Galvanic Isolators

[chris w]

you must be living on another planet Get some real qualifications

[Gibbo]

you must be living on another planet Get some real qualifications

 

Not even worthy of comment.

 

Gibbo (BSc)

[chris w]

If you want to test your theory on the boat with apparent 300V RFI voltages try the following. Buy or borrow a second Galvanic Isolator and wire it in series with the original one.

 

Then wire an LED (with a 50 ohm resistor in series with it) across the pair of GI's. Then wire a second LED across the pair of GI's (again with a 50 ohm resistor in series) BUT with opposite polarity.

 

A DC galvanic voltage greater than about 2.4v will cause one of the LEDs to light. An AC voltage greater than about 2.4v peak (~1.7v rms [assuming a sine wave]) will illuminate both LED's.

 

The LEDs will be protected against excess voltage by the clamping action of the GI diodes regardless of the input voltage. However, they will only be protected against excess current (to about 35mA) by the series resistor up to an AC level of about 6v rms (~9v peak). Increasing the series resistor to 100 ohm would double the current protection at the expense of a dimmer LED on excess galvanic action.

 

At least this way you will know for sure whether you have a problem.

 

Chris

[Grasshopper]

I suspect that other members of the forum may be losing the will to live after reading this level of detail.

 

 

Errr - around about page 2 in my case..........

 

But I've been forcing myself to try and follow you as it has real world implications fow what I do.

 

Hi by the way. I'm new here and have recently bought my first narrowboat. Small and ALUMINIUM. Currently it has no shore power circuit and I've no interest in fitting 240V appliances even if there was room. The attraction of the waterways is getting away from such things in my case. Obviously I'm not gong to live on the boat.

 

But the marina does have the availability of shorepower and I was thinking - hey - that would be useful to keep the battery charged whilst I'm away. I was aware of galvanic corrosion but was glibly thinking a GI would be the solution.

 

Then I started reading this....................

 

So could I just interupt the debate for a moment with some practical questions?

 

I accept that an Isolating Transformer is a safer solution but even allowing for the high cost for my limited needs, I don't think that I could physically fit one in. Unless there are any smaller ones than I've seen advertised.

 

If I'm following you correctly, those of you who think that the GI does not provide adequate protection see this happening in two potential circumstances:

 

1. The use of switched mode devices. Query: Would a standard marine battery charger be a switched mode device? And if so would it in itself (given no other AC equipment) be enough to make the GI conduct?

 

2. Where the diodes in the GI itself fail in conductive mode. Query: Do you have a view as to the probability of that happening? Is it less likely with a higher quality GI? Earlier there was mention of GI's to ABYC standards that have a capacitor wired in parallel to the diodes. Would one of these reduce the risk?

[Keeping Up]

The LEDs will be protected against excess voltage by the clamping action of the GI diodes regardless of the input voltage. However, they will only be protected against excess current (to about 35mA) by the series resistor up to an AC level of about 6v rms (~9v peak). Increasing the series resistor to 100 ohm would double the current protection at the expense of a dimmer LED on excess galvanic action.

The idea of using LEDs in this way seems a good one, certainly much easier and safer (for a layman in particular) than hooking up a scope with a floating earth. But it does need a little bit of further development.

 

The statement here about protection against excess current is complete rubbish - "I think you are getting a little confused Chris about currents and voltages". The GIs will limit the voltage across the LED/resistor combination to about 2.4 volts, and thereafter will carry any additional current themselves.

 

2.4v is not guaranteed to be enough to light the LEDs properly. The forward conducting voltage of many LEDs is close to this figure and they may turn on only for a fraction of the peak of the cycle (and therefore look very dim) or may not turn on at all; for reliable results you would need to use 3 GIs (giving 3.6v); you could use a series resistor of 50 ohms to give a peak current of 20-25mA.

 

Rather than using a separate resistor for each LED it is better practice to use a single resistor, and to put the two LEDs across each other (in opposite directions) so that each will clamp the reverse voltage applied to the other. Your configuration with two resistors will be OK in this circuit because of the clamping effect of the GIs, but in other applications offers no reverse voltage protection and the application of 6v rms (9v peak) as you describe would immediately destroy both LEDs.

