer should be mounted shoreside for maximum safety/protection so that all cables on the boat issolated from the mains supply.

Absolutely so yes

[nicknorman]

20 minutes ago, ditchcrawler said:

I was under the impression that an Isolation Transformer should be mounted shoreside for maximum safety/protection so that all cables on the boat issolated from the mains supply.

Yes. Only trouble is you may then create a trip hazard or end up in hospital with a bad back! Nothing is 100% risk free! At our marina with its long but relatively narrow pontoons, putting an IT on the pontoon would be extremely antisocial and I'm sure you would be asked (told!) not to.