 

What we need is an LED with a forward conduction voltage of less than 1.2 volts, but I don't know of one. I wonder if you could use a couple of complementary transistors to sense the voltage across the GIs and drive the LEDs? Maybe it's worth making up a little box containing 2 transistors with resistors in the base connection, two LEDs with resistors, and possibly a pair of PP3 batteries (one for the PNP transistor and one for the NPN transistor). plus two long leads with probes to touch across the GI - I shall go away and scribble on the back of an envelope to see how it would look, then may go down to the boat and try it out

 

Allan

[Keeping Up]

Welcome to the debate Grasshopper. Please help us make sure we stay in touch with the real world and don't drift too far into the theoretical (it's dificult keeping your feet on the ground when that ground may be live!)

 

Query: Would a standard marine battery charger be a switched mode device? And if so would it in itself (given no other AC equipment) be enough to make the GI conduct?

It may or may not be. And it may or may not do so.

 

Certainly a cheap charger would not be a switched mode device, but most cheap chargers are not intelligent enough to avoid cooking your battery while you are away.

 

Where the diodes in the GI itself fail in conductive mode. Query: Do you have a view as to the probability of that happening? Is it less likely with a higher quality GI? Earlier there was mention of GI's to ABYC standards that have a capacitor wired in parallel to the diodes. Would one of these reduce the risk?

I believe that they are very unlikely to fail in conductive mode although I don't have any figures to support this belief. They are much more likely to fail to an open-circuit condition, which is their most worrying feature because it is a condition that would not be noticed for some time but would leave your boat without an earth connection. It is a rare occurrence but I wish I knew a solution to it (other than using an isolating transformer which is undoubtedly the best solution except that they are very big and very expensive)

 

How can you assess what is a high-quality GI? Not necessarily by the price!

 

Gibbo ended up in an extremely long and very technical debate on uk.rec.waterways about the merits or otherwise of putting a capacitor across a GI. It wouldn't significantly reduce the risk of a failure, but there is an outside chance it may reduce the level of RF current passing through the GI, although theory says it's pointless unless you can experiment on your own particular setup until (and if) you find something that works. This would mean that you'd need a tester of some sort.

 

With my own setup on Keeping Up, the only device left connected for any period of time is the battery charger, which I know is a switched mode device. As a temporary feature I have connected its earth wire to the shore side of the GI, a situation which is fine unless the GI fails open-circuit in which case the charger would be earthed and the boat wouldn't, which would be potentially lethal if I hadn't placed the charger inside a (vented) wooden box that you cannot remove without taking out the mains connection. Maybe when all the tests are completed I'll wire it differently

 

Allan

[John Orentas]

Hi Grasshopper.

 

Life really is worth living so please don't give up on it, let us not lose sight of where all this started. There are some people who lie awake at night with images of millions of iron atoms detaching themselves from their hull which is currently preventing their duvet from getting wet, swimming a few yards along the canal and attaching themselves to the neighbouring boat thus enhancing it at their expense.

 

The vast majority of all corrosion or erosion of a steel boat is a result of chemical action, the fact that a voltage can be measured here and there is largely incidental and if your pride and joy is coated in a layer of paint the steel will last virtually forever, they do say that if you were to take one probe of a multimeter a 1,000 metres up in the air you will measure 100 volts or so. But so what, who cares, what effect does it have, probably none.

 

Reading some of these threads we might get the idea that if we were to connect torch battery to a lump of steel we could all stand there and watch it disappear into the ether in a day or two. Years ago I used to visit electroplating companies in my area as a potential customer, what stuck in my mind at the time was the massive amount of electrical power that was required for so little effect, to coat a small component with a few thou' of nickle would take a couple of days and the output of the local substation. A couple of incidental volts won't do much.

 

Those 'Switch mode power supplies' someone has persuaded himself that they are something new and revolutionary. I was using them twenty years ago, they are a more efficient and lightweight alternative to the use of transformers but the problem with them is that the are extremely 'noisy' in terms of radiation, I do remember that the navy would not have them on their ships in any form for that reason alone.

 

As I have mentioned before the first rule of the installation of mains electrical systems as any electrician learns on his first day at college is that all exposed metallic components should be earth bonded. No if's or but's. Bonded to a true earth no-one ever said "but you can put a couple of diodes in series with it if you want". If you want RCD or other protection devices to work earths must be properly bonded.

 

And what of isolation transformers, quite commonly used in industry they are simply a transformer with 230 volts in and 230 volts out with no connection between, someone rather missed the point, those yellow units that are used on building sites are isolation transformers but their safely does not come from that, they output only 110 volts, thats what makes them safe.

 

So if we want to fit large, rather expensive transformers in our boats, well why not, but let us not pretend they enrich our lives. If we want to fit a rather less expensive plastic box with a couple diodes in it albeit a bit dodgy on safety grounds again why not, but I have a suspicion that the £90 would be better invested in a gallon or two of bitumastic paint.

[bottle]

Gentlemen

 

Thank you for your discussions, fascinating, cannot say I kept up, my thoughts are that, the problem of galvanic corrosion is possibly only for those that are connected to the 'mains' for long periods of time in a marina.

 

As John says maybe a thicker coat of bitumen would be just as good (especially on the water line.)

 

Once again I have enjoyed your discussions and think I may have learnt/understood more about the perceived problem.

[chris w]

I have a suspicion that the £90 would be better invested in a gallon or two of bitumastic paint[

 

Don't forget to paint over your anodes then cos they will disappear fast!!!

 

Galvanic corrosion is very real and can be a very fast acting agent for people plugged continuously into shore power. If you're not a regular shore mains user then it won't affect you.

 

If you want RCD or other protection devices to work earths must be properly bonded.

 

Actually, the beauty of an RCD is that it detects current imbalance between the Live and Neutral lines, so will work if any imbalance arises even if the diverted current is not to "earth"

 

 

Chris

[Keeping Up]

Actually, the beauty of an RCD is that it detects current imbalance between the Live and Neutral lines, so will work if any imbalance arises even if the diverted current is not to "earth"

Correct point well made Chris. I'll be interested to read Gibbo's response to this (are you there Gibbo?) because on his website he states "They simply do not operate as designed without the neutral-earth bond" which I have always disagreed with. And if the statement is incorrect it could have implications on the conclusions later in the page.

 

Once upon a time we had ELCBs, which actually relied on the current in the earth to trip them. Thankfully they have disappeared now. Our house many years ago had been re-wired by an incompetent amateur, with an ELCB wrongly connected so that if you shorted live to earth it blew the trip which disconnected the neutral and earth, leaving everything (including all the metalwork in the house) live at 240v ac while you stunbled around in the darkness. And yet the Electricity people had passed it as safe!

 

Allan

[chris w]

Allan

 

With reference to your circuit points, you must take into account that forward voltage drops of diodes are not constant but vary with current. The figures quoted typically for diodes (eg: 0.6v for silicon, ~1.8v for regular LEDs) are not fixed. The diodes will start to conduct some time before these values are reached and will have much higher Vf's as current increases so the clamping action is not fixed either.

 

The idea of using a transistor to detect the change occurred to me too (in order to obviate the need for a second GI). The issue would be getting some hard switching to occur based on the rather flaky characteristic of diode forward voltage drops. It's interesting that the one commercial unit to incorporate LEDs, that I have come across, also effectively uses 2 GI's in series. http://www.galvanicisolators.com/

 

Chris

[Grasshopper]

Thanks guys for your answers........I think

 

It's part of the human condition I guess that we crave certainty where none is to be had.

 

In my case, whatever the speed of galvanic corrosion, having paid a premium for an aluminium alloy boat (Sea Otter) which is more resistant than a steel boat to normal corrosion, it just feels wrong to now do something that introduces the problem afresh. Obviously it would be different if I was going to live on it but in my case it would be nothing more than a convenience.

 

Perhaps I'll strike a compromise and just consider the shore power to keep the batteries charged if I'm working on the boat and don't want to run the engine. That way it wouldn't be connected for long periods of time and the problem goes away. It just means that I need to bring the batteries away for recharge at intervals during the winter I guess. In the scheme of things not a big deal since I will be checking on the boat regularly anyway.

 

Anyway, thanks again for your thoughts.

 

Great site by the way so I'll be sticking around.

[chris w]

Grasshopper

 

Also note that even if you're not using shore power, there are other sources of electrolysis that you can't correct. The boats each side of you in the marina may be connected together through the shore ground lead and one may be eating up the anodes rapidly on the other (as described ad nauseum above).

 

If your boat sits between them, this current may take a short cut by going in an item near one boat, and exiting via your anode near the other. This will eat up your anode too even though you are not connected to them. The best solution here us to use a zinc "fish" while you are at the marina. They are large lumps of zinc, often cast in the shape of a fish, that are cheaper and easier to replace than the anodes on your hull.

 

The "fish" come with a copper wire already attached which is also used to hang them in the water. They have an alligator clip on the end of the wire and this should be connected to the negative bonding circuit on your boat. (If it is not conveniently available in the cockpit in the vicinity of the prop, you might consider installing a stainless bolt for clipping it to, with the head of the bolt inside the deck connected to the negative bonding system).

 

Also (ref your Sea Otter) while aluminium is less prone to chemical corrosion than steel, aluminium and aluminium alloys are actually above steel in the galvanic list of metals. ie: aluminium is more susceptible than steel to galvanic corrosion.

 

Chris

Edited November 11, 2006 by chris w

[Audi Fan]

I know i risk getting caught in the crossfire, but while you're reloading can i ask another question?

 

If a 110v transfomer is £53 from Screwfix, Why are the isolating transformers so expensive?

surely you'd get the same effect by connecting two of these together?

 

[alan_fincher]

I know i risk getting caught in the crossfire, but while you're reloading can i ask another question?

 

If a 110v transfomer is £53 from Screwfix, Why are the isolating transformers so expensive?

surely you'd get the same effect by connecting two of these together?

 

I believe such things are generally what I used to know as an "autotransformer", and have only one winding in them that is shared by input and output circuit, (the latter being tapped off from half way along the winding, to provide the reduced voltage).

 

So they don't isolate the output from the mains at all.

 

Edit

 

Yep, didn't think to Google for it, before posting - Silly!

 

Try this link, for example...

 

Link to Wilkipedia

Edited November 11, 2006 by alan_fincher

[chris w]

Re price of isolation transformers:

 

I would guess for 2 reasons:

 

1. Market volumes are significantly lower for isolation transformers, I would guess, than 240-110v transformers.

 

2. Most things are ultimately priced on the price the market will bear. Anything involving boats carries a premium it seems. Compare even buying a simple thing like a nice, highly polished brass cabin hook from a chandler and from Screwfix (around £7.50 in a chandler and £1.49 from Screwfix - identical product).

 

Chris

[John Orentas]

I believe such things are generally what I used to know as an "autotransformer", and have only one winding in them that is shared by input and output circuit, (the latter being tapped off from half way along the winding, to provide the reduced voltage).

 

So they don't isolate the output from the mains at all.

 

Edit

 

 

 

 

 

Just consulted the RS Book.

There is a range of Safety Site Transformers. 230v input, 230v output example : 1650 VA

Double wound with safety screen between winding. BSEN 61558-2-23 all fitted in a nice enclosure.

 

£80.00 each and this supplier is always the most expensive.

 

There is also a range of 110v output types. £60 to £80

[chris w]

John

 

Do you have the RS part number?

 

cheers

 

Chris

[GUMPY]

Just consulted the RS Book.

There is a range of Safety Site Transformers. 230v input, 230v output example : 1650 VA

Double wound with safety screen between winding. BSEN 61558-2-23 all fitted in a nice enclosure.

 

£80.00 each and this supplier is always the most expensive.

 

There is also a range of 110v output types. £60 to £80

 

Be vary careful with these

 

The 1650va will be a "tool" rating not continious, they will be about 1KW continious or 4 amps.

 

Allthough they are isolating transformers they will also have a centre tapped earth on the secondary so you will always have 125v betwen "neutral"and earth and "live" and earth, all site transformers are built this way.

 

Julian

[John Orentas]

John

 

Do you have the RS part number?

 

cheers

 

Chris

 

 

The 1650 unit I mentioned; model CM30002/230

 

RS 436-8805 .................... £80.00

 

I used to buy specialised transformers for various purposes, I often found that I could have a transformer manufactured to my own specification at lower cost than the proprietary types.

[chris w]

I just checked the rating and, in fact, the 1600VA is continuous. The tool rating is 3000VA.

 

see http://rswww.com/cgi-bin/bv/rswww/searchBr...e&Nr=avl:uk

 

Chris

[Audi Fan]

ooh i was so close!

 

I'm sure the 'centre tapped earth' could easily be modified in favor of a neutral-earth bond

[John Orentas]

I just checked the rating and, in fact, the 1600VA is continuous. The tool rating is 3000VA.

 

see http://rswww.com/cgi-bin/bv/rswww/searchBr...e&Nr=avl:uk

 

Chris

 

Surely you would not use any rating other than the 'Coninuous' one for use on a boat.

Edited November 11, 2006 by John Orentas

[GUMPY]

I just checked the rating and, in fact, the 1600VA is continuous. The tool rating is 3000VA.

 

 

Indeed that is correct

I hadnt looked, However at 1600VA its still 2000VA less than you should have for a 16amp supply.

It would just about cope with the battery charger here, if the charger was running at anything other than float.

 

J

[John Orentas]

Going off topic just a little I am always a little surprised why some of you chaps don't go for a 110 volt system or a 2 phase 230 volt (110 - 0 -110) system as is common in many other countries.

 

Though it is not in any way common, domestic electrocution only happens in the UK. Single phase 230 volts, the product of a diseased brain